# Lesson 1: Conceptualizing Health, Illness, and Disease

*Companion-podcast transcript • Sarah & Kiffer*  
*Office Hours episode to listen to after working through the lesson*

---

**Sarah:** Welcome back to Office Hours. I'm Sarah.

**Kiffer:** And I'm Kiffer. This episode goes with Lesson one, Conceptualizing Health, Illness, and Disease.

**Sarah:** It's meant for after you've worked through the lesson. We'll bring in some perspective on why these ideas matter in practice, offer some critique, work through the questions students tend to find thorny in this material, and do some extra worked examples, including a couple that are harder than the ones on the lesson page.

**Kiffer:** Along the way there are a few places where we'll ask you to work something out before we give the answer. When that happens, you'll hear a few seconds of quiet, and you can pause the audio if you'd like more time to think it through.

**Sarah:** We'll take five questions. First, how do you tell disease, illness and sickness apart when they point in different directions? Second, how much does the definition of health change the number of people we call healthy? Third, can a wrong theory still produce good public health, and what does that tell us about the models in use today? Fourth, should we keep the World Health Organization's 1948 definition of health? Kiffer and I see that one differently. And fifth, what does cultural safety ask of a health service, and who decides whether it has been achieved?

**Kiffer:** Question one. Disease, illness and sickness are easy to define one at a time, and the difficulty starts when a real person is in front of you and the three point in different directions. So let's try a made-up case. Imagine a thirty-four-year-old warehouse worker we'll call Marco. Eight months ago he had COVID-19. Since then he has had heavy fatigue and trouble concentrating, and he feels worse for a day or two after any hard physical effort. His blood work and scans come back normal, and nobody has given him a diagnosis. He has missed eleven shifts, and his employer has started a disciplinary process over the absences.

**Sarah:** So here's the question for everyone. Which of the three concepts apply to Marco, and how? Think about disease, illness and sickness one at a time. Take a few seconds.

*(Pause)*

**Sarah:** I'll try it. Illness is clearly present, because he feels unwell, and illness is the lived experience. Disease is absent, because the tests are normal and there's no diagnosis. And sickness is present too, because he's off work. He's missing shifts, so he's in the sick role.

**Kiffer:** Two of those need another look. Let's start with sickness. What did Talcott Parsons say the sick role actually involves?

**Sarah:** A temporary release from normal obligations, granted on the condition that the person seeks help and tries to recover. Granted. Okay, I see it. Being absent from work and being excused from work are two different things. His employer is treating the absences as misconduct, so the institution hasn't granted him the role at all.

**Kiffer:** Right. Marco has the absences without the exemption, which is about the hardest position to be in. Sickness is about how other people and institutions respond, and here they're declining to recognize him as sick. Now, what about your call on disease?

**Sarah:** I said disease is absent. The more careful thing to say is that no disease has been detected. Normal results tell us about the tests that were run.

**Kiffer:** That's the right correction. No detectable disease describes the current state of the diagnostic tools, and the long history of ME slash CFS is a warning about treating that as the last word. So the careful summary is clear illness, contested disease, and a sickness role that's being denied. Each of those points to a different response. A clinician might keep investigating, a peer support group might work on the experience of the illness, and a workplace accommodation policy would work on the sickness side.

**Sarah:** There's Canadian survey data on exactly this gap, too.

**Kiffer:** Statistics Canada asked adults about symptoms that were still present three months or more after a confirmed or suspected COVID-19 infection. As of June 2023, about 6.8 percent of Canadian adults, roughly 2.1 million people, were still living with long-term symptoms. Among those people, fewer than six percent had received a diagnosis of post COVID-19 condition. And among adults who had ever had long-term symptoms and were working or in school, about twenty-two percent had missed days, twenty-four days on average.

**Sarah:** So the survey counts the illness directly. The diagnosis, which is the disease label, is rare, and the missed days give a rough view of the sickness side.

**Kiffer:** Yes, and the gap between those numbers matters for what services get planned. A health system that counted only diagnoses would see a small fraction of the people who report the illness, and it would plan its clinics around that small fraction.

**Sarah:** Question two. The lesson argues that the definition of health decides what gets measured. I'd like to see how big that effect can be, so let's put numbers on it.

**Kiffer:** Here's a hypothetical survey of two thousand adults. Nine hundred of them report at least one diagnosed chronic condition, and the other eleven hundred report none. Everyone also answers the self-rated health question: in general, would you say your health is excellent, very good, good, fair or poor? Among the nine hundred with a condition, five hundred and forty rate their health good or better. Among the eleven hundred without a condition, nine hundred and ninety rate it good or better. What share of the sample is healthy under the negative definition, and what share rates their health good or better? Take a few seconds.

*(Pause)*

**Sarah:** Under the negative definition, healthy means no diagnosed condition, so that's eleven hundred out of two thousand, which is fifty-five percent. For self-rated health, I add the two groups who said good or better. Five hundred and forty plus nine hundred and ninety is fifteen hundred and thirty. Divided by two thousand, that's 0.765, or about seventy-seven percent.

**Kiffer:** So the same two thousand people are fifty-five percent healthy on one definition and about seventy-seven percent healthy on another. And look at who moves between them. Five hundred and forty people with a chronic condition count as unhealthy under the negative definition and healthy by their own rating.

**Sarah:** And a hundred and ten people with no diagnosis rate their health fair or poor. The negative definition calls them healthy.

**Kiffer:** Those hundred and ten are the people I'd want to follow up. Idler and Benyamini reviewed twenty-seven community studies in 1997 and found that self-rated health predicted mortality in nearly all of them, even after the studies took account of many specific health measures. So a fair or poor rating with no diagnosis may be picking up something the medical chart hasn't caught yet.

**Sarah:** And the World Health Organization's definition, complete physical, mental and social wellbeing? Where does that leave our sample?

**Kiffer:** Suppose that in our made-up sample, only a hundred and forty people report no condition, rate their health excellent, and report no limits or distress in any area of life. That's a hundred and forty out of two thousand, or seven percent. So for the same people, the share who are healthy runs from seven percent to about seventy-seven percent, depending only on the definition we choose.

**Sarah:** For real Canadian figures, Statistics Canada reported that in 2023 about fifty-two percent of adults aged eighteen and older rated their health very good or excellent, down from about sixty-one percent in 2020. In the same report, about forty-six percent of adults had at least one of a set of selected chronic diseases, such as arthritis, high blood pressure and diabetes.

**Kiffer:** It's worth noticing what you can't do with those two numbers. You can't subtract one from the other, or assume that the people without a chronic disease are the people rating their health highly. They answer different questions, the rating uses a higher cut-off than the good-or-better line in our example, and the overlap between the two groups doesn't appear in those headline figures.

**Sarah:** Here's the harder example, and it has a trap in it. Imagine a made-up support program for five hundred people newly diagnosed with type 2 diabetes. At the start, three hundred of them, sixty percent, rate their health good or better. Two years later, the program reports that sixty-eight percent rate their health good or better, and it describes that as an eight-point improvement in health.

**Kiffer:** Before anyone credits the program, here's a detail from the follow-up. Only four hundred and fifty people answered the second survey, and the fifty who didn't answer had all rated their health fair or poor at the start. If nobody changed their answer at all, what percentage of the four hundred and fifty would rate their health good or better? Take a few seconds.

*(Pause)*

**Sarah:** If nobody changed, the three hundred who said good or better at the start are all still in the sample, because the fifty who left were all in the fair or poor group. So it's three hundred out of four hundred and fifty, which is two thirds, or about sixty-seven percent.

**Kiffer:** So the figure rises from sixty to about sixty-seven percent with nobody's health changing at all. Now, how many people does sixty-eight percent of four hundred and fifty actually represent?

**Sarah:** 0.68 times four hundred and fifty is three hundred and six. So compared with the three hundred we started with, the net gain is six people. Most of the headline comes from the people with poorer health leaving the sample.

**Kiffer:** That's the dropout trap. And even those six people can be read in three ways, which is where the definitions come back in. One reading is real improvement. Another is response shift. People living with a chronic condition often recalibrate what good health means to them, so the yardstick moves while the body may stay the same. And under Huber's adaptive definition, someone who has learned to manage the condition and now rates their health higher may well be healthier, because adapting and self-managing is what health means on that definition.

**Sarah:** So the same six people could be a measurement problem under one definition and the whole point of the program under another.

**Kiffer:** Yes. A good evaluation would report who dropped out and how they differed, add a measure that doesn't depend on self-rating, such as blood sugar control or days of usual activity, and state which definition of health the program was trying to change.

**Sarah:** Question three. Miasma theory was wrong about how disease spreads, and it still gave us sewers, drainage and cleaner water. Students sometimes draw a puzzling conclusion from that, that the theory doesn't matter as long as the intervention works. Is that fair?

**Kiffer:** It's half right. Chadwick's 1842 report and the sanitary measures that followed did save lives, and they did it for the wrong reason. The trouble is that a wrong theory can only take you as far as its overlap with the real cause. A miasma theorist judged danger by smell. A well that smelled fine while sitting downhill from a cesspool gave a strict miasma theorist little reason for concern. The sanitary reforms worked where clearing foul smells and clearing contaminated water happened to be the same job.

**Sarah:** So when an intervention works, that's good evidence for the intervention, and much weaker evidence for the theory behind it.

**Kiffer:** Right, and the same logic applies to the models in use today. Here's a made-up case to test it. A city sees rising emergency visits for childhood asthma, and three proposals reach city council. The first funds a clinic that tests children's lung function and makes sure each child has the right inhaler. The second funds repairs for mould and dampness in social housing. The third funds a school program in which children with asthma learn to recognize early warning signs, plan for bad days, and feel in charge of their condition.

**Sarah:** The first is the biomedical model, a specific disease with a specific treatment. The second is the biopsychosocial model, leaning on its social side, since housing conditions enter the causal story. The third sounds salutogenic. It builds the child's sense that the condition is understandable and manageable, which covers two of the three parts of Antonovsky's sense of coherence.

**Kiffer:** Good. And notice that all three accept the same biology of asthma. So the council's disagreement is about where to intervene, and the biology alone doesn't settle that. It's the same point the lesson makes about germ theory and tuberculosis.

**Sarah:** And a city could fund all three.

**Kiffer:** It could, and many programs do mix them. The skill is noticing which model each proposal comes from, so that each one is judged by the outcome it was built to change. The clinic should be judged on asthma control in its patients, the housing repairs on mould exposure and emergency visits across whole buildings, and the school program on how confident children feel managing their asthma.

**Sarah:** Question four, and this is where Kiffer and I disagree. The World Health Organization's constitution, adopted in 1946 and in force from 1948, defines health as a state of complete physical, mental and social wellbeing, and adds that health is more than the absence of disease or infirmity. I think that for practical public health work, it's time to retire it.

**Kiffer:** Make your case first, Sarah, and then I'll make the other one, because I think both sides have a point.

**Sarah:** The word complete is the problem. Almost nobody is completely well at any moment, so almost everyone is unhealthy, and a definition that puts nearly everyone in the same category can't guide measurement. Huber and her colleagues made that argument in the BMJ in 2011. It also turns ordinary grief and stress into health deficits. I'd use Huber's definition, the ability to adapt and self-manage in the face of social, physical and emotional challenges. It fits a population in which so many adults live with a chronic condition.

**Kiffer:** I agree with every one of those criticisms as a description of a measuring tool. My view is that the 1948 definition was always a poor measuring tool, and that its value is as a statement of what health systems are for. The breadth that makes it impossible to measure is the same breadth that puts mental health and social conditions inside the mandate of health ministries.

**Sarah:** A statement of aims that nobody can meet can also justify almost any expansion of what counts as a health problem. That's one of the main critiques the lesson lists.

**Kiffer:** That's a fair risk. I'd point to a risk on your side as well. An adaptive definition can drift toward asking people to cope with conditions that ought to be changed. If a family adapts well to living in damp, crowded housing, they may count as healthy under Huber's definition, and the housing problem drops out of view. It's a version of the worry the lesson raises about wellness.

**Sarah:** I'd answer that adapting and self-managing depend on having the resources to do it, and a public health team can measure whether people have those resources.

**Kiffer:** That's reasonable, and it's the version of the adaptive definition I'd be comfortable with. I'd keep the 1948 definition as a statement of aims, and use something narrower and measurable for any particular study or program.

**Sarah:** I'd still drop it from teaching as a working definition, but we land on the same practical rule. For any study, survey or policy, write down which definition of health you're using, say why it fits the question, and say what it leaves out.

**Kiffer:** That's a rule I'd be glad to see applied to every definition of health in this lesson.

**Sarah:** Question five. Cultural safety is the concept in this lesson that I think gets flattened most often. People hear the phrase and picture a training workshop.

**Kiffer:** The flattening usually runs three ideas together. Cultural awareness recognizes that cultures differ. Cultural competence builds skills for working across those differences. Cultural safety, the concept that Irihapeti Ramsden developed in nursing education in New Zealand, looks at the power between providers and the people they serve, and it gives the recipient of care the authority to judge whether care was safe.

**Sarah:** Let's test that with a made-up clinic. A community health clinic makes three changes. First, all staff attend a two-hour session on the history of the local First Nations. Second, staff are trained to work with interpreters and to ask patients about traditional medicines without judgment. Third, the clinic sets up an Indigenous patients' council with real authority over how complaints are handled, and it starts asking Indigenous patients directly whether they felt safe during their visits. Which change is awareness, which is competence, and which moves toward cultural safety? Take a few seconds.

*(Pause)*

**Kiffer:** The history session builds awareness. The interpreter and traditional medicine training builds competence, because it's a set of skills. The third change is the one that touches power. It moves some authority to the people receiving care and builds in a way to hear their judgment.

**Sarah:** And here's the trap I'd want people to catch. Even the third change doesn't make the clinic culturally safe by itself. The clinic can't award itself that label. Whether care is safe is decided by the patients' experience, which is why the asking matters as much as the council.

**Kiffer:** Exactly. And there's Canadian evidence of why this matters. In November 2020, the In Plain Sight review, led by Mary Ellen Turpel-Lafond, reported on Indigenous-specific racism in British Columbia's health care system. Nearly twenty-eight hundred people filled out its survey for Indigenous peoples, and eighty-four percent of the Indigenous people who took part reported experiencing some form of discrimination in health care. Those are the patients' own judgments, which is the standard cultural safety asks us to use.

**Sarah:** And if the clinic wanted to survey First Nations patients more formally, the OCAP principles would apply to that data.

**Kiffer:** Right. OCAP stands for ownership, control, access and possession. First Nations communities own their collective information and control how it's collected and used. So even a well-meant patient survey has to be designed with the community, and the community decides how the results are stored and shared.

**Sarah:** Let's pull it together with three things to take away from this episode.

**Kiffer:** First, when disease, illness and sickness point in different directions, name each one separately. Normal test results mean that no disease has been detected, and being off work is a different thing from being granted the sick role.

**Sarah:** Second, the definition of health decides the number. In our made-up survey of two thousand adults, the share counted as healthy ran from seven percent to about seventy-seven percent, depending only on the definition.

**Kiffer:** And third, look for the model behind every claim and every program. A wrong theory can produce a good intervention, a debate about facts is often a debate about models, and whether care is culturally safe is judged by the people receiving it.

**Sarah:** If you'd like more practice, rework the diabetes program example with your own numbers. Change how many people drop out and how they rated their health at the start, and see how much of an apparent improvement the dropouts alone can produce.

**Kiffer:** Next time, it's Lesson two, The Rise of Public Health, where we follow how public health grew into an organized field.

**Sarah:** Take care, everyone.

**Kiffer:** See you in Lesson two.
