HSCI 180 · Lesson 10

Why People Use Substances and Why Risks Differ

Drugs and Society

Learning objectives for this lesson:

  1. Distinguish motives for substance use from risk factors, describe Cooper’s four drinking motives and summarize what reviews have found about the links between coping and enhancement motives and drinking problems.
  2. Classify the reasons for use reported by BC youth in the 2023 BC Adolescent Health Survey and identify the limits of survey data on reasons.
  3. Explain social learning theory, outcome expectancies and the self-medication hypothesis, and evaluate the evidence for each.
  4. Describe the four levels of the social ecological model and define risk and protective factors.
  5. Sort Canadian evidence on risk and protective factors for youth substance use into ecological levels and explain how levels combine.
  6. Distinguish the factors that predict substance use from those that predict harm, and explain why substance use harms are distributed more unequally than substance use, with reference to the alcohol harm paradox.
  7. Describe how poverty, housing, childhood adversity, colonialism, racism, dislocation and criminalization shape substance use harms in Canada.
  8. Calculate and interpret a rate ratio from a group’s share of deaths and share of the population.
  9. Describe the distinct vulnerabilities and needs of youth, older adults, pregnant people, Indigenous peoples, 2SLGBTQ+ people, workers in high-risk trades and people leaving prison, and apply the lesson to a Canadian case.

This course was developed by Dr. Kiffer G. Card.

Lesson 10 · HSCI 180

Why People Use Substances and Why Risks Differ

A short guided walkthrough before you work through the lesson at your own pace.

Drugs and Society
Why this matters

Similar use, very different harm

Two groups can use substances at similar rates and experience very different levels of injury and death.

Explaining that gap requires attention to reasons, relationships, communities and policies as well as pharmacology.

The road map

Four sections

1 · Motives and theories

Why people say they use, and where those reasons come from.

2 · Social ecological model

Risk and protective factors at four levels, with BC youth data.

3 · Social determinants

Why harm falls unevenly, and how to measure the gap.

4 · Populations and needs

Seven groups whose vulnerabilities and needs differ.

Real Canadian evidence

Where the evidence comes from

38,500BC students in the 2023 Adolescent Health Survey
4levels of the social ecological model
7populations with distinct needs

Sources include the McCreary Centre Society, the First Nations Health Authority, the BC Coroners Service and the Public Health Agency of Canada.

Section 1 of 5

Motives and Psychological Theories of Substance Use

⏱ Estimated reading time: 35 minutes
Section 1 of 5

Motives and Psychological Theories of Substance Use

What people say they use for, and four accounts of where those reasons come from.

Cooper’s model

Two dimensions, four motives

Enhancement

Positive and internal: to feel good or excited.

Social

Positive and external: to celebrate or be sociable.

Coping

Negative and internal: to forget worries or sadness.

Conformity

Negative and external: to fit in or avoid rejection.

Four-factor model (Cooper, 1994), building on an earlier motivational model (Cox & Klinger, 1988).

What motives predict

Coping and enhancement predict problems

Social

Most commonly reported; linked to moderate drinking.

Enhancement

Linked to heavy drinking and to drinking problems.

Coping

Reported by few; linked to drinking problems.

A review (Kuntsche et al., 2005) and a meta-analysis (Bresin & Mekawi, 2021); few of the studies followed people over time.

What BC youth say

Experimenting, fun and friends

53%wanted to experiment
46%wanted to have fun
30%said friends were doing it

Youth who used only cannabis, 2023 BC Adolescent Health Survey. Among youth who used daily, 51% cited stress (Smith et al., 2025).

Social learning theory

Use is learned from others

Differential association

Learning values and techniques from family, friends and peers.

Definitions

The attitudes a person attaches to the behaviour.

Differential reinforcement

The balance of rewards and costs that follows it.

Imitation

Copying behaviour seen in others.

Theory from Akers (1977). A meta-analysis found moderate support in substance use research, strongest for differential association (Kruis et al., 2020).

Two more accounts

Expectancies and self-medication

Outcome expectancies

Beliefs about a drug’s effects predict when young people start drinking and how they drink later, and drinking also changes them (Smit et al., 2018).

Self-medication

Using substances to relieve emotional pain (Khantzian, 1985); reported by about a fifth to a quarter of people with mood or anxiety disorders (Turner et al., 2018).

Causation can run both ways: substances can also cause the symptoms they seem to relieve.

Carry forward

From the person to the context

  • Motives differ by valence and source; reviews link coping and enhancement motives to drinking problems.
  • In the 2023 BC survey, curiosity, fun and friends were the most common reasons for use.
  • Social learning, expectancies and self-medication each have support; research on what comes first is still evolving.

Next: the social ecological model.

Introduction

People use psychoactive substances for many reasons. The First Nations Health Authority’s harm reduction toolkit for friends and family, With Open Arms, states that “substance use is shaped by personal, social, cultural, and historical experiences”, and it lists stress, curiosity, trauma, fun, boredom and peer pressure among the reasons people use (First Nations Health Authority [FNHA], 2025). The list joins pleasant reasons with painful ones and personal reasons with social ones, and a useful account of substance use has to make room for all of them.

Lessons 8 and 9 explained how drugs act on the brain’s reward and learning systems and why some people are more vulnerable to addiction than others. This lesson asks why people use substances in the first place and why the risks that follow differ so much between people and groups. Section 1 begins with the individual: the motives people report and four psychological accounts of where those motives come from. Explaining use is a different task from judging it. Lesson 1 showed that most substance use does not meet the definition of a substance use disorder, and the theories below apply to a glass of wine with friends as much as to daily use of an unregulated opioid. HSCI 214 develops the psychology of stress and mental illness in depth.

Learning Objectives

  • 1.1 Distinguish a motive from a risk factor and explain why motivational models treat motives as the most proximal predictors of use.
  • 1.2 Describe Cooper’s four-factor model of drinking motives (enhancement, social, coping and conformity) and the valence and source dimensions that organize it.
  • 1.3 Summarize what reviews have found about the links between each motive and patterns of drinking and harm, and explain why coping motives may be linked to drinking problems.
  • 2.1 Classify the reasons for substance use reported by young people in the 2023 BC Adolescent Health Survey, explain why survey reasons do not map neatly onto any single model, and identify the limits of survey data on reasons.
  • 3.1 Explain Akers’s social learning theory and its four mechanisms: differential association, definitions, differential reinforcement and imitation.
  • 3.2 Explain outcome expectancies and how the balanced placebo design separates the effects of a drug from the effects of believing one has taken it.
  • 3.3 Evaluate the self-medication hypothesis using Canadian and international epidemiological evidence.

1.1 Why Reasons for Use Matter

Researchers distinguish the reasons a person gives for using a substance from the background characteristics that make use more likely. The two kinds of explanation answer different questions.

A motive is the purpose a person hopes a substance will serve on a particular occasion: to feel good, to relax with friends, to stop worrying or to avoid being left out. A risk factor, which Section 2 examines, is a characteristic of a person or their surroundings that is statistically associated with a higher chance of use or harm across a population. Motivational models treat motives as the most proximal predictors of drinking, the last step through which personality, family history, peers and setting influence whether and how much a person drinks (Sherman et al., 2025). Two people who each drink two beers, one at a celebration and one alone every night to stop feeling sad, consume the same amount, but only the second pattern is tied to distress that returns each day.

1.2 Four Motives for Drinking

A widely used classification of drinking motives was developed by the psychologist M. Lynne Cooper. It organizes motives along two dimensions and produces four types, each with a distinct link to drinking and harm.

Building on the motivational model of alcohol use proposed by Cox and Klinger (1988), Cooper (1994, as described by Sherman et al., 2025, and Ng Tseung-Wong et al., 2026) developed and tested a four-factor model of drinking motives among adolescents. Valence describes whether a person drinks to obtain a positive outcome, such as pleasure, or to reduce or avoid a negative one, such as distress. Source describes whether the outcome is internal, a change in the person’s own mood, or external, a social reward or the avoidance of rejection (Sherman et al., 2025; Ng Tseung-Wong et al., 2026). Crossing the two dimensions gives the four motives arranged in Figure 10.1.

Two dimensions define four drinking motives Internal source (the person’s own mood) External source (other people) Positive valence seek a reward Negative valence reduce or avoid a bad outcome Enhancement to feel good or excited positive and internal Social to celebrate or be sociable positive and external Coping to forget worries or sadness negative and internal Conformity to fit in or avoid rejection negative and external
Figure 10.1. Cooper’s four drinking motives arise from crossing valence (seeking a positive outcome or avoiding a negative one) with source (the person’s own mood or other people). Based on the descriptions of Cooper (1994) in Sherman et al. (2025) and Ng Tseung-Wong et al. (2026).

Table 10.1 defines each motive and gives an illustrative reason. The reasons were written for this lesson and are not items from Cooper’s questionnaire.

Table 10.1. The four motives in Cooper’s model, with the valence and source of each and an illustrative reason written for this lesson.

MotiveValence and sourceWhat the person seeksIllustrative reason
EnhancementPositive, internalMore positive feelings, excitement or fun“I like the buzz.”
SocialPositive, externalSocial rewards, such as celebrating with a group“It makes the party more fun.”
CopingNegative, internalRelief from worry, stress or sadness“It helps me forget a bad day.”
ConformityNegative, externalAvoiding social rejection or disapproval“I didn’t want to be the only one not drinking.”

1.3 What Each Motive Predicts

Motives are measured because they predict different outcomes. In a 2005 review, Kuntsche and colleagues examined fifteen years of research on the drinking motives of adolescents and young adults. Most young people reported drinking for social motives, some for enhancement and only a few for coping. Social motives were associated with moderate drinking, enhancement motives with heavy drinking and coping motives with alcohol-related problems (Kuntsche et al., 2005). A 2021 meta-analysis of 229 studies by Bresin and Mekawi found that enhancement and coping motives were the strongest predictors of drinking problems, and that enhancement motives were the strongest predictor of how much people drank. Few of those studies followed people over time, so the links are consistent across reviews but rest mainly on data collected at one point in time (Bresin & Mekawi, 2021).

The link between coping and problems has a plausible mechanism. Lesson 8 described negative reinforcement, in which a behaviour is strengthened because it removes something unpleasant: drinking to quiet sadness brings relief, the relief makes drinking more likely when the feeling returns, and the underlying problem remains. Enhancement motives are linked to heavy drinking, and Canada’s Guidance on Alcohol and Health states that more than two standard drinks on one occasion is associated with a higher risk of harm to self and others, including injuries and violence (Canadian Centre on Substance Use and Addiction [CCSA], 2023). Kuntsche and colleagues also found that measurement varied widely, with studies grouping 10 to 40 questions into 2 to 10 dimensions, so comparisons across surveys need care.

2.1 What Young People in British Columbia Report

Canadian survey data allow the motivational model to be checked against what young people say about their own use.

The BC Adolescent Health Survey (BC AHS) has been conducted by the McCreary Centre Society about every five years since 1992. In the spring of 2023 it was completed by about 38,500 students aged 12 to 19 in 59 of the province’s 60 school districts; it does not reach students who were absent, attended independent schools or were home-schooled (Smith et al., 2024). The McCreary report Blunt Talk III analyzes the 38,277 participants aged 12 to 18, since cannabis is legal at 19. Because the survey asks about reasons for substance use in general, the report examined reasons among the 5 percent of youth who had used cannabis and no other substance, so that their answers refer to cannabis (Smith et al., 2025). Figure 10.2 shows the results.

Reasons for the most recent use of cannabis, BC youth, 2023 Wanted to experiment53% Wanted to have fun46% Friends were doing it30% Because of stress21% Felt down or sad15% There was nothing else to do6% To manage physical pain5% Thought it would help them focus3% Felt pressured or to fit in3% Didn’t mean to2% Percentage of youth who used only cannabis; each person could choose more than one reason
Figure 10.2. Reasons for their most recent cannabis use given by the 5 percent of BC students aged 12 to 18 who had used cannabis and no other substance, 2023 BC Adolescent Health Survey. Data from Smith et al. (2025), Blunt Talk III.

The three most common reasons were to experiment, to have fun and because friends were doing it. Reasons differed between groups: females were more likely than males to report using because they were stressed (32 percent compared with 15 percent), and youth aged 15 or older were more likely than younger youth to report using because they felt down or sad (17 percent compared with 8 percent) (Smith et al., 2025).

The pattern changes among young people who use often. Most youth who used cannabis daily also used another substance daily, so their reasons refer to substance use in general. Their most common reason was to have fun (75 percent), but 51 percent cited stress, 44 percent feeling down or sad and 34 percent addiction (Smith et al., 2025). Reasons tied to distress are far more common in this group, which is consistent with the link between coping motives and problems. Because the data are cross-sectional, they cannot show whether distress led to frequent use, frequent use added to distress, or both.

Reading survey data on reasons for use

Four features of the BC AHS data shape how they can be read. Young people could choose more than one reason, so the percentages in Figure 10.2 add up to more than 100. The question refers to the most recent occasion, which may differ from a person’s usual reasons. The cannabis analysis uses a small subgroup whose answers may differ from those of youth who used several substances. And the survey’s categories were not designed to match Cooper’s model, an example of the measurement variation that Kuntsche et al. (2005) described.

Use Interactive 10.1 to match six reasons from Figure 10.2 to the motive in Cooper’s model that fits best. Some reasons fit more than one motive, and some fit none.

Interactive 10.1
Match reasons reported by BC youth to Cooper’s motives

Percentages are the shares of BC youth who used only cannabis who gave each reason for their most recent use in 2023 (Smith et al., 2025). Choose “Outside the four motives” when no motive fits well, then check your answers.

“Wanted to have fun” (46 percent)
Enhancement: having fun is a positive change in the person’s own experience.
“Friends were doing it” (30 percent)
Social or conformity. The source is external, but the wording does not reveal the valence: joining in to share a good time is social, and joining in to avoid being left out is conformity.
“Because of stress” (21 percent)
Coping: the person uses to reduce an unpleasant internal state, one of the two motives that reviews link most closely to drinking problems.
“Felt pressured into doing it or to fit in” (3 percent)
Conformity: negative valence and an external source. Few young people chose this reason, despite the common belief that peer pressure drives youth substance use.
“Wanted to experiment” (53 percent)
Outside the four motives is best, and enhancement is defensible. Curiosity is the most common reason in the survey, yet Cooper’s model has no curiosity motive.
“To manage physical pain” (5 percent)
Outside the four motives is best, and coping is defensible. Cooper’s coping motive concerns emotions; relieving a physical symptom is closer to self-medication (3.3 The Self-Medication Hypothesis).

The exercise shows the value and the limits of the motivational model: it sorts most reasons in a way that predicts later harm, but the most common reason in this survey, curiosity, sits outside it. Motives also leave open where the underlying beliefs and habits come from, the question addressed by the three theories that follow.

3.1 Social Learning Theory

People learn what a substance is for, how to use it and whether using it is acceptable from the people around them. Social learning theory describes how that learning happens.

The sociologist Ronald Akers (1977, as described by McCloud & Abdullah, 2025) developed a social learning theory of deviant behaviour that drew on key elements of Edwin Sutherland’s differential association theory and of Albert Bandura’s social learning theory, adding the principle that behaviour is shaped by its rewards and costs (Mui et al., 2014). It proposes that substance use is learned through four mechanisms (Mui et al., 2014), which the flip cards below define.

Differential associationClick to explore
DefinitionsClick to explore
Differential reinforcementClick to explore
ImitationClick to explore

The theory fits several Canadian findings. In the 2023 BC AHS, “friends were doing it” was the third most common reason for recent cannabis use, and use was usually social: among youth who had used cannabis, 59 percent said the cannabis they last used was shared among a group of friends and 35 percent said a friend or family member gave it to them (Smith et al., 2025). In workshops held to discuss the results, some youth said that pressure to fit in and be part of a social group drove vaping, especially among younger students (McCreary Centre Society, 2025). Mui and colleagues (2014) concluded that the more a person associates with others who use a drug, the more ordinary the behaviour comes to seem. The same mechanisms can protect: families and friends who model moderate use or no use teach those patterns too. A 2020 meta-analysis of 83 studies by Kruis and colleagues found moderate support for social learning theory in research on substance use, strongest for differential association, but few of those studies could show whether the associations came before the use (Kruis et al., 2020).

Try it 10.1: Social learning in a real finding

The 2023 BC AHS found that 17 percent of youth who felt their parents or guardians knew what they were doing in their free time most or all of the time had used cannabis, compared with 40 percent of youth who felt their parents rarely or never knew (Smith et al., 2025). In two or three sentences, explain how two of the four mechanisms of social learning could produce this difference. Then name one explanation that does not depend on social learning at all.

3.2 Outcome Expectancies

A second account focuses on what people believe a substance will do for them. These beliefs are learned before most people ever use a substance, and they shape both the decision to use and the experience of using.

An outcome expectancy is a belief about the effects of a substance, such as “alcohol makes me more outgoing”. Expectancies grow out of childhood experiences connected with alcohol, such as watching how adults behave when they drink. Christiansen and colleagues measured alcohol expectancies among students in their seventh and eighth years of school and followed them for a year. Five of seven expectancy scores distinguished students who stayed non-problem drinkers from those who began problem drinking, and expectancies accounted for a large share of the variation in later drinking (Christiansen et al., 1989). Because expectancies were measured before the drinking, the study strengthened the case that expectancies come before drinking and help to shape it. A 2018 systematic review of 43 studies by Smit and colleagues found that expectancies predict when young people start drinking and how they drink later, and also that starting to drink changes expectancies, so the influence appears to run in both directions (Smit et al., 2018).

Lesson 3 introduced expectancy and placebo responses as part of “set and setting”. To separate the pharmacological effect of alcohol from the effect of believing one has drunk it, alcohol researchers use the balanced placebo design, for which a 1973 study by Marlatt and colleagues provided the model (Marlatt et al., 1973, as described by Brockdorf et al., 2025, and Sayette et al., 2012). The design crosses what participants are told with what they receive (Table 10.2).

Table 10.2. The four conditions of the balanced placebo design, which varies the drink a participant receives independently of what the participant is told (Palmer & Brandon, 2019).

Receives alcoholReceives no alcohol
Told the drink contains alcoholDrug effect plus expectancyExpectancy alone (placebo)
Told the drink contains no alcoholDrug effect alone (antiplacebo)Neither (control)

Comparing rows shows the effect of expectancy, and comparing columns shows the effect of the drug. A meta-analysis of balanced placebo studies by Hull and Bond (1986), as summarized by Sayette et al. (2012), concluded that social behaviours were more strongly affected by what people expected than by the alcohol they drank. The design has a practical limit: people given even a moderate dose often notice it, so the antiplacebo condition is hard to carry out (Sayette et al., 2012). It has since been applied to nicotine in vaping research (Palmer & Brandon, 2019). Changing what people expect a substance to do may therefore be one way to change how they use it, a strategy HSCI 402 examines in its discussion of prevention.

3.3 The Self-Medication Hypothesis

The third account links substance use to emotional pain and mental illness. Researchers at the University of Manitoba have contributed population evidence used to test it.

The psychiatrist Edward Khantzian proposed the self-medication hypothesis from clinical work with people dependent on heroin and cocaine. He argued that painful emotional states and related psychiatric conditions predispose people to addiction, and that the drug a person comes to prefer reflects an interaction between the drug’s effects and the feelings the person struggles with: opioids muted rage and aggression, and cocaine relieved distress associated with depression and hyperactivity (Khantzian, 1985).

Population surveys can test whether self-medication is common. Bolton, Robinson and Sareen (2009), working at the University of Manitoba, analyzed a nationally representative survey of 43,093 adults in the United States. Almost one quarter (24.1 percent) of people with a mood disorder reported using alcohol or drugs to relieve their symptoms, the share was highest for bipolar I disorder (41.0 percent), and men were more than twice as likely as women to self-medicate. A later review by the same Manitoba group found that 21.9 to 24.1 percent of people with mood or anxiety disorders reported self-medicating, and that longitudinal data supported a sequence in which the mood or anxiety disorder came first and the substance use disorder followed among people who self-medicated (Turner et al., 2018). The accordion below sets out the strengths and limits of the hypothesis.

What the evidence supportsv

Self-medication is common among people with mood and anxiety disorders, and among them mental illness often comes before a substance use disorder (Turner et al., 2018). In the 2023 BC AHS, 27 percent of youth who felt they managed stress poorly used cannabis to manage stress, compared with 12 percent of those who managed it well (Smith et al., 2025). Clinically, the hypothesis supports treating mental health and substance use together.

What it does not explainv

Most people with a mood or anxiety disorder do not report self-medicating, and in the review and survey described above most young people drank or used for social or enhancement reasons, or out of curiosity.

The problem of directionv

Substances can cause or worsen the symptoms they are said to relieve; Lesson 7 showed that frequent use of high-potency cannabis is associated with psychosis. A report that a person uses to feel better is genuine evidence about motive, but it cannot by itself show that the mental illness caused the substance use.

Table 10.3 compares the four psychological accounts introduced in this section.

Table 10.3. Four psychological accounts of why people use substances, with their central claim, an example of supporting evidence and a main limitation.

AccountCentral claimExample of evidenceMain limitation
Motivational modelPeople use to gain positive outcomes or avoid negative onesCoping and enhancement motives predict drinking problemsCuriosity does not fit the four motives
Social learning theoryUse is learned from others through four mechanismsMost BC youth who had used cannabis last used cannabis shared among friends (2023)Friends may be chosen because they already use
Outcome expectanciesBeliefs about effects shape use and experienceExpectancies predict starting to drink and later drinking (2018 review)Says less about why beliefs differ
Self-medicationPeople use to relieve emotional painAbout a fifth to a quarter of people with mood or anxiety disorders report self-medicatingSubstances can cause the symptoms

Summary and Transition

This section explained substance use from the standpoint of the individual. Cooper’s model sorts motives by valence and source, and reviews link coping and enhancement motives most closely to drinking problems. In the 2023 BC AHS, youth most often reported experimenting, having fun and joining friends, while reasons linked to stress and sadness were much more common among youth who used daily. Social learning theory explains how motives are acquired from others, expectancy theory how beliefs shape use, and the self-medication hypothesis how use can follow emotional pain. Each has support, although research on which comes first, the motive or the use, is still evolving.

Each account keeps pointing outward, to friends, families, schools and communities. Section 2 adopts a framework built to describe those wider influences, the social ecological model, and uses Canadian youth data to sort risk and protective factors into its levels.

Key Takeaways

  • 1.1 A motive is the purpose a person hopes a substance will serve on a particular occasion, while a risk factor is a characteristic of a person or their surroundings that is statistically associated with use or harm across a population. Motivational models treat motives as the most proximal predictors of drinking, the last step through which personality, family history, peers and setting influence whether and how much a person drinks.
  • 1.2 Cooper’s model crosses valence (seeking a positive outcome or avoiding a negative one) with source (the person’s own mood or other people). The result is four motives: enhancement, social, coping and conformity.
  • 1.3 In the review by Kuntsche et al. (2005), most young people reported social motives, which were associated with moderate drinking, while enhancement motives were associated with heavy drinking and coping motives with alcohol-related problems. Negative reinforcement is one explanation for the link between coping motives and problems: relief from distress makes use more likely when the feeling returns, and the underlying problem remains.
  • 2.1 BC youth who used only cannabis most often reported experimenting (53 percent), having fun (46 percent) and joining friends (30 percent), while reasons tied to stress and sadness were much more common among youth who used daily. Curiosity, the most common reason in this survey, sits outside Cooper’s model, and the survey data are cross-sectional, allow several answers, refer to the most recent occasion and use categories not designed to match the model.
  • 3.1 Akers’s social learning theory holds that substance use is learned through differential association, definitions, differential reinforcement and imitation. It fits the 2023 finding that most BC youth who had used cannabis last used cannabis shared among friends, although friends may also be chosen because they already use.
  • 3.2 Outcome expectancies are beliefs about a substance’s effects that are learned before most people use it, and in one study they predicted problem drinking a year later. The balanced placebo design crosses what participants are told with what they receive, separating the effect of the drug from the effect of expecting it.
  • 3.3 Khantzian’s self-medication hypothesis holds that people use substances to relieve emotional pain, and about a fifth to a quarter of people with mood or anxiety disorders report self-medicating. In the review and survey described in this lesson, however, most young people used for other reasons, and substances can cause the symptoms they are said to relieve, so a report of self-medication cannot by itself establish the direction of causation.

Reflection

The 2023 BC Adolescent Health Survey asked students why they used substances the last time they did so. Among BC students aged 12 to 18 who had used cannabis and no other substance, the most common reasons were to experiment (53 percent), to have fun (46 percent) and because friends were doing it (30 percent); 21 percent used because of stress and 15 percent because they felt down or sad. Among youth who used cannabis daily, most of whom also used other substances daily, 75 percent last used to have fun, 51 percent because of stress, 44 percent because they felt down or sad and 34 percent because of an addiction. A review by Kuntsche et al. (2005) linked social motives with moderate drinking, enhancement motives with heavy drinking and coping motives with alcohol-related problems. In 150 to 250 words, (a) classify at least four of these reasons using Cooper’s four motives (enhancement, social, coping and conformity), noting any reason that does not fit; (b) explain why coping reasons may be more common among youth who use daily, using negative reinforcement or the self-medication hypothesis; and (c) name one limitation of these survey data that affects your interpretation.

Model answer

Having fun is an enhancement motive, because the person seeks a positive change in their own mood. Using because of stress or because they felt down or sad is a coping motive: the valence is negative and the source is internal. “Friends were doing it” has an external source, so it is a social motive if the young person joined in to share a good time and a conformity motive if they joined in to avoid being left out; the survey wording cannot tell us which. “Wanted to experiment”, the most common reason, does not fit Cooper’s model well, since it has no curiosity motive, although some would treat it as enhancement.

Coping reasons may be more common among youth who use daily because of negative reinforcement. If cannabis relieves stress or sadness, the relief makes using more likely the next time those feelings return, so use can become daily while the source of the distress remains. The self-medication hypothesis makes a similar prediction for young people whose distress comes from a mental health condition.

The main limitation is that the data are cross-sectional, so they cannot show whether distress led to daily use or daily use added to distress. The reasons given by youth who used daily also refer to substance use in general, since most of them used other substances daily, and every answer describes only the most recent occasion.

Minimum 20 characters required.

✓ Reflection saved
Knowledge Check: this section

Question 1: In Cooper’s four-factor model of drinking motives, which motive has a positive valence and an external source?

A social motive seeks a positive outcome (positive valence) from other people (external source), such as celebrating with friends. Enhancement is also positive, which makes it tempting, but its source is internal: the person seeks a change in their own mood.

Question 2: According to the review by Kuntsche et al. (2005), which drinking motive was associated with alcohol-related problems?

The review linked coping motives with alcohol-related problems, consistent with negative reinforcement. Enhancement motives are the tempting distractor, but this review linked them with heavy drinking (a 2021 meta-analysis later linked them to drinking problems as well), while social motives were linked with moderate drinking.

Question 3: Among BC youth who used only cannabis, what was the most common reason given for their most recent use in the 2023 BC Adolescent Health Survey?

Wanting to experiment was chosen by 53 percent, followed by having fun (46 percent) and friends doing it (30 percent). Friends doing it is tempting because peer influence is often assumed to drive youth use, but it ranked third, and feeling pressured to fit in was chosen by only 3 percent.

Question 4: What does the balanced placebo design allow researchers to do?

The design crosses what participants are told with what they actually receive, so the effect of the drug and the effect of expectancy can be estimated separately. Predicting drinking a year later is tempting, but that was the longitudinal design of Christiansen et al. (1989), which measured expectancies without manipulating the drink.
Section 2 of 5

The Social Ecological Model: Risk and Protective Factors Across Levels

⏱ Estimated reading time: 35 minutes
Section 2 of 5

The Social Ecological Model

Risk and protective factors at individual, relationship, community and societal levels.

Origins

Four nested levels

Individual

Stress, mental health, age at first use, plans.

Relationship

Family, friends and partners.

Community

Schools, workplaces, neighbourhoods, nearby stores.

Societal

Laws, economic policy and cultural norms.

Ecological model (McLeroy et al., 1988); CDC four-level model.

Risk and protection

Risk factors describe groups

Risk factor

Associated with a higher chance of an outcome.

Protective factor

Associated with a lower chance, or a weaker effect of risk.

Community
Family
School
Peer and individual

Risk-focused approach (Hawkins et al., 1992); domains from Communities That Care (Center for Communities That Care, 2013).

BC youth, 2023

Findings by level

Relationship

18% who felt respected by family had tried cannabis, compared with 41%.

Community

34% at rural schools, compared with 21% at urban schools.

Community

More cannabis stores near a school, more regular use.

Individual

52% with post-traumatic stress disorder, compared with 21%.

2023 BC Adolescent Health Survey, Blunt Talk III (Smith et al., 2025).

How levels combine

Risk accumulates, protection buffers

54%used regularly with none of three basic needs met
24%used regularly with all three basic needs met

Among youth who had experienced homelessness, 44% with three or more caring adults used regularly, compared with 57% with fewer. Associations from one survey (Smith et al., 2025).

Use and harm

Different factors predict harm

  • Harm depends on what is in the substance, how and where it is used, and who is present.
  • Each level can be read for harm: tolerance, someone to respond, services and laws.
  • Groups with similar use can suffer very different harm.
Carry forward

From factors to determinants

  • Four levels: individual, relationship, community and societal.
  • Risk and protective factors are associations, and risk accumulates.
  • Harm depends on conditions beyond how often people use.

Next: the social determinants of harm.

Introduction

Section 1 explained substance use through the motives, beliefs and learning of individuals, and each account pointed beyond the individual: to friends who were using, parents who knew or did not know where their children were, and communities with little to do. The social ecological model organizes those wider influences into nested levels, from the individual at the centre to the society around them. This section traces the model’s origins, introduces risk and protective factors, applies both to Canadian data on young people’s cannabis use, and ends by separating the factors that make use more likely from those that make it more harmful.

Learning Objectives

  • 4.1 Describe the four levels of the social ecological model (individual, relationship, community and societal) and explain how the model grew out of McLeroy and colleagues’ ecological model for health promotion.
  • 4.2 Define risk factors and protective factors, explain why a risk factor is not necessarily a cause, and describe the domains of risk identified by Hawkins, Catalano and Miller (1992) and used by the Communities That Care framework.
  • 5.1 Classify risk and protective factors for youth substance use reported in the 2023 BC Adolescent Health Survey into ecological levels.
  • 5.2 Explain how factors at different levels combine and buffer one another, using Canadian evidence on basic needs and supportive adults.
  • 6.1 Distinguish the factors that predict substance use from the factors that predict harm.

4.1 From Individual Behaviour to Ecology

Ecological models arose from a criticism of health promotion that focused only on changing individual behaviour. That criticism explains the model’s shape.

4.1.1 McLeroy and the Ecological Model of Health Promotion

In 1988, Kenneth McLeroy and colleagues noted that critics had accused lifestyle interventions of promoting a “victim-blaming” ideology by neglecting the social influences on health. Their response was an ecological model that treats both individual and social environmental factors as targets for health promotion, directing attention to the interpersonal, organizational, community and public policy factors that support and maintain unhealthy behaviours. The model assumes that changes in the social environment produce changes in individuals, and that individuals’ support is needed to put environmental changes in place (McLeroy et al., 1988).

4.1.2 The Four-Level Social Ecological Model

A simpler four-level version is used by the United States Centers for Disease Control and Prevention (CDC), which cites as its source a chapter of the World Health Organization’s 2002 World Report on Violence and Health (CDC, 2025). Although the CDC applies it to violence prevention, its levels describe the influences on substance use equally well. Figure 10.3 draws the levels as nested rings with Canadian examples.

Four nested levels of influence on substance use Societal Laws such as the minimum age to buy cannabis, economic and housing policy, cultural norms about drinking, colonial policies Community Schools, workplaces and neighbourhoods; rural or urban setting; stores near schools; caring adults and activities in the community Relationship Family connection and monitoring, friends who use, dating partners Individual Age at first use, stress and mental health, motives, expectancies, plans for the future
Figure 10.3. The four levels of the social ecological model, with examples relevant to substance use in Canada. Level definitions follow CDC (2025); examples are drawn from this lesson.

The CDC defines the levels as follows (CDC, 2025). The individual level covers biological and personal history factors such as age, education, income, substance use and a history of abuse. The relationship level covers a person’s closest social circle of peers, partners and family. The community level covers the settings in which relationships take place, such as schools, workplaces and neighbourhoods, and their features, such as neighbourhood poverty or a high density of alcohol outlets. The societal level covers broad factors such as health, economic and educational policies and social and cultural norms, including policies that help to maintain economic or social inequalities between groups. The CDC treats each level as both a source of influence and a point for prevention, describes prevention strategies for each level, and states that prevention requires action across several levels at the same time. Table 10.4 maps McLeroy’s five levels onto the CDC’s four.

Table 10.4. The five levels of McLeroy et al. (1988) and the four levels of the CDC social ecological model, with an example of each relevant to substance use.

McLeroy et al. (1988)CDC modelExample relevant to substance use
Individual (intrapersonal)IndividualA young person’s expectancies about cannabis
Interpersonal processes and primary groupsRelationshipFriends who use, or a parent who knows how a teenager spends free time
Organizational (institutional)CommunityA school’s vaping policy or a workplace’s approach to injuries
CommunityCommunityThe number of cannabis stores near a school
Public policySocietalProvincial minimum ages for buying alcohol and cannabis

The model’s main strength is the one McLeroy and colleagues intended: it keeps an explanation from stopping at the individual. Its limits are that the levels overlap, so the same factor can be placed in more than one level depending on how it is measured, and that it lists where influences come from without saying how they work. For mechanisms, it has to be combined with theories such as those in Section 1.

4.2 Risk and Protective Factors

The levels of the model are filled with risk and protective factors, concepts from epidemiology and prevention science that need precise definitions.

A risk factor is a characteristic of a person or their environment that is associated with a higher probability of an outcome, such as early substance use or later problems. A protective factor is associated with a lower probability, or weakens the effect of a risk factor when both are present. In an influential review, J. David Hawkins, Richard Catalano and Janet Miller (1992) argued that the most promising route to preventing adolescent alcohol and drug problems was a risk-focused approach: identify the risk factors, identify methods that have reduced them, and test those methods in controlled studies. The Communities That Care prevention system at the University of Washington later organized these factors into community, family, school, and peer and individual domains, listed in the accordion below (Center for Communities That Care, 2013). HSCI 402 examines the prevention programs built on them.

Community domainv

Availability of drugs; community laws and norms favourable toward drug use; transitions and mobility; low neighbourhood attachment and community disorganization; extreme economic deprivation; and media portrayals of the behaviour.

Family domainv

A family history of the problem behaviour; family management problems, such as unclear expectations and poor supervision; family conflict; and favourable parental attitudes toward, or involvement in, the behaviour.

School domainv

Academic failure beginning in late elementary school, and a lack of commitment to school, in which a young person no longer finds the role of student meaningful.

Peer and individual domainv

Early and persistent antisocial behaviour; rebelliousness; friends who engage in the behaviour; favourable attitudes toward it; early initiation; and constitutional factors such as sensation-seeking and a lack of impulse control.

A risk factor is not necessarily a cause

A risk factor is defined by a statistical association, which can arise in several ways. The factor may cause the outcome; it may be a marker of something else, as when a family history of addiction reflects both shared genes and a shared household; or the direction may run the other way, as when early substance use leads to falling behind at school. Risk factors also describe groups: most young people with any single risk factor never develop a substance use problem. Lesson 9 showed how twin and adoption studies separate genetic from environmental explanations, and similar care is needed before treating any factor in this section as a cause. Risk also tends to accumulate. A 2013 review of research on child development concluded that exposure to several risk factors has larger effects than exposure to any single one (Evans et al., 2013), and in a 1986 study of high school students the number of risk factors predicted increases in substance use over the following year (Newcomb et al., 1986).

5.1 Canadian Youth Evidence from the 2023 BC Adolescent Health Survey

The BC Adolescent Health Survey, introduced in Section 1, provides detailed Canadian evidence on risk and protective factors for youth substance use. This part summarizes recent trends and then sorts the findings by level.

5.1.1 Recent Trends

Substance use among BC students has declined over three decades. In 2023, 38 percent of students aged 12 to 19 had tried alcohol beyond a few sips, the lowest share in 30 years, compared with 67 percent in 1992 (Smith et al., 2024). Among students aged 12 to 18, the share who had tried cannabis fell from 25 percent in 2018 to 22 percent in 2023, the lowest in 25 years, and 26 percent had ever vaped (Smith et al., 2024; Smith et al., 2025). The decline is not uniform: among youth who had tried cannabis, the share who said they needed help for their use in the past year rose to 17 percent, from 10 percent in 2018 and 2013 (Smith et al., 2025).

5.1.2 Factors by Level

The McCreary report Blunt Talk III compares cannabis use between groups of youth defined by many personal, family, school and community characteristics (Smith et al., 2025). Table 10.5 sorts a selection by ecological level. Each comparison is a percentage difference from a cross-sectional survey, an association of the kind described in 4.2 Risk and Protective Factors.

Table 10.5. Selected risk and protective factors for cannabis use among BC students aged 12 to 18, 2023 BC Adolescent Health Survey, by ecological level. Percentages are the share who had ever tried cannabis unless stated otherwise. Data from Smith et al. (2025).

LevelFactorTypeFinding
IndividualPost-traumatic stress disorderRisk52% compared with 21% without
IndividualPlans for post-secondary educationProtective21% compared with 36% without such plans
RelationshipFelt their family respected them quite a bit or very muchProtective18% compared with 41% who felt respected a little or not at all
RelationshipIn a dating relationship in the past yearRisk36% compared with 12%
CommunityFelt like a part of their schoolProtective17% compared with 35%
CommunityAttended a school in a rural areaRisk34% compared with 21% in urban areas
CommunityMore cannabis stores within a kilometre of the schoolRiskUse on 6 or more days in the past month: 42% with four nearby stores, 28% with one
SocietalProvincial law bars sales of cannabis to people under 19Protective in intentAmong youth who had used cannabis, 10% had most recently bought it from a store

Relationship-level protective factors show some of the largest differences: youth who felt respected by their family were less than half as likely to have tried cannabis as those who did not. Community settings also matter. The rural and urban difference has persisted across 15 years of surveys, and the more cannabis stores there were near a school, the more likely its students were to have tried cannabis and to use it regularly, which the Communities That Care framework would classify as availability (Smith et al., 2025). Use Interactive 10.2 to place eight factors into the four levels; the feedback explains factors that could sit in more than one.

Interactive 10.2
Sort risk and protective factors into ecological levels

Each card describes a factor reported for BC youth in the 2023 BC AHS (Smith et al., 2025). Choose the level of the social ecological model where it fits best, then check your answers.

The young person felt that parents or guardians knew what they did in their free time most or all of the time
Protective: 17% had tried cannabis, compared with 40% of youth who felt their parents rarely or never knew.
Relationship. Parents and guardians belong to a young person’s closest social circle, which the CDC places at the relationship level.
Attended a school in a rural area
Risk: 34% had tried cannabis, compared with 21% at urban schools.
Community. Rurality describes a setting; rural youth linked their use to boredom and few local activities.
Had post-traumatic stress disorder
Risk: 52% had tried cannabis, compared with 21% without the condition.
Individual. A diagnosis is part of personal history, although the trauma behind it may have arisen elsewhere (Section 3).
Provincial law bars the sale of cannabis to anyone under 19
Intended as protective; even so, among youth who had used cannabis, 10% had most recently bought it from a store.
Societal. A law is a public policy; whether local stores follow it is a community-level question.
Felt like a part of their school
Protective: 17% had tried cannabis, compared with 35% of youth who did not.
Community. Schools are community settings in the CDC model; a single caring teacher could also count as a relationship.
Cultural norms about cannabis and alcohol
Asian youth who reviewed the findings said cannabis use was strongly discouraged and stigmatized in Asian communities, whereas alcohol was normalized.
Societal is the best fit, and community is defensible: the CDC places cultural norms at the societal level, but norms shared within one cultural community also act locally.
Worked 21 or more hours a week at a paid job during the school year
Risk: 54% had tried cannabis, compared with 14% of youth who did not work; the finding held when age was controlled.
Community is the best fit, and individual is defensible. Workplaces are community settings, and youth linked long hours to stress, spending money and time away from family.
Legalization of adult cannabis use under the Cannabis Act, in force from 17 October 2018
Between the 2018 and 2023 surveys, the share of BC youth who had tried cannabis fell from 25% to 22%.
Societal. Timing alone cannot show cause: when the survey found a shift toward earlier first use, its research director noted the same shift for alcohol and tobacco and concluded legalization did not explain it (UBC School of Nursing, 2024). Lesson 12 examines legalization.

5.2 How Levels Combine and Interact

The levels of the model do not act independently. Risk accumulates across levels, and protection at one level can soften risk that arises at another.

5.2.1 Basic Needs and a Gradient of Risk

The BC AHS measured three basic needs: food security, housing stability and the absence of material deprivation. Each was protective on its own; for example, 21 percent of youth with stable housing had tried cannabis, compared with 55 percent of youth who had experienced homelessness in the past year (Smith et al., 2025). Figure 10.4 shows what happened when the needs were combined.

Regular cannabis use falls as more basic needs are met 54%None of 3 42%1 of 3 33%2 of 3 24%All 3 Basic needs met: food security, housing stability, no material deprivation Used on 6 or more days
Figure 10.4. Share of BC youth who had tried cannabis and used it on 6 or more days in the past month, by the number of three basic needs met, 2023 BC Adolescent Health Survey. Data from Smith et al. (2025).

Regular use was more than twice as common among youth with none of the three needs met as among youth with all three met. Food, housing and material resources are measured for each young person, but whether they are met depends largely on household income, housing markets and income support, which belong to the community and societal levels. A factor can be measured at one level and produced at another, a point Section 3 develops.

5.2.2 Protection That Buffers Risk

In the BC data, protective factors were associated with less regular use even among young people facing serious adversity. Case 10.1 brings together BC findings on supportive adults and connection among youth facing adversity.

Case 10.1: Supportive adults for BC youth facing adversity

Among BC youth with experience of government care who had tried cannabis, 43 percent of those who had an adult to help them make or get to appointments had used cannabis on 6 or more days in the past month, compared with 55 percent of those without such an adult. Among youth who had experienced homelessness and had used cannabis, the figures were 44 percent for those with three or more caring adults in their community (an estimate the report flags as less precise) and 57 percent for those with fewer. Among youth who had tried cannabis and had no adult in their family to talk to about their problems, those who felt a teacher cared about them were less likely to use on 6 or more days (26 percent compared with 39 percent) (Smith et al., 2025).

A related analysis found that among trans and nonbinary youth who used alcohol or other substances, high family connectedness was associated with much lower odds of three or more negative consequences of use, including 95 percent lower odds for trans boys and 93 percent lower odds for nonbinary youth (Saewyc et al., 2024).

These are associations from one survey, and they fit the idea that relationship and community protection can partly offset risks that arise elsewhere, although youth who have caring adults may differ in other ways from youth who do not. They also show why the levels are best read together: a caring teacher is a relationship, the school that makes room for that relationship is a community setting, and school funding is a matter of public policy.

Try it 10.2: Reading a protective factor across levels

Choose one protective factor from Table 10.5 or Case 10.1. In three or four sentences, (a) explain how it might reduce regular cannabis use, using one theory from Section 1; (b) name a factor at a different level that would make it more or less available to young people; and (c) suggest one other explanation for the association that would not involve the factor protecting anyone.

6.1 Factors That Predict Use and Factors That Predict Harm

Most evidence in this section concerns who tries a substance and who uses it often. Who is harmed is a separate question, and the difference shapes the rest of this lesson.

In the 2023 BC AHS, more youth had tried alcohol (38 percent) than cannabis (22 percent) or vaping (26 percent) (Smith et al., 2024; Smith et al., 2025), and Section 1 showed that most young people use for social and enhancement reasons. Harm depends on more than whether and how often a person uses: on what is in the substance, how it is taken, where and with whom, and what help is available. Lesson 5 showed that Canada’s unregulated drug supply has become unpredictable and highly potent, so a single use can be fatal for a person with low tolerance. Each level can be read for harm as well as for use: tolerance and health at the individual level; whether someone is present to respond to an overdose at the relationship level; access to services such as naloxone and drug checking at the community level, which HSCI 402 develops; and laws on regulation and criminalization at the societal level. Because harm depends so heavily on these conditions, groups with similar levels of use can experience very different levels of harm.

Summary and Transition

This section introduced the social ecological model, which McLeroy and colleagues proposed to keep health promotion from blaming individuals for behaviour shaped by their surroundings, and the CDC’s four-level version of it. Hawkins, Catalano and Miller’s risk-focused approach fills the levels with risk and protective factors. In the 2023 BC AHS, family respect, parental knowledge and school belonging were associated with less cannabis use, while trauma, rural schooling, nearby cannabis stores and long working hours were associated with more; risk rose steeply as basic needs went unmet, and supportive adults were associated with less regular use among youth facing adversity.

Section 3 follows the thread on harm. It examines the social determinants of substance use harms, including poverty, housing, childhood adversity, colonialism, racism, dislocation and criminalization, and shows why the burden of harm in Canada falls so unevenly.

Key Takeaways

  • 4.1 McLeroy and colleagues (1988) proposed an ecological model so that health promotion would not blame individuals for behaviour shaped by their surroundings, and the CDC’s four-level version places influences at the individual, relationship, community and societal levels. The levels overlap and say where influences come from rather than how they work, so the model is best combined with theories such as those in Section 1.
  • 4.2 A risk factor is associated with a higher probability of an outcome, and a protective factor with a lower probability or a weaker effect of risk. Because these are statistical associations, a risk factor may be a cause, a marker of something else or a consequence of the outcome. Hawkins, Catalano and Miller’s risk-focused approach, organized by Communities That Care into community, family, school, and peer and individual domains, fills the levels with such factors.
  • 5.1 In the 2023 BC AHS, family respect, parental knowledge and school belonging were associated with less cannabis use, while post-traumatic stress disorder, rural schooling, nearby cannabis stores and long working hours were associated with more. Youth substance use in British Columbia has declined over three decades, although more youth who had tried cannabis said they needed help.
  • 5.2 Risk tends to accumulate across levels: regular cannabis use was more than twice as common among youth with none of three basic needs met as among youth with all three met. Supportive adults were associated with less regular use among youth facing adversity, which fits the idea that protection at one level can buffer risk that arises at another.
  • 6.1 Most evidence on risk and protective factors concerns who tries a substance and who uses it often. Harm also depends on what is in the substance, how and where it is taken, who is present and what help is available, so groups with similar levels of use can experience very different levels of harm.

Reflection

The 2023 BC Adolescent Health Survey reported these findings for BC students aged 12 to 18. Of youth who felt their family respected them quite a bit or very much, 18 percent had tried cannabis, compared with 41 percent of those who felt respected only a little or not at all. Of youth at rural schools, 34 percent had tried cannabis, compared with 21 percent at urban schools. Among youth who had tried cannabis, 54 percent of those with none of three basic needs met (food security, housing stability and no material deprivation) had used it on 6 or more days in the past month, compared with 24 percent of those with all three needs met. Among youth who had experienced homelessness and used cannabis, 44 percent of those with three or more caring adults in their community used it on 6 or more days in the past month, compared with 57 percent of those with fewer such adults. In 150 to 250 words, (a) place the first three findings in levels of the social ecological model (individual, relationship, community or societal), explaining one placement that could be argued; (b) explain what the last finding suggests about how levels interact; and (c) use one finding to explain why a risk or protective factor is not necessarily a cause.

Model answer

Family respect belongs at the relationship level, because it describes the young person’s closest social circle. Attending a rural school belongs at the community level, since it describes the setting in which young people live, and youth linked rural use to boredom and few local activities. The basic needs gradient can be argued two ways. Food, housing and material resources are measured for each young person, which suggests the individual level, but whether they are met depends on household income, housing markets and income support, which places their causes at the community and societal levels.

The last finding suggests that protection at one level can buffer risk that arises at another. Homelessness is a serious risk, yet youth who had several caring adults in their community were less likely to use cannabis regularly, so relationships and community settings partly offset a hardship produced by wider conditions.

The family respect finding shows why a protective factor is not necessarily a cause. Youth who feel respected may differ in other ways, such as household income or mental health, and the direction could also run backwards if a young person’s cannabis use led to conflict that reduced their sense of being respected. A cross-sectional survey cannot rule out these explanations.

Minimum 20 characters required.

✓ Reflection saved
Knowledge Check: this section

Question 1: In the CDC’s four-level social ecological model, at which level does a school or workplace belong?

The CDC places the settings in which social relationships take place, such as schools, workplaces and neighbourhoods, at the community level. The relationship level is tempting because a teacher or co-worker is a relationship, but the setting itself is part of the community level.

Question 2: Why did McLeroy and colleagues (1988) propose an ecological model of health promotion?

McLeroy and colleagues responded to critics who said lifestyle interventions promoted victim-blaming by neglecting social influences, and their model made interpersonal, organizational, community and policy factors targets for change. The claim about individual behaviour reverses their argument.

Question 3: Among BC youth who had tried cannabis in 2023, 54 percent of those with none of three basic needs met used it on 6 or more days in the past month, compared with 24 percent of those with all three met. Which interpretation is best supported?

The steady gradient shows that risk accumulates as more needs go unmet, and whether needs are met depends on income, housing and support policies at the community and societal levels. The data are cross-sectional associations, so options claiming that meeting needs stops use or that housing policy is the main cause go beyond the evidence.

Question 4: Which statement about risk factors is correct?

A risk factor is defined by a statistical association, which may reflect causation, a marker of another cause or reverse causation. The idea that most people with a risk factor develop the outcome is tempting, but risk factors describe groups, and most young people with any single risk factor never develop a substance use problem.
Section 3 of 5

Social Determinants of Substance Use Harms

⏱ Estimated reading time: 40 minutes
Section 3 of 5

Social Determinants of Substance Use Harms

Why harm falls unevenly, and how to measure the gap.

Harm is patterned

More unequal than use

Alcohol harm paradox

More alcohol harm among people with low socioeconomic status from similar or lower drinking; alcohol use explained only part of the gap in deaths (Probst et al., 2020).

BC toxic drug deaths

Downtown Eastside death rate 10 times the provincial average, yet about 85% of deaths occurred elsewhere (BC Coroners Service, 2023).

Poverty and housing

Where and how people use

458deaths of people experiencing homelessness in BC, 2023
86%of those deaths due to unregulated drug toxicity
48%of BC unregulated drug deaths in 2025 occurred in a private residence

Coroner data (BC Coroners Service, 2025, 2026).

Childhood adversity

Toxic stress and later drug problems

Toxic stress

Prolonged activation of the stress response without supportive relationships to calm it.

Adverse childhood experiences

In the original ACE study, five or more: 7 to 10 times more likely to report drug use problems (Dube et al., 2003). Meta-analyses also link adversity to substance use (Hughes et al., 2017; Zhu et al., 2025).

Colonialism and racism

Naming the causes

5.4 timesFirst Nations toxic drug death rate compared with other BC residents, 2025 (FNHA, 2026)
26%of Indigenous respondents always treated as if drunk or asked about substance use in health care (In Plain Sight; Turpel-Lafond, 2020)

The ratio describes the size of a gap that colonial policies, residential schools and racism help to explain.

Rate ratio

Measuring the gap

Equation 10.1
\[ RR = \frac{D \div P}{(100 - D) \div (100 - P)} \]

Worked example: 15.8 ÷ 3.4 = 4.65; 84.2 ÷ 96.6 = 0.872; 4.65 ÷ 0.872 = 5.3 (published figure 5.4).

Dislocation and criminalization

Disconnection and using alone

Dislocation

The sustained absence of psychosocial integration is painful and linked to addiction (Alexander, 2008); a broad theory that is hard to test.

Criminalization

Fear of criminal sanction leads people to hide their substance use (BC Coroners Service, 2023); hidden use can mean using alone.

Carry forward

From determinants to populations

  • In this section’s examples, harm was distributed more unequally than use.
  • Seven determinants help explain where harm concentrates; the evidence ranges from consistent reviews to theory that is hard to test.
  • A rate ratio measures the gap when the comparison is chosen well.

Next: populations with unique vulnerabilities or needs.

Content note. This section discusses drug toxicity deaths, childhood abuse, residential schools and racism in health care. If you would like support, the 9-8-8 Suicide Crisis Helpline is available by call or text at 9-8-8, the National Overdose Response Service can be reached at 1-888-688-6677, and the Hope for Wellness Help Line for Indigenous people can be reached at 1-855-242-3310. The First Nations Health Authority also lists cultural support from the Tsow-Tun Le Lum Society at 1-888-403-3123 (limited hours).

Introduction

Section 2 separated the factors that predict substance use from those that predict harm. This section examines the second set. The World Health Organization defines the social determinants of health as the conditions in which people are born, grow, live, work and age, and people’s access to power, money and resources (World Health Organization [WHO], n.d.). Applied to substance use, they help explain why groups that use substances at similar rates can experience very different rates of injury, illness and death.

Social epidemiologists describe substance use as shaped by the sociocultural context, such as social networks and family functioning; by the physical and built environment, such as neighbourhood poverty; and by access to health care and education (Galea et al., 2004, as summarized by McCuistian et al., 2021). A 2023 death review panel convened by the BC Coroners Service put the point directly: substance use “often results from unaddressed concurrent challenges related to chronic pain, homelessness, poverty, racism, and mental health”, many of them “multi-generational” and “rooted in unaddressed physical, emotional, and psychological trauma” (BC Coroners Service, 2023). HSCI 214 develops the stress response, adverse childhood experiences and colonialism in depth, and HSCI 402 develops overdose mortality.

Learning Objectives

  • 6.2 Define the social determinants of substance use harms and explain why harms are distributed more unequally than use, with reference to the alcohol harm paradox.
  • 7.1 Describe how poverty and housing shape substance use harms, using BC Coroners Service data on deaths among people experiencing homelessness.
  • 7.2 Explain the stress response and toxic stress, and summarize the evidence from the Adverse Childhood Experiences study and later reviews on childhood adversity and drug use.
  • 7.3 Explain how colonialism and racism shape the toxic drug crisis among First Nations people in British Columbia, using First Nations Health Authority data and naming structural causes.
  • 7.4 Describe Alexander’s dislocation theory and identify its main limitation.
  • 7.5 Explain how criminalization adds to substance use harms.
  • 8.1 Calculate a rate ratio from a group’s share of deaths and share of the population, and explain why the choice of comparison population matters.

6.2 Why Substance Use Harms Are Socially Patterned

If harm depended only on consumption, use and harm would rise and fall together across social groups. They do not, as the alcohol harm paradox shows.

The alcohol harm paradox

People with low socioeconomic status experience disproportionately more alcohol-attributable harm than people with high socioeconomic status from similar or lower amounts of drinking. Probst and colleagues, a team that included researchers at Toronto’s Centre for Addiction and Mental Health, systematically reviewed ten longitudinal studies of more than 400,000 adults to examine this alcohol harm paradox. Alcohol use explained up to 27 percent of socioeconomic inequalities in mortality; the share depended on drinking pattern, with heavy episodic drinking having particular explanatory value; and there was some, although scarce, evidence that socioeconomic status and alcohol use interact, so that the same drinking carries more risk for people with fewer resources (Probst et al., 2020).

The toxic drug crisis shows a related pattern. The BC Coroners Service’s 2023 death review panel reported that the death rate in Vancouver’s Downtown Eastside was 10 times the provincial average, yet about 85 percent of deaths occurred outside the neighbourhood, in every health authority and among people of all socioeconomic groups; there is “no single stereotype” of the people who have died (BC Coroners Service, 2023). Conditions in one neighbourhood concentrate risk, while a toxic supply makes any use risky wherever it happens.

Try it 10.3: Concentrated risk and widespread deaths

Using the two figures from the death review panel (a Downtown Eastside death rate 10 times the provincial average, and about 85 percent of deaths elsewhere), write three sentences explaining how both can be true. Then give one reason why a response focused only on the Downtown Eastside would miss most deaths, and one reason why the neighbourhood still needs a response of its own.

7.1 Poverty and Housing

Income and housing are measured for individuals but produced by labour markets, housing markets and income support, so their causes sit at the community and societal levels.

Lesson 5 showed that the unregulated supply is unpredictable, so the conditions in which a person uses can decide whether an overdose is survived; people with low incomes and unstable housing usually have less control over those conditions and fewer routes into care. The death review panel noted that people seeking treatment in British Columbia meet waiting times and barriers that include precarious housing, employment and childcare, and that securing stable housing and pathways to employment improves health outcomes (BC Coroners Service, 2023). Case 10.2 describes the deaths of people experiencing homelessness in the province.

Case 10.2: Deaths of people experiencing homelessness in British Columbia, 2016 to 2023

The BC Coroners Service reported in March 2025 that at least 458 people experiencing homelessness died in British Columbia in 2023, an increase of 23 percent from 373 in 2022 and nearly three times the 155 deaths in 2020; 1,940 such deaths were reported between 2016 and 2023. In 2023, 86 percent of these deaths (394) were due to accidental unregulated drug toxicity. More than half of the people who died were aged 30 to 49, 79 percent were male, and about 46 percent were unsheltered, living outdoors, in a makeshift shelter, a vehicle or another place not intended for habitation (BC Coroners Service, 2025).

The Coroners Service cautioned that it does not investigate every death of a person experiencing homelessness and that the data are preliminary. Even so, drug toxicity is by far the leading cause of the deaths it investigates among people without housing.

Housing shapes harm among people who are housed as well. In 2025, 48 percent of unregulated drug deaths in British Columbia occurred in a private residence and 20 percent outdoors (BC Coroners Service, 2026). A private residence offers safety from exposure and theft but can also mean using alone, a risk that 7.5 Criminalization returns to.

7.2 Childhood Adversity and the Stress Response

Some determinants act long before a person first uses a substance. Childhood adversity is one of them, and its effects are thought to run partly through the body’s stress response.

When a person faces a threat, the stress response raises heart rate and blood pressure and releases stress hormones such as cortisol. The Center on the Developing Child at Harvard University distinguishes positive stress, a brief normal reaction; tolerable stress, a more severe reaction buffered by supportive relationships; and toxic stress, which occurs when stress responses stay activated at high levels for long periods without supportive relationships to calm them, and which can disrupt the development of the brain and other body systems and raise the risk of stress-related disease into adulthood (Center on the Developing Child, 2025). A report from the American Academy of Pediatrics argued that toxic stress links early adversity to later problems in learning, behaviour and health, and that disparities associated with poverty, discrimination or maltreatment could be reduced by alleviating toxic stress in childhood (Shonkoff & Garner, 2012).

Early evidence linking childhood adversity to drug use came from the Adverse Childhood Experiences (ACE) study in California. In a 2003 analysis, Dube and colleagues reported on 8,613 adults attending a primary care clinic about ten categories of abuse, neglect and household dysfunction in childhood. Each adverse experience increased the likelihood of starting illicit drug use early by two to four times, and risk rose step by step with the number of experiences. Compared with people who reported none, people who reported five or more were 7 to 10 times more likely to report illicit drug use problems, addiction and injection drug use (Dube et al., 2003). The study relied on adults recalling their childhoods and on one clinic’s patients, and its estimate that adverse experiences accounted for 56 to 67 percent of these problems assumes the associations are causal. Later meta-analyses found the same direction of association across many studies: in a 2017 meta-analysis of 37 studies, people who reported four or more adverse experiences had more than seven times the odds of problematic drug use (Hughes et al., 2017), and a 2025 meta-analysis of 102 studies linked adverse childhood experiences to smoking, alcohol, cannabis and other drug use (Zhu et al., 2025). The association is consistent across reviews, but adversity cannot be assigned in an experiment, so how much of it is causal is still being studied. Canadian youth data point the same way: in the 2023 BC AHS, 42 percent of youth with experience of government care had used cannabis, compared with 21 percent of other youth (Smith et al., 2025). Lesson 9 examined related evidence on early adversity, social rank and vulnerability to addiction. Figure 10.5 draws these ideas together, showing how societal conditions reach individuals through living conditions and early experiences, change patterns of use, and shape harm.

From social conditions to substance use harms Structural conditions colonialism racism poverty housing policy criminalization Living conditions and experiences childhood adversity toxic stress unstable housing discrimination in care Patterns of use coping motives earlier first use more frequent use using alone or in hiding Harms drug toxicity deaths injury and illness substance use disorder Toxic unregulated supply raises the harm of any use harms feed back into poverty and family loss
Figure 10.5. A conceptual pathway, drawn for this lesson, from structural conditions through living conditions and patterns of use to harm. The toxic unregulated supply adds risk at the final step, and harms such as bereavement and lost income feed back into the conditions that produced them.

7.3 Colonialism and Racism

Colonialism and racism act on particular peoples, and they help explain why the toxic drug crisis has fallen so heavily on First Nations people in British Columbia. This part describes the policies and their consequences factually; HSCI 214 develops colonialism as a determinant of health.

7.3.1 Colonial Policies and Their Effects

For more than 150 years, First Nations, Inuit and Métis children were taken from their families and communities to attend residential schools. More than 150,000 children attended, and many never returned. By the 1880s the federal government funded the schools with the explicit intent of separating children from their families and cultures, and in 1920 the Indian Act made attendance compulsory for Treaty-status children aged 7 to 15 (National Centre for Truth and Reconciliation [NCTR], 2026). Lesson 2 described the liquor offences that the Indian Act applied only to Status Indians, one of which the Supreme Court of Canada declared inoperative in R v Drybones (1970).

The effects continue across generations. A 2026 report by the McCreary Centre Society’s Young Indigenous Research Team, based on the 2023 BC AHS, found that First Nations youth with a family history of residential school attendance were more likely than other First Nations youth to have experienced physical abuse (28 percent compared with 14 percent), government care (21 percent compared with 10 percent) and material deprivation, and to have been bereaved by an overdose (32 percent compared with 11 percent). The majority of First Nations youth (54 percent) had experienced discrimination in the past year (McCreary Centre Society’s Young Indigenous Research Team, 2026).

7.3.2 The Toxic Drug Crisis among First Nations People

The First Nations Health Authority (FNHA) reports toxic drug deaths among First Nations people in British Columbia using data from the BC Centre for Disease Control, the BC Coroners Service, BC Emergency Health Services and the BC Ministry of Health. Its report for 2025 states that First Nations individuals, families and communities “continue to be disproportionately affected by the toxic drug crisis”. In 2025, 289 First Nations people died from toxic drug poisoning, a decrease of 33.3 percent from 433 in 2024. First Nations people make up 3.4 percent of the province’s population but accounted for 15.8 percent of toxic drug deaths (FNHA, 2026). Figure 10.6 shows the resulting rate ratios.

Toxic drug death rate, First Nations people compared with other BC residents All First Nations, 20246.5 times All First Nations, 20255.4 times First Nations females, 20259.1 times First Nations males, 20254.5 times 1 = equal rates Rate ratio (First Nations people compared with other BC residents)
Figure 10.6. Rate ratios for toxic drug poisoning deaths among First Nations people compared with other BC residents, 2024 and 2025. Female and male ratios compare First Nations people with other residents of the same sex. Data from the First Nations Health Authority (FNHA, 2026).

First Nations people died at 5.4 times the rate of other BC residents in 2025, down from 6.5 times in 2024, and First Nations females died at 9.1 times the rate of other female residents (FNHA, 2026). The FNHA notes that its data include only Status First Nations people and their status-eligible descendants. The death review panel added that many of the underlying determinants also apply to other Indigenous peoples, and that a specific strategy must be developed with Indigenous peoples to respect their right to self-determination (BC Coroners Service, 2023).

Naming the causes of a gap

A rate ratio of 5.4 describes the size of a gap that colonial policies, racism, poverty, an unequal health system and the toxic supply help to explain. The risk lies in the conditions created by residential schools, the removal of children and discrimination in care, so describing First Nations people as a “risk group” would misplace the cause. The data also show change: deaths fell by a third between 2024 and 2025. Section 4 describes the strengths and culturally grounded harm reduction that First Nations organizations and youth identify.

7.3.3 Racism in Health Care

Racism shapes harm through the services meant to reduce it: a person who expects to be judged or dismissed may delay care or avoid it. Case 10.3 describes the British Columbia review that documented this pattern.

Case 10.3: In Plain Sight (2020)

In June 2020, after allegations that staff in some BC emergency departments were guessing the blood alcohol levels of Indigenous patients, British Columbia’s Minister of Health appointed Mary Ellen Turpel-Lafond to lead an independent review of Indigenous-specific racism in the province’s health care system. The review heard from almost 9,000 people, including 2,780 respondents to an Indigenous Peoples’ Survey, 80 percent of whom were Indigenous. Only 16 percent of Indigenous respondents reported no discrimination in health care on any of eight factors related to stereotyping, and 26 percent reported “always” being treated as if they were drunk, or being asked about substance use, when accessing care, compared with 5 percent of non-Indigenous respondents (Turpel-Lafond, 2020).

The review described consequences that included under-treated pain and symptoms taken less seriously on the assumption that the patient was seeking medication (Turpel-Lafond, 2020).

Racism also shapes who is policed and punished for drug use, which Lesson 11 examines through evidence on cannabis arrests by race and the overrepresentation of Black and Indigenous people in prisons.

7.4 Dislocation

A broader account links these determinants to the way modern societies are organized. It comes from the Simon Fraser University psychologist Bruce Alexander, whose Rat Park experiments Lesson 9 examined.

In The Globalization of Addiction, Alexander (2008) proposed a dislocation theory of addiction. Its first principle holds that psychosocial integration is an essential part of human well-being and that dislocation, the sustained absence of psychosocial integration, is “excruciatingly painful” (Alexander, 2008, p. 86, as cited by Ross et al., 2022). Psychosocial integration describes a balance between the need to belong and the equally strong need for autonomy and achievement. Alexander linked addiction in Canada and elsewhere to a growing sense of alienation and disconnection produced by free-market economies and globalization, so that the remedy lies in restoring connection, social inclusion and belonging (Ross et al., 2022).

Applied to this lesson, the theory directs attention to policies and conditions that break the ties between people, families, cultures and land, such as residential schools, the removal of children into government care, the loss of housing and work, and the isolation that criminalization produces. It fits the youth findings in Section 2, where connection was associated with less regular cannabis use. Its central concepts are, however, broad and hard to measure, which makes the theory difficult to test against competing explanations, and Lesson 9 described the contested replication record of the Rat Park experiments that first prompted it. The theory therefore remains largely speculative, although it offers a way of connecting the other determinants.

7.5 Criminalization

The final determinant is the legal status of the drugs themselves. Treating possession as a crime changes how and where people use, and it adds harms of its own.

The death review panel listed the harms associated with substance use as including stigma and shame, criminal justice involvement, using alone and high-risk consumption patterns, as well as blood-borne infections and drug toxicity deaths (BC Coroners Service, 2023). Criminalization contributes to several of these. The panel wrote that fear of criminal sanction or public disapproval leads people to hide their substance use, and that recognizing substance use as a health issue encourages them to seek help (BC Coroners Service, 2023). Hidden use often means using alone and in private, where no one can respond to an overdose. A 2022 Ontario report on opioid toxicity deaths among people who had worked in construction found that fewer than 20 percent of those deaths occurred where someone was known to be present to intervene, and suggested that stigma and fear of losing employment may contribute to using alone (Gomes et al., 2022). Lessons 11 and 12 examine policing and British Columbia’s decriminalization exemption (31 January 2023 to 31 January 2026). Table 10.6 summarizes the seven determinants.

Table 10.6. Seven social determinants of substance use harms, the main pathway through which each raises harm, and evidence presented in this section.

DeterminantMain pathway to harmEvidence in this section
PovertyThe same use carries more risk with fewer resourcesAlcohol harm paradox (Probst et al., 2020)
HousingUsing in unsafe places; barriers to treatment458 deaths of people experiencing homelessness in BC in 2023
Childhood adversityToxic stress; earlier first use42% of BC youth with government care experience had used cannabis
ColonialismIntergenerational trauma and family separationFirst Nations death rate 5.4 times that of other BC residents (2025)
RacismDiscrimination and under-treated pain in care26% of Indigenous respondents always treated as if drunk or asked about substance use
DislocationLoss of belonging, identity and purposeConnection associated with less regular youth cannabis use
CriminalizationUsing alone and in hiding; stigmaBC death review panel: fear of criminal sanction leads people to hide substance use

8.1 Calculating a Rate Ratio

The FNHA’s finding in 7.3.2, that First Nations people died at 5.4 times the rate of other BC residents in 2025, is a rate ratio, a measure that summarizes the size of a gap between groups. The FNHA’s rate ratio can be approximated from the two percentages its report gives, the group’s share of deaths and its share of the population, because total deaths and total population cancel out. Equation 10.1 gives the formula.

Rate ratio from shares of deaths and population
RR = (D ÷ P) ÷ [(100 − D) ÷ (100 − P)]Eq 10.1
the rate ratio equals the group’s percentage share of all deaths divided by its percentage share of the population, divided in turn by the same ratio for everyone else.

The first part, D ÷ P, is a representation ratio: how many times larger a group’s share of deaths is than its share of the population. Dividing by the same ratio for everyone else gives a rate ratio, which compares the group’s death rate with the rate among all other people.

Worked Example 10.1: First Nations toxic drug deaths in British Columbia, 2025

First Nations people made up 3.4 percent of the population and 15.8 percent of toxic drug deaths in 2025 (FNHA, 2026). The representation ratio is 15.8 ÷ 3.4 = 4.65. Everyone else made up 96.6 percent of the population and 84.2 percent of deaths, a ratio of 84.2 ÷ 96.6 = 0.872. The rate ratio is 4.65 ÷ 0.872 = 5.33, or about 5.3.

The FNHA’s published figure is 5.4. The difference reflects rounding: a population share of 3.35 percent, which rounds to 3.4, would give 5.41. A ratio rebuilt from rounded percentages is therefore an approximation.

Use Calculator 10.1 to repeat the worked example and to explore a second case. The Ontario construction report cited in 7.5 found that people who had worked in construction accounted for 7.9 percent of opioid toxicity deaths in Ontario between July 2017 and December 2020, while construction workers made up 3.6 percent of Ontario’s population and 7.2 percent of its employed people in 2021 (Gomes et al., 2022). Section 4 examines that report. Notice how much the answer changes with the comparison population.

The two Ontario presets give rate ratios of about 2.3 and 1.1. Neither is exactly right: the deaths include anyone who had ever worked in construction, while the population shares count only people working in 2021, and construction workers are mostly working-age men, the group most affected by opioid deaths overall. Choosing a fair comparison population, and adjusting for age and sex where possible, is part of reading any rate ratio; HSCI 402 develops these methods.

Summary and Transition

In this section’s examples, from the alcohol harm paradox to British Columbia’s death data, substance use harms were distributed far more unequally than substance use. The alcohol harm paradox describes more harm among people with fewer resources from similar or lower drinking, and British Columbia’s toxic drug deaths are both concentrated and widespread. Poverty and housing shape where and how people use; childhood adversity is consistently associated with later drug problems across reviews, possibly in part through toxic stress; and colonialism and racism, acting through residential schools, the removal of children and discrimination in care, help explain why First Nations people died at 5.4 times the rate of other BC residents in 2025. Alexander’s dislocation theory, which remains largely speculative, links these determinants through the loss of psychosocial integration, and the 2023 death review panel wrote that fear of criminal sanction leads people to hide their substance use, which can mean using alone.

Section 4 turns from determinants to people. It examines seven populations whose vulnerabilities or needs differ: youth, older adults, pregnant people, Indigenous peoples, 2SLGBTQ+ people, workers in high-risk trades and people leaving prison.

Key Takeaways

  • 6.2 The social determinants of health are the conditions in which people are born, grow, live, work and age, and their access to power, money and resources. In the alcohol harm paradox, people with low socioeconomic status experience more alcohol-attributable harm from similar or lower drinking, and British Columbia’s toxic drug deaths are both concentrated, at 10 times the provincial rate in the Downtown Eastside, and widespread, with about 85 percent occurring elsewhere.
  • 7.1 Income and housing are measured for individuals but produced by labour markets, housing markets and income support. At least 458 people experiencing homelessness died in British Columbia in 2023, 86 percent of them from unregulated drug toxicity, and housing shapes harm among people who are housed too, since 48 percent of 2025 deaths occurred in a private residence.
  • 7.2 Toxic stress occurs when stress responses stay highly activated for long periods without supportive relationships to calm them. In the ACE study, people who reported five or more adverse childhood experiences were 7 to 10 times more likely to report illicit drug use problems, although the study relied on adults’ recall and its estimates of attributable risk assume the associations are causal.
  • 7.3 Residential schools, the removal of children and Indigenous-specific racism in health care help explain why First Nations people in British Columbia died of toxic drug poisoning at 5.4 times the rate of other residents in 2025. The risk lies in these conditions rather than in First Nations people as a “risk group”, and deaths fell by a third between 2024 and 2025.
  • 7.4 Alexander’s dislocation theory holds that the sustained absence of psychosocial integration is painful and linked to addiction, which directs attention to policies that break the ties between people, families, cultures and land. Its concepts are broad and hard to measure, which makes the theory difficult to test against competing explanations, so it remains largely speculative.
  • 7.5 The BC Coroners Service’s 2023 death review panel wrote that fear of criminal sanction leads people to hide their substance use, and hidden use often means using alone, where no one can respond to an overdose. Fewer than 20 percent of opioid toxicity deaths among Ontario construction workers occurred where someone was known to be present to intervene.
  • 8.1 A rate ratio can be approximated as (D ÷ P) ÷ [(100 − D) ÷ (100 − P)] from a group’s share of deaths and share of the population; for First Nations people in 2025 this gives about 5.3, close to the published 5.4. The answer depends on the comparison population: Ontario construction workers give about 2.3 against the whole population and 1.1 against employed people, so the comparison must be justified and, where possible, adjusted for age and sex.

Reflection

The First Nations Health Authority (FNHA) reported that 289 First Nations people died from toxic drug poisoning in British Columbia in 2025, a decrease of 33.3 percent from 433 in 2024. First Nations people made up 3.4 percent of the province’s population and 15.8 percent of toxic drug deaths, and the FNHA reported that they died at 5.4 times the rate of other BC residents (9.1 times for First Nations females compared with other female residents). The FNHA data include only Status First Nations people and their status-eligible descendants. An independent review led by Mary Ellen Turpel-Lafond, In Plain Sight (2020), found that 26 percent of Indigenous respondents reported always being treated as if they were drunk, or being asked about substance use, when accessing health care, compared with 5 percent of non-Indigenous respondents. More than 150,000 First Nations, Inuit and Métis children attended residential schools. In 150 to 250 words, (a) use the formula RR = (D ÷ P) ÷ [(100 − D) ÷ (100 − P)] to approximate the rate ratio and explain why it differs slightly from 5.4; (b) explain, using at least two social determinants, which structural conditions help to explain the gap that the ratio describes; and (c) note one limitation of the FNHA data.

Model answer

With D = 15.8 and P = 3.4, the representation ratio is 15.8 ÷ 3.4 = 4.65. For everyone else it is 84.2 ÷ 96.6 = 0.872, so the rate ratio is 4.65 ÷ 0.872 = 5.33. The published 5.4 is slightly higher because the percentages are rounded: a population share of 3.35 percent, which also rounds to 3.4, would give about 5.4, so a ratio rebuilt from rounded percentages is only an approximation.

The ratio describes the size of a gap, and the conditions created by what has been done to First Nations people help to explain it. Colonial policies separated more than 150,000 children from their families through residential schools, and McCreary’s Young Indigenous Research Team found that First Nations youth with a family history of residential school attendance were more likely to have experienced abuse, government care and bereavement by overdose. Racism in health care, documented by In Plain Sight in 2020, can deter people from seeking care when they expect to be treated as drunk or questioned about substance use, and it can lead to under-treated pain. Combined with poverty and a toxic supply, these conditions raise the risk of death. The fall in deaths between 2024 and 2025 also shows that the gap can change.

One limitation is that the data count only Status First Nations people and their status-eligible descendants, so they omit First Nations people without status and other Indigenous peoples, such as Métis and Inuit people living in British Columbia.

Minimum 20 characters required.

✓ Reflection saved
Knowledge Check: this section

Question 1: What is the alcohol harm paradox?

The paradox is that people with low socioeconomic status experience more alcohol-attributable harm from similar or lower amounts of drinking. The heavy episodic drinking option is tempting, but Probst et al. (2020) found that alcohol use explained up to 27 percent of inequalities in mortality, not all of them.

Question 2: A group makes up 3.4 percent of a population and 15.8 percent of deaths. Using Equation 10.1, what is the approximate rate ratio compared with everyone else?

The representation ratio is 15.8 ÷ 3.4 = 4.65, the ratio for everyone else is 84.2 ÷ 96.6 = 0.872, and 4.65 ÷ 0.872 = 5.3. The answer 4.6 is tempting because it is the first step of the calculation, but it compares the group with the whole population, including the group itself.

Question 3: In the Adverse Childhood Experiences study (Dube et al., 2003), how did people who reported five or more adverse experiences compare with people who reported none?

People with five or more adverse childhood experiences were 7 to 10 times more likely to report illicit drug use problems, addiction and injection drug use. The doubling option is tempting, but the study found that each single adverse experience increased early initiation by two to four times, and risk rose with the number of experiences.

Question 4: How does this lesson interpret the FNHA finding that First Nations people died at 5.4 times the rate of other BC residents in 2025?

The lesson reads the ratio as describing the size of a gap, and it names colonial policies, racism, poverty and the toxic supply as the structural conditions that help to explain it. The claim that the gap has widened is wrong: the ratio fell from 6.5 in 2024 to 5.4 in 2025, and the FNHA data cover only Status First Nations people and their descendants in British Columbia.
Section 4 of 5

Populations with Unique Vulnerabilities or Needs

⏱ Estimated reading time: 40 minutes
Section 4 of 5

Populations with Unique Vulnerabilities or Needs

Seven groups, the evidence behind their risks, and what they need.

Two terms

Vulnerabilities and needs

Vulnerability

A circumstance that raises the chance of harm, such as lost tolerance after prison.

Need

What a service must provide to reach a group, such as care without fear of losing a child.

Youth
Older adults
Pregnant people
Indigenous peoples
2SLGBTQ+ people
Workers in trades
People leaving prison
Youth and older adults

Risk at both ends of life

Youth

In the 2023 BC survey, among 16- to 18-year-olds, 30% who first used cannabis at 12 or younger needed help, compared with 8% who started at 15 or older (Smith et al., 2025).

Older adults

Share of Canadian opioid deaths among people 60 or older: 9% in 2016, 15% in 2025 (PHAC, 2026). Slower alcohol metabolism and medication interactions add risk (Butt et al., 2020).

Pregnant people and Indigenous peoples

Care that keeps people connected

Pregnant people

Fear, stigma and possible loss of custody are common barriers to care (Choi et al., 2022). Rooming-in is linked to less withdrawal treatment and shorter stays (MacMillan et al., 2018).

Indigenous peoples

The FNHA weaves culture, teachings, community and land into harm reduction (FNHA, 2026).

2SLGBTQ+ people

Minority stress and family connection

Patterns

Higher substance use among sexual minority youth, with the largest gaps among plurisexual youth and among girls and young women (Mereish et al., 2026).

Explanation

Stigma, prejudice and discrimination create a stressful environment (Meyer, 2003); support for parts of the model is mixed (Pellicane et al., 2025). Family connection is linked to fewer harms (Saewyc et al., 2024).

Workers and people leaving prison

Hidden risk and lost tolerance

7.9%of Ontario opioid deaths among people who had worked in construction, 2017 to 2020 (Gomes et al., 2022)
1 in 10Ontario adult drug toxicity deaths within a year of release from prison, 2006 to 2013 (Groot et al., 2016)

Across eight countries, death rates after release were highest in the first week, led by alcohol and other drug poisoning (Borschmann et al., 2024).

Bringing it together

Before the final assessment

  • Work through the Applying this lesson task on Ontario construction workers.
  • Complete the reflection and the knowledge check for this section.
  • Then take the final assessment for Lesson 10.

Content note. This section discusses drug toxicity deaths, including deaths of young people and of people recently released from prison, and refers to residential schools. If you would like support, the 9-8-8 Suicide Crisis Helpline is available by call or text at 9-8-8, the National Overdose Response Service can be reached at 1-888-688-6677, and the Hope for Wellness Help Line for Indigenous people can be reached at 1-855-242-3310.

Introduction

Public health also has to plan for particular groups whose circumstances make them more vulnerable to harm, or whose needs differ from those of the general population. This section examines seven such groups: youth, older adults, pregnant people, Indigenous peoples, 2SLGBTQ+ people, workers in high-risk trades and people leaving prison.

Learning Objectives

  • 9.1 Explain what it means for a population to have unique vulnerabilities or needs, and why the idea should be applied without blaming groups.
  • 9.2 Describe how age at first use and adolescent development shape risks for youth, using BC survey and coroner data.
  • 9.3 Explain why older adults face distinct risks from alcohol, medications and the unregulated supply, using national surveillance data.
  • 9.4 Describe the needs of pregnant people who use substances, including the effect of fear of child protection involvement, and one Canadian model of care.
  • 9.5 Describe the strengths and self-determined approaches that First Nations organizations and youth identify.
  • 9.6 Describe the patterns of substance use and the role of minority stress among 2SLGBTQ+ people.
  • 9.7 Explain the risks facing workers in high-risk trades.
  • 9.8 Explain the risks facing people leaving prison.
  • 9.9 Compare the vulnerabilities and needs of the seven populations and apply the lesson to a Canadian case.

9.1 Vulnerabilities and Needs

Two terms need care. A vulnerability is a circumstance that raises the chance of harm, such as a fall in tolerance after release from prison or a body that clears alcohol more slowly with age. A need is something a service or policy must provide for a group to be reached, such as care that does not lead to a child’s removal or a clinician who does not assume an Indigenous patient has been drinking. Because Section 3 showed that many vulnerabilities are produced by social conditions, describing a group’s vulnerability says nothing about its members’ character. HSCI 402 develops each population in depth.

9.2 Youth

Adolescence is the period in which most people first try alcohol, cannabis or nicotine. Lesson 9 described adolescent brain maturation and why age at first use matters; this part adds Canadian evidence on what early use predicts.

The 2023 BC AHS found a steep relationship between age at first cannabis use and later problems. Among youth aged 16 to 18 who had used cannabis, 30 percent of those who first used it at 12 or younger reported that their use had reached a point in the past year where they needed help, compared with 23 percent of those who started at 13 or 14 and 8 percent of those who started at 15 or older. Youth aged 16 to 18 who delayed their first use until at least 15, or who had not used recently, reported better health and well-being than those who started earlier or used recently (Smith et al., 2025). Early use may be a marker of other difficulties as well as a cause of later problems.

The toxic supply reaches young people too. In 2025, 26 youth aged 18 and younger died of suspected unregulated drug toxicity in British Columbia, up from 21 in 2024 (BC Coroners Service, 2026). Young people themselves have described what they need. In workshops to discuss the 2023 survey results, youth asked for regular, non-judgemental education about substances, including their effects on the brain and body, delivered at school, and one participant said that adults should “think of the person behind the substance” (McCreary Centre Society, 2025).

9.3 Older Adults

The age profile of harm in Canada has been shifting older, and bodies, medications and social circumstances all change with age in ways that alter risk.

National surveillance shows the shift. Figure 10.7 compares the age distribution of apparent opioid toxicity deaths in Canada in 2016, the first year of national reporting, and 2025. The share of deaths among people aged 60 or older rose from 9 percent to 15 percent, and the share among people in their twenties fell (Public Health Agency of Canada [PHAC], 2026).

Apparent opioid toxicity deaths in Canada by age group, 2016 and 2025 2016 2025 220 to 19 181320 to 29 272630 to 39 212540 to 49 221950 to 59 91560 or older Age group (years); bar labels are percentages of all apparent opioid toxicity deaths in that year
Figure 10.7. Percentage of apparent opioid toxicity deaths in Canada by age group, 2016 and 2025. In 2025 there were 5,704 such deaths. Percentages may not sum to 100 because of rounding. Data from the Public Health Agency of Canada’s opioid- and stimulant-related harms surveillance, data reported as of 12 August 2026 (PHAC, 2026).

Alcohol and prescribed medications raise distinct risks with age. The Canadian Coalition for Seniors’ Mental Health notes that, because of physiological changes, older adults are much more vulnerable to the negative effects of substance use, and that alcohol, cannabis and the medications commonly prescribed to older adults for pain, anxiety, sleeplessness and depression can affect cognition, emotions and physical health (Canadian Coalition for Seniors’ Mental Health [CCSMH], 2024). Its guidelines on alcohol use disorder in older adults, commissioned by Health Canada, explain that older adults generally metabolize alcohol more slowly than younger adults, so alcohol has stronger effects even at low levels of drinking, and that many older adults take several medications that can interact harmfully with alcohol. The guidelines advise older adults who drink to keep well below the national low-risk drinking guidelines then in force (since replaced by Canada’s Guidance on Alcohol and Health in 2023) and recommend that primary care providers review alcohol use every year in a simple, neutral way that makes the conversation normal (Butt et al., 2020). Lesson 4 described the risks of combining depressants and a trial of benzodiazepine deprescribing in older adults.

9.4 Pregnant People

The needs of pregnant people who use substances are shaped as much by how services respond as by the substances themselves.

Canada’s Guidance on Alcohol and Health states that when a person is pregnant or trying to get pregnant, there is no known safe amount of alcohol use (CCSA, 2023). Many pregnant people who use substances, however, face barriers to care that come from fear. A 2022 scoping review of 41 articles by Choi and colleagues found that fear, stigma and the possibility of losing custody of children were among the main barriers to treatment for pregnant women and mothers (Choi et al., 2022). A 2026 qualitative study based on seven interviews with parents receiving opioid agonist therapy in Newfoundland and Labrador shows related concerns in Canada: participants described inconsistent prenatal education about newborn withdrawal and hospital care, concerns about child protection involvement and judgment shaped what they needed to know, and stigma was among the barriers they met in hospital (Hand et al., 2026). Case 10.4 describes one Canadian model that addressed this fear directly.

Case 10.4: Rooming-in at BC Women’s Hospital

In the early 2000s, BC Women’s Hospital in Vancouver cared for opioid-exposed newborns in the same room as their mothers, a practice called rooming-in. Abrahams and colleagues compared 32 women in Vancouver who had used heroin or methadone during pregnancy and whose babies roomed in between October 2001 and December 2002 with a historical group of 38 women and a concurrent group of 36 women cared for at a neighbouring community hospital (Abrahams et al., 2007).

Rooming-in was associated with a lower need for treatment of neonatal abstinence syndrome (adjusted relative risk 0.40 compared with the historical group and 0.39 compared with the concurrent group), with shorter newborn hospital stays, and with a higher likelihood that babies were discharged in their mothers’ custody (relative risk 2.23 and 1.52). The study was small and retrospective, so the groups may have differed in ways the analysis could not measure (Abrahams et al., 2007). A 2018 meta-analysis of six studies found consistent evidence that rooming-in, compared with care in a neonatal intensive care unit, reduced the use of medication for withdrawal and shortened hospital stays (MacMillan et al., 2018). HSCI 402 develops prenatal substance use care and fetal alcohol spectrum disorder.

9.5 Indigenous Peoples

Section 3 described the colonial policies and racism behind the disproportionate burden of the toxic drug crisis among First Nations people in British Columbia. This part turns to needs and strengths, drawing on First Nations organizations and youth.

Indigenous peoples in Canada include First Nations, Inuit and Métis peoples, each with distinct histories, cultures and governance. Most of the British Columbia data in this lesson concern First Nations people, and the death review panel noted that a specific strategy must be developed with Indigenous peoples to respect their right to self-determination (BC Coroners Service, 2023). The FNHA, an Indigenous-led health authority, describes harm reduction as offering “compassionate, accessible health services” and states that “a First Nations approach strengthens conventional harm reduction by weaving in culture, teachings, community, and connections to the land and one another”. In 2025 it distributed 6,157 nasal naloxone kits to 87 First Nations communities, provided 95 harm reduction grants of up to $50,000, supported 1,524 virtual sessions with psychiatrists and addictions specialists, and reported that 3,968 First Nations people accessed opioid agonist therapy (FNHA, 2026). Its With Open Arms toolkit lists culturally supportive options such as Nation-based counselling, ceremonies, teachings and the guidance of Elders (FNHA, 2025).

Strengths identified by First Nations youth

The McCreary Centre Society’s Young Indigenous Research Team analyzed the 2023 BC AHS responses of First Nations youth and emphasized strengths alongside the effects of colonialism. Youth who could speak at least a few words of an Indigenous language were more likely to feel proud of their background, eat traditional foods and take part regularly in cultural or traditional activities. Youth who felt quite or very proud of their background were more likely to feel hopeful for their future (52 percent compared with 43 percent), as were youth who felt quite or very much a part of their community (69 percent compared with 33 percent) and youth who had a teacher who cared about them (61 percent compared with 27 percent). The team concluded that First Nations youth are resilient and that, with support and connection, particularly to culture, community and language, they can experience more positive health and well-being (McCreary Centre Society’s Young Indigenous Research Team, 2026). These findings concern well-being and hope, and they match the protective role of cultural activities and community connection described in Section 2.

9.6 2SLGBTQ+ People

Two-Spirit, lesbian, gay, bisexual, transgender, queer and other sexual and gender minority (2SLGBTQ+) people are a diverse population whose patterns of substance use differ by group.

A 2026 meta-analysis of 304 studies by Mereish and colleagues found higher substance use among sexual minority youth than among heterosexual youth, with larger gaps among bisexual and other plurisexual youth and among girls and young women (Mereish et al., 2026). Canadian survey data show related patterns among adults: a 2025 analysis of the Canadian Community Health Survey from 2009 to 2014 found elevated odds of binge drinking, illicit drug use and cannabis use among lesbian and bisexual women compared with heterosexual women, the largest health disparities overall among bisexual women, and decreasing substance use over time among gay and bisexual men (Bellows et al., 2025). Among BC youth in 2023, about one in three trans boys and nonbinary youth had tried cannabis, significantly more than cisgender youth, while trans girls and questioning youth did not differ significantly from cisgender youth (Saewyc et al., 2024).

The minority stress model offers one explanation. Ilan Meyer (2003) proposed that stigma, prejudice and discrimination create a hostile and stressful social environment for lesbian, gay and bisexual people, and he described the stress processes involved: experiences of prejudice, expectations of rejection, hiding and concealing one’s identity, internalized homophobia and the coping processes that can soften these effects. A Canadian study of 222,548 adults aged 18 to 59 supported part of this account: sexual minority respondents were more likely to describe their lives as stressful and their sense of community belonging as weak, and perceived life stress partly explained their higher odds of mood and anxiety disorders and of these disorders co-occurring with heavy drinking (Pakula et al., 2016). Support for other parts of the model is mixed: a 2025 meta-analysis of 36 studies of transgender and gender-diverse people found small associations between discrimination and substance use, and no significant associations for expected rejection, concealment or internalized transphobia (Pellicane et al., 2025). These were cross-sectional associations, and research on minority stress and substance use is still evolving. Coping motives, introduced in Section 1, offer one possible route from minority stress to substance use. Section 2 showed the protective side: high family connectedness was associated with much lower odds of negative consequences of substance use among trans and nonbinary youth (Saewyc et al., 2024). HSCI 214 develops minority stress theory.

9.7 Workers in High-Risk Trades

Public discussion of the toxic drug crisis often overlooks people who are employed. Coroner data in two provinces show that workers in physically demanding trades carry a large share of deaths.

In British Columbia, among 2025 unregulated drug deaths where the industry of employment was known, the two most common industries of current or past employment were trades, transport and equipment operators, and sales and service (BC Coroners Service, 2026). In Ontario, a 2022 report by the Ontario Drug Policy Research Network, the Office of the Chief Coroner and Public Health Ontario examined opioid toxicity deaths among people who had worked in construction. Its authors proposed several reasons for the industry’s vulnerability: physically demanding work that leads to injuries and chronic pain, long hours in precarious and stressful jobs, a workforce made up mostly of men, and an industry culture in which job insecurity can lead workers to under-report injuries and return to work quickly, leaving pain unresolved. Temporary workers, who often lack health benefits, face these pressures most (Gomes et al., 2022). The task card at the end of this section uses the report’s findings.

9.8 People Leaving Prison

Release from custody is a dangerous period for a person who uses drugs. Lesson 5 explained the pharmacology: tolerance falls during time in custody, so a dose a person once tolerated can stop their breathing after release.

A review of coroners’ files in Ontario identified 6,978 drug toxicity deaths among adults between 2006 and 2013 and matched them with provincial correctional records. Of these deaths, 702, or one in ten, occurred within a year of release from provincial incarceration; 20 percent of those 702 deaths occurred within one week of release, and 77 percent involved one or more opioids. In 318 of the opioid-involved deaths, the authors judged that intervention by another person may have been possible (Groot et al., 2016). A 2024 meta-analysis of 18 cohort studies in eight countries, including Canada, found that death rates after release were highest in the first week, when alcohol and other drug poisoning was the leading cause of death (Borschmann et al., 2024). The first days and weeks after release therefore need planning before a person leaves custody. Lesson 11 examines incarceration and the overrepresentation of Indigenous and Black people in Canadian prisons, and HSCI 402 develops harm reduction in custody and care after release.

9.9 Comparing Populations and Applying the Lesson

The seven populations differ in the source of their vulnerability and in what services must offer to reach them. Table 10.7 brings the evidence from this section together.

Table 10.7. Vulnerabilities, needs and Canadian evidence for seven populations discussed in this section.

PopulationMain vulnerabilityA distinct needCanadian evidence
YouthEarly first use during brain developmentNon-judgemental education and support30% of BC youth who started cannabis at 12 or younger needed help, compared with 8% who started at 15 or older
Older adultsSlower metabolism; several medicationsRoutine, neutral review of drinking and medicationsShare of opioid deaths among people 60 or older rose from 9% (2016) to 15% (2025)
Pregnant peopleFear of child protection involvement deters careCare that keeps parents and newborns together where safeRooming-in in Vancouver (2001 to 2002) associated with more babies discharged in mothers’ custody
Indigenous peoplesColonial policies, racism in careSelf-determined, culturally grounded servicesFirst Nations people died at 5.4 times the rate of other BC residents (2025)
2SLGBTQ+ peopleMinority stressAffirming services and family supportElevated binge drinking and drug use among lesbian and bisexual women (survey data, 2009 to 2014)
Workers in tradesInjury, pain, precarious work, using alonePain care and harm reduction that reach workers7.9% of Ontario opioid deaths (2017 to 2020) among construction workers
People leaving prisonLost tolerance after custodyPlanning before releaseOne in ten Ontario adult drug toxicity deaths within a year of release

Use Interactive 10.3 to test your recall of the evidence by matching each population to the finding that describes it.

Interactive 10.3
Match each population to its evidence

Each statement summarizes Canadian evidence from this section. Choose the population it describes, then check your answers. Each population is used once.

Twenty percent of these deaths occurred within one week, and most involved opioids.
People leaving prison. Groot et al. (2016) found that 20 percent of the drug toxicity deaths within a year of release in Ontario occurred in the first week, when tolerance is lowest.
Alcohol is metabolized more slowly, so its effects are stronger even at low levels, and several medications can interact with it.
Older adults. The Canadian guidelines on alcohol use disorder among older adults describe slower metabolism and medication interactions (Butt et al., 2020).
Injuries, chronic pain and job insecurity that discourages reporting injuries; most deaths occurred at home.
Workers in high-risk trades. The Ontario construction report linked injuries, pain and precarious work to opioid deaths, most of which occurred in private residences (Gomes et al., 2022).
Concerns about child protection involvement and judgment shaped what parents receiving opioid agonist therapy needed to know.
Pregnant people. Parents receiving opioid agonist therapy in Newfoundland and Labrador described these concerns (Hand et al., 2026), and rooming-in at BC Women’s Hospital is one model that keeps families together.
Those who started at 12 or younger were almost four times as likely to need help as those who started at 15 or older.
Youth. Among BC youth aged 16 to 18, 30 percent of those who first used cannabis at 12 or younger needed help in the past year, compared with 8 percent of those who started at 15 or older (Smith et al., 2025).
Stigma, expectations of rejection and concealment create stress that is linked to higher odds of heavy drinking alongside mood and anxiety disorders.
2SLGBTQ+ people. This is the minority stress model (Meyer, 2003), partly supported in Canadian data by Pakula et al. (2016).
An approach that weaves culture, teachings, community and connections to the land into harm reduction.
Indigenous peoples. This is how the First Nations Health Authority describes a First Nations approach to harm reduction (FNHA, 2026).
Applying this lesson: Opioid deaths among Ontario construction workers

A 2022 report by the Ontario Drug Policy Research Network, the Office of the Chief Coroner for Ontario and Public Health Ontario found that 428 people with a history of work in construction died of opioid toxicity in Ontario between July 2017 and December 2020, 7.9 percent of all opioid toxicity deaths in the province, although construction workers made up 3.6 percent of Ontario’s population in 2021. Among those who died between 2018 and 2020, 98.4 percent were male, 57.7 percent were employed when they died (compared with 11.7 percent of people without construction work history), 37.2 percent had a history of chronic pain, and about one in four had received a prescription opioid for pain in the year before death. Unregulated fentanyl and cocaine were the substances most often involved, and alcohol directly contributed to about one in five deaths. Most deaths (78.7 percent) occurred in a private residence, usually the person’s own home, fewer than 2 percent at construction sites, and fewer than 20 percent occurred where someone was known to be present to intervene (Gomes et al., 2022).

Using these findings: (1) name one motive or psychological account from Section 1 that could explain part of the pattern, and the evidence that supports it; (2) place one factor at each of the four levels of the social ecological model; (3) identify two social determinants from Section 3 at work, including criminalization; and (4) use Calculator 10.1 to compare the result with the two denominators, population and employed people, and explain which comparison you think is fairer and why.

Summary

This section applied the lesson to seven populations. In the 2023 BC survey, early first use was strongly associated with later need for help among youth; slower metabolism and multiple medications raise risk for older adults, whose share of Canadian opioid deaths grew between 2016 and 2025; and a review identified fear, stigma and possible loss of custody as common barriers to care for pregnant people. First Nations organizations and youth identify culturally grounded harm reduction and strengths in culture, language and community. Minority stress may help explain some higher rates of substance use among 2SLGBTQ+ people, although support for parts of the model is mixed. Workers in high-risk trades combine injury, pain, precarious work and using alone, and people leaving prison face sharply higher risk in the first weeks after release.

Across the lesson, the same thread runs from motives to populations. People use substances for reasons that are learned in relationships and shaped by communities and policies, and the harm that follows depends heavily on the social conditions around them. The reflection and knowledge check below review this section, and the final assessment brings the four sections together.

Key Takeaways

  • 9.1 A vulnerability is a circumstance that raises the chance of harm, and a need is something a service or policy must provide for a group to be reached. Because many vulnerabilities are produced by social conditions, describing a group’s vulnerability says nothing about its members’ character.
  • 9.2 In the 2023 BC survey, among youth aged 16 to 18 who had used cannabis, 30 percent of those who first used it at 12 or younger needed help in the past year, compared with 8 percent of those who started at 15 or older, although early use may also be a marker of other difficulties. The toxic supply reaches young people too, and youth ask for regular, non-judgemental education about substances.
  • 9.3 The share of Canadian apparent opioid toxicity deaths among people aged 60 or older rose from 9 percent in 2016 to 15 percent in 2025. Older adults metabolize alcohol more slowly and often take several medications that can interact with it, so Canadian guidelines recommend a routine, neutral yearly review of drinking in primary care.
  • 9.4 There is no known safe amount of alcohol in pregnancy, and a 2022 review identified fear, stigma and possible loss of custody as common barriers to care for pregnant people. Rooming-in at BC Women’s Hospital in 2001 to 2002 was associated with less treatment for neonatal abstinence syndrome and more babies discharged in their mothers’ custody, although the study was small and retrospective.
  • 9.5 The FNHA describes a First Nations approach that strengthens harm reduction by weaving in culture, teachings, community and connections to the land, and strategies must be developed with Indigenous peoples to respect their self-determination. First Nations youth connected to culture, community and language reported more pride and hope for their future.
  • 9.6 Patterns of substance use differ within 2SLGBTQ+ populations, with elevated use among lesbian and bisexual women and among trans boys and nonbinary youth. Meyer’s minority stress model links these patterns to stigma, prejudice and discrimination, although support for parts of the model is mixed, and family connectedness is associated with fewer negative consequences of use.
  • 9.7 Workers in physically demanding trades combine injury, chronic pain, long hours, precarious work and an industry culture that can discourage reporting injuries. Most opioid deaths among Ontario construction workers occurred at home, usually with no one present to intervene.
  • 9.8 Tolerance falls during custody, so the period after release is especially dangerous: one in ten adult drug toxicity deaths in Ontario between 2006 and 2013 occurred within a year of release, and 20 percent of those within the first week. Planning for those first days needs to begin before a person leaves custody.
  • 9.9 The seven populations differ in the source of their vulnerability and in what services must offer to reach them, so a single general message is unlikely to reach all of them. Applying the lesson to a case such as opioid deaths among Ontario construction workers means combining motives, ecological levels, social determinants and a carefully chosen comparison population.

Reflection

Two Ontario studies describe groups at high risk of drug toxicity death. Groot et al. (2016) found that of 6,978 adult drug toxicity deaths in Ontario between 2006 and 2013, 702 (one in ten) occurred within a year of release from provincial incarceration; 20 percent of these occurred within the first week after release, and 77 percent involved opioids. Gomes et al. (2022) found that people who had worked in construction accounted for 7.9 percent of Ontario’s opioid toxicity deaths between July 2017 and December 2020, while construction workers made up 3.6 percent of the population in 2021. Among construction workers who died between 2018 and 2020, 98.4 percent were male, 57.7 percent were employed when they died, 37.2 percent had a history of chronic pain, 78.7 percent died in a private residence, and fewer than 20 percent died where someone was known to be present to intervene. In 150 to 250 words, compare the main vulnerability of each group, identify one distinct need for each, and explain why a single general overdose prevention message would likely fail to reach both groups.

Model answer

The main vulnerability of people leaving prison is lost tolerance. During custody their tolerance to opioids falls, so a dose they once used can stop their breathing, which helps explain why one in five of the deaths within a year of release happened in the first week. Their distinct need is planning that starts before release, so that care and support are in place for those first days.

Construction workers face a different combination. Physically demanding work produces injuries and chronic pain, more than a third of those who died had chronic pain, and job insecurity can discourage workers from reporting injuries or taking time to recover. Most were employed and died at home, usually with no one present, and the report’s authors suggested that stigma and fear of losing work may contribute to using alone. Their distinct need is pain care and harm reduction that reach working men through workplaces and trades organizations, without threatening their jobs.

A single general message would probably miss both groups. It would reach people leaving prison too late if it were not linked to the moment of release, and it would not reach employed workers who do not see themselves as people who use drugs and who may not use services designed mainly for people experiencing homelessness. Each group needs a message delivered at the right time, in the right setting, in terms that fit its circumstances.

Minimum 20 characters required.

✓ Reflection saved
Knowledge Check: this section

Question 1: Among BC youth aged 16 to 18 who had used cannabis, what share of those who first used it at 12 or younger said they had needed help for their use in the past year?

Thirty percent of early starters needed help, compared with 23 percent of those who started at 13 or 14 and 8 percent of those who started at 15 or older. Seventeen percent is tempting, but it is the share of all youth who had tried cannabis who reported needing help.

Question 2: How did the share of apparent opioid toxicity deaths in Canada among people aged 60 or older change between 2016 and 2025?

Public Health Agency of Canada data show the share among people aged 60 or older rose from 9 percent in 2016 to 15 percent in 2025. The option of 2 percent is tempting because it matches the share among people aged 0 to 19, which stayed the same.

Question 3: In the Ontario study by Groot et al. (2016), what share of adult drug toxicity deaths occurred within a year of release from provincial incarceration?

Of 6,978 adult drug toxicity deaths between 2006 and 2013, 702, or one in ten, occurred within a year of release. One in three is tempting if it is confused with other large shares in the lesson, but the study’s own conclusion was one in ten, with 20 percent of those deaths in the first week.

Question 4: Which description of the minority stress model (Meyer, 2003) is accurate?

Meyer proposed that stigma, prejudice and discrimination create a hostile and stressful social environment that harms health. Concealment is tempting because it is one of the stress processes Meyer described, but it is only one of several, alongside prejudice events, expected rejection and internalized stigma.
Section 5 of 5

Final Assessment

⏱ Estimated time: 25 minutes

Bringing It All Together

This lesson asked why people use substances and why the risks that follow differ so much. Section 1 began with individuals. Cooper’s model sorts drinking motives into enhancement, social, coping and conformity motives, and reviews link coping and enhancement motives to drinking problems, with negative reinforcement one explanation for the coping link. In the 2023 BC Adolescent Health Survey, young people most often reported experimenting, having fun and joining friends, while reasons tied to stress and sadness were common among those who used daily. Social learning theory, outcome expectancies and the self-medication hypothesis each have support in explaining how motives are learned and how use can follow emotional pain, although research on which comes first is still evolving.

Section 2 placed individuals inside relationships, communities and society. The social ecological model and the risk-focused approach of Hawkins and colleagues organize the evidence, and the 2023 BC Adolescent Health Survey showed that family respect, parental knowledge and school belonging were associated with less cannabis use, while trauma, rural schooling, nearby stores and long working hours were associated with more. In Section 3’s examples, harm was distributed more unequally than use. Poverty, housing, childhood adversity, colonialism, racism, dislocation and criminalization help explain where harm concentrates, with evidence that ranges from consistent reviews to theory that is hard to test, and conditions created by colonial policies and racism help explain why First Nations people in British Columbia died at 5.4 times the rate of other residents in 2025.

Section 4 applied these ideas to seven populations whose vulnerabilities or needs differ. The same message runs through every section: people use substances for reasons learned in relationships and shaped by communities and policies, and the harm that follows depends heavily on the social conditions around them.

Key Takeaways from this lesson

  1. A motive is the purpose a person hopes a substance will serve, distinct from a risk factor. Cooper’s model sorts drinking motives by valence and source into enhancement, social, coping and conformity motives, and reviews link coping and enhancement motives to drinking problems, with negative reinforcement one explanation for the coping link.
  2. In the 2023 BC Adolescent Health Survey, curiosity, fun and friends were the most common reasons young people in British Columbia gave for use, and curiosity sits outside Cooper’s model. Survey data on reasons are cross-sectional and refer to the most recent occasion, so they cannot show whether distress led to frequent use or the reverse.
  3. Social learning theory, outcome expectancies and the self-medication hypothesis each have support in explaining how motives are acquired and why use can follow emotional pain, although research on which comes first is still evolving.
  4. The social ecological model places influences at individual, relationship, community and societal levels and was proposed to keep health promotion from blaming individuals. Risk and protective factors are statistical associations that fill those levels, and a risk factor is not necessarily a cause.
  5. In the 2023 BC Adolescent Health Survey, family respect, parental knowledge and school belonging were associated with less youth cannabis use. Risk tends to accumulate across levels, and supportive adults were associated with less regular cannabis use among youth facing adversity.
  6. Harm depends on the substance, the setting and the help available as well as on use, so in this lesson’s examples, from the alcohol harm paradox to the distribution of toxic drug deaths in British Columbia, substance use harms were distributed more unequally than substance use.
  7. Poverty, housing, childhood adversity, colonialism, racism, dislocation and criminalization help explain where harm concentrates, and structural conditions created by colonial policies and racism help explain the 5.4-fold toxic drug death rate among First Nations people in British Columbia in 2025.
  8. A rate ratio can be approximated from a group’s share of deaths and share of the population, but the comparison population must be justified.
  9. Youth, older adults, pregnant people, Indigenous peoples, 2SLGBTQ+ people, workers in high-risk trades and people leaving prison each have distinct vulnerabilities and needs, and applying the lesson to a Canadian case means bringing motives, ecological levels and social determinants together.

Core Concepts Reviewed

Section 1: motives, valence and source; enhancement, social, coping and conformity motives; negative reinforcement; social learning theory and its four mechanisms; outcome expectancies and the balanced placebo design; the self-medication hypothesis.

Section 2: the social ecological model and its individual, relationship, community and societal levels; risk and protective factors; the Communities That Care domains; accumulation and buffering of risk; factors that predict use compared with factors that predict harm.

Section 3: social determinants of health; the alcohol harm paradox; housing and homelessness; toxic stress and adverse childhood experiences; colonialism and Indigenous-specific racism; rate ratios and representation ratios; dislocation and psychosocial integration; criminalization and using alone.

Section 4: vulnerabilities and needs; early first use among youth; alcohol and medications in older adults; fear of child protection and rooming-in; culturally grounded, self-determined harm reduction; minority stress; injury and pain among workers in trades; lost tolerance after release from prison.

The final reflection asks you to bring all four sections together for one Canadian population.

Reflection

Consider young people in British Columbia with experience of government care. The 2023 BC Adolescent Health Survey found that 42 percent of youth with government care experience had used cannabis, compared with 21 percent of youth without such experience. Among youth with government care experience who had tried cannabis, 43 percent of those who had an adult to help them make or get to appointments had used it on 6 or more days in the past month, compared with 55 percent of those without such an adult. A 2026 report by the McCreary Centre Society’s Young Indigenous Research Team found that 19 percent of First Nations youth had experience of government care or alternatives to care (including care through Indigenous Child and Family Service Agencies), compared with 4 percent of non-First Nations youth, and that First Nations youth with a family history of residential school attendance were more likely to have been in government care (21 percent compared with 10 percent). A 2003 analysis of the Adverse Childhood Experiences study found that adults who reported five or more adverse childhood experiences were 7 to 10 times more likely to report illicit drug use problems than adults who reported none. In 200 to 250 words, integrate the four sections of this lesson: (1) identify a motive or psychological account that could explain part of the higher cannabis use; (2) place at least three factors in different levels of the social ecological model; (3) explain which social determinants are at work, naming structural causes; and (4) identify one need specific to this population.

Model answer

Young people who enter government care have usually experienced serious adversity, such as abuse, neglect or the loss of a home. A coping motive, or self-medication, could explain part of their higher cannabis use: cannabis may relieve stress or sadness, and negative reinforcement makes use more likely each time those feelings return. Social learning may add to this if placements expose young people to peers who use.

Several levels are involved. At the individual level are a history of adversity and the stress it produces. At the relationship level is whether a young person has a trusted adult; youth with an adult to help with appointments were less likely to use regularly. At the community level are the placements, schools and services young people move between, and at the societal level are child welfare law and funding.

The main determinants are childhood adversity, acting through toxic stress, and colonialism. First Nations youth are almost five times as likely as non-First Nations youth to have been in care, a pattern linked to residential schools and the continuing removal of children, which also reflects dislocation from family, culture and land.

A specific need for this population is stable, caring adult relationships that continue across placements and after young people leave care, with culturally grounded support for First Nations youth that respects their communities’ self-determination.

Minimum 30 characters required.

✓ Reflection saved

Final Knowledge Assessment

Final Assessment, this lesson: Why People Use Substances and Why Risks Differ (15 Questions)

Question 1: A hypothetical student says she drinks at parties “because everyone else is drinking and I don’t want to stand out.” Which motive does this express, and which theory best explains how such motives are acquired?

Drinking to avoid standing out has a negative valence and an external source, which is a conformity motive, and social learning theory explains how such motives are learned from peers. A social motive is tempting, but social motives seek a positive reward, and the balanced placebo design is a research method.

Question 2: Which pairing of an ecological level with a 2023 BC Adolescent Health Survey finding is correct?

Family respect describes the closest social circle, the relationship level. A diagnosis is individual, stores near a school are community-level availability, and dating partners belong to the relationship level, so the dating option is tempting but misplaced.

Question 3: Why does this lesson distinguish the factors that predict substance use from those that predict harm?

Harm depends on what is in the substance, how and where it is used, whether someone is present and what services exist, so groups with similar use can suffer very different harm. The claim that harm depends only on frequency is tempting, but the toxic supply makes even occasional use dangerous.

Question 4: The BC Coroners Service’s death review panel reported that the death rate in Vancouver’s Downtown Eastside was 10 times the provincial average, yet about 85 percent of deaths occurred elsewhere. What does this imply?

A high rate shows concentrated risk, while the wide spread of deaths shows that a toxic supply makes use risky everywhere. The option that the neighbourhood accounts for most deaths confuses a high rate with a large share of the total.

Question 5: Using Equation 10.1, construction workers’ 7.9 percent share of Ontario opioid deaths gives a rate ratio of about 2.3 when compared with their 3.6 percent share of the population, and about 1.1 when compared with their 7.2 percent share of employed people. What is the best conclusion?

The two denominators give very different answers, and neither is exact, because the deaths include past workers and construction workers are mostly working-age men. Treating the employed comparison as proof of no extra risk is tempting, but it ignores those mismatches.

Question 6: Which description matches toxic stress as defined by the Center on the Developing Child at Harvard University?

Toxic stress occurs when stress responses stay activated at high levels for long periods without supportive relationships to calm them. The buffered severe reaction is tempting, but that describes tolerable stress.

Question 7: Among BC youth who had used cannabis, 59 percent said the cannabis they last used was shared among a group of friends. Which mechanism of Akers’s social learning theory does this most directly illustrate?

Differential association is the learning of values, techniques and motives through interaction with family, friends and peers, and shared use among friends is a direct example. Conformity is tempting, but it is a motive in Cooper’s model, not one of Akers’s four mechanisms.

Question 8: Which finding from the Manitoba research on self-medication best supports the self-medication hypothesis?

Turner et al. (2018) found that longitudinal data supported a sequence in which a mood or anxiety disorder came first and a substance use disorder followed among people who self-medicated. The equal rates option is wrong, since Bolton et al. (2009) found men were more than twice as likely as women to self-medicate.

Question 9: Why do Canadian guidelines advise older adults who choose to drink to keep well below the national limits?

Butt et al. (2020) explain that older adults generally metabolize alcohol more slowly and often take several medications that interact with it. The treatment option is wrong: the same guidelines describe medications and psychosocial treatment for alcohol use disorder in older adults.

Question 10: Abrahams et al. (2007) studied rooming-in for opioid-exposed newborns at BC Women’s Hospital. With what was rooming-in associated?

Rooming-in was associated with a lower need for withdrawal treatment, shorter stays and more babies discharged in their mothers’ custody. The no-difference option is tempting because the study was small and retrospective, but it did find differences, which still need cautious interpretation.

Question 11: Which statement summarizes Alexander’s dislocation theory as presented in this lesson?

Alexander held that psychosocial integration is essential to well-being and that its sustained absence, dislocation, is painful and linked to addiction in free-market societies. The pharmacology option is the view the theory was written to challenge.

Question 12: How does this lesson explain criminalization as a determinant of substance use harm?

The BC Coroners Service’s 2023 death review panel wrote that fear of criminal sanction or public disapproval leads people to hide their substance use, and hidden use often means using alone, where no one can respond to an overdose. The outdoor option is wrong: in 2025, 48 percent of BC unregulated drug deaths occurred in private residences and 20 percent outdoors.

Question 13: Which statement about First Nations youth in the 2023 BC Adolescent Health Survey is consistent with this lesson?

Youth who felt proud of their background were more likely to feel hopeful about the future (52 percent compared with 43 percent). The discrimination option is wrong, since 54 percent had experienced discrimination in the past year.

Question 14: Christiansen et al. (1989) measured alcohol expectancies in young adolescents and followed them for a year. Why did this design strengthen a causal interpretation?

Measuring expectancies first showed that they came before the drinking, which makes it less likely that drinking produced the beliefs. Random allocation is tempting, but the study was observational, so other explanations remain possible.

Question 15: Which pairing of a population with a need is supported by the evidence in this lesson?

One in five deaths within a year of release occurred in the first week, so planning must begin before release. The worksite option is tempting, but most construction worker deaths occurred at home, and fewer than 2 percent occurred at construction sites.
✦ Complete the final reflection above before submitting

Congratulations!

You have successfully completed this lesson: Why People Use Substances and Why Risks Differ.

You can now explain the main reasons people give for using substances and the psychological theories behind them, sort risk and protective factors into the levels of the social ecological model using Canadian youth data, explain how social determinants such as poverty, colonialism, racism and criminalization concentrate harm, calculate and interpret a rate ratio, and describe the distinct vulnerabilities and needs of seven populations.

Lesson 11 turns from people to policy. It examines the international drug control system, the war on drugs in the United States and Canada, policing, incarceration and race, and the iron law of prohibition, which helps explain how the unregulated supply described in this course became so toxic.

Reference

Glossary: Key Terms, People & Frameworks

📚 Reference page, available throughout the lesson

Key terms, frameworks and people from this lesson on why people use substances and why the risks that follow differ.

Core Concepts
Motive The purpose a person hopes a substance will serve on a particular occasion, such as having fun, relaxing with friends or relieving worry.
Valence In Cooper’s model, whether a person uses to obtain a positive outcome or to reduce or avoid a negative one.
Source (of a motive) In Cooper’s model, whether the desired outcome is internal (a change in one’s own mood) or external (a social reward or avoiding rejection).
Enhancement motive Using a substance to increase positive feelings or excitement; positive valence and internal source. Linked with heavy drinking and, in a 2021 meta-analysis, with drinking problems.
Social motive Using a substance to obtain social rewards, such as celebrating with friends; positive valence and external source. The most commonly reported drinking motive among young people in a 2005 review.
Coping motive Using a substance to reduce or avoid negative feelings such as stress or sadness; negative valence and internal source. Linked with alcohol-related problems.
Conformity motive Using a substance to avoid social rejection or disapproval; negative valence and external source.
Negative reinforcement Strengthening of a behaviour because it removes something unpleasant, as when drinking relieves worry and so becomes more likely when worry returns.
Outcome expectancy A belief about the effects a substance will have, such as the belief that alcohol makes a person more outgoing.
Differential association In Akers’s social learning theory, the learning of values, techniques and motives connected with a behaviour through interaction with family, friends and peers.
Self-medication hypothesis Khantzian’s proposal that people use substances to relieve painful emotional states and psychiatric symptoms, and that the drug they prefer reflects the feelings they struggle with.
Risk factor A characteristic of a person or their environment that is associated with a higher probability of an outcome, such as substance use problems. An association is not necessarily a cause.
Protective factor A characteristic associated with a lower probability of an outcome, or one that weakens the effect of a risk factor, such as a caring adult in a young person’s life.
Social determinants of health The conditions in which people are born, grow, live, work and age, and people’s access to power, money and resources (World Health Organization).
Alcohol harm paradox The finding that people with low socioeconomic status experience more alcohol-attributable harm than people with high status from similar or lower amounts of drinking.
Toxic stress Prolonged, high activation of the body’s stress response systems without supportive relationships to calm them, which can disrupt development and raise later disease risk.
Adverse childhood experiences Abuse, neglect and household dysfunction in childhood. In the ACE study and in later meta-analyses, more such experiences were associated with higher risks of later substance use and drug use problems.
Indigenous-specific racism Stereotyping, bias and prejudice directed at Indigenous peoples, rooted in colonialism; documented in British Columbia health care by the In Plain Sight review (2020).
Rate ratio The death or disease rate in one group divided by the rate in a comparison group. A ratio of 5.4 means the group’s rate is 5.4 times as high.
Representation ratio A group’s share of deaths divided by its share of the population; it shows how over-represented the group is among deaths.
Psychosocial integration In Alexander’s theory, a balance between a person’s need to belong and their need for autonomy and achievement.
Dislocation In Alexander’s theory, the sustained absence of psychosocial integration, which he argued is painful and linked to addiction.
Minority stress Stress that arises from stigma, prejudice and discrimination directed at a minority group, including expected rejection and concealment of identity.
Rooming-in Caring for a newborn in the same room as the parent, a practice studied at BC Women’s Hospital for opioid-exposed newborns in the early 2000s.
Unsheltered homelessness Living outdoors, in a makeshift shelter, a vehicle or another place not intended for habitation, as defined by the BC Coroners Service.
Frameworks & Tools
Four-factor model of drinking motives Cooper’s (1994) model that crosses valence and source to produce enhancement, social, coping and conformity motives.
Social learning theory Akers’s theory that substance use is learned through differential association, definitions, differential reinforcement and imitation.
Balanced placebo design An experimental design that crosses what participants are told they receive with what they actually receive, to separate drug effects from expectancy effects.
Social ecological model A framework that places influences on health behaviour at individual, relationship, community and societal levels, used by the CDC and derived from earlier ecological models.
Communities That Care A prevention system from the University of Washington that organizes risk factors into community, family, school, and peer and individual domains.
BC Adolescent Health Survey A survey of BC students aged 12 to 19 run by the McCreary Centre Society about every five years since 1992; about 38,500 students took part in 2023.
Dislocation theory Alexander’s (2008) theory linking addiction to the loss of psychosocial integration in free-market societies; it is broad and hard to test, so it remains largely speculative.
Key People
M. Lynne Cooper Psychologist who developed and tested the four-factor model of drinking motives among adolescents (1994).
Ronald Akers Sociologist who developed a social learning theory of deviant behaviour (1977) drawing on Sutherland’s differential association theory and Bandura’s social learning theory.
Edward Khantzian Psychiatrist who proposed the self-medication hypothesis of addictive disorders (1985).
Kenneth McLeroy Lead author of the 1988 ecological model of health promotion, which responded to criticism that lifestyle programs blamed individuals.
J. David Hawkins Prevention researcher who, with Catalano and Miller (1992), set out a risk-focused approach to preventing adolescent alcohol and drug problems.
Bruce Alexander Simon Fraser University psychologist who led the Rat Park experiments and proposed the dislocation theory of addiction (2008).
Ilan Meyer Researcher who proposed the minority stress model to explain higher rates of mental disorders among lesbian, gay and bisexual people (2003).
No matching entries. Try a different search term.