What prevalence tells us

Prevalence is an epidemiological measure: the proportion of a defined population experiencing a health problem at a particular point or during a specified period. A population is the defined group of people about whom a conclusion is drawn. It may be defined by place, age or occupation, or by another relevant characteristic.

Prevalence counts existing cases rather than simply tracking the appearance of new ones. It may rise when more people develop the problem or experience it for longer. It can also increase when fewer affected people leave the population through recovery, migration or death. The reverse processes can reduce prevalence.

A point-prevalence estimate refers to one specified moment, whereas a period-prevalence estimate covers a stated interval. They answer slightly different questions and shouldn't be treated as interchangeable.

Explaining change within a population

Change over time may show a genuine shift in the health problem. Exposure to risk may have altered, or protective conditions may have become stronger or weaker. Other possible explanations include changes in prevention, improved treatment, shifts in population composition and migration.

Yet recorded prevalence may change even if the underlying health problem hasn't changed by the same amount. Revised diagnostic criteria and improved detection can increase the number of identified cases. So can greater access to services, increased willingness to report symptoms and more complete record keeping. By contrast, reduced access or greater reluctance to report can hide cases.

A time trend therefore calls for two levels of explanation. First, what may have changed in people’s health? Second, what may have changed in the way cases were defined, found and recorded? The line graph below shows why the same time period can reveal both change within populations and differences between them.

Explaining differences between populations

Differences between populations may come from unequal exposure to harmful conditions or unequal access to protection or treatment. Demographic composition, social expectations and environmental circumstances may also contribute, as may differences in reporting and diagnosis. Which explanation is relevant depends on the health problem selected; no single factor accounts for every prevalence gap.

A comparison is meaningful only when sufficiently similar case definitions, time periods, sampling procedures and measurement methods were used to study the populations. Otherwise, what looks like a health difference may partly reflect a difference in measurement.

Prevalence data establish a pattern, not its cause. When a population with greater exposure to a suspected factor also has higher prevalence, that factor may contribute. However, reverse causation, additional variables and selection effects are still possible. Strong causal reasoning needs a plausible mechanism, along with evidence that competing explanations have been considered.

The central claim

Social learning theory is a psychological theory that explains how people acquire and maintain behaviour by observing others and seeing the consequences linked to that behaviour. It can help explain a health problem when socially learned behaviour contributes to its development, continuation or reduction.

A model is someone whose behaviour, and the consequences of that behaviour, are observed by another person. The model might be a family member, a peer or another socially significant person. Simply observing the behaviour doesn’t guarantee that it will be copied. The observer has to notice and remember it, be capable of performing it and have a reason to do so.

Observation and consequences

Observational learning is a learning process in which someone acquires behaviour by watching a model. When a model performs behaviour connected with a health problem, the observer may learn the action, when to perform it and what outcome to expect.

Imitation is the performance of behaviour that resembles behaviour previously observed in a model. It depends partly on the observer’s relationship with the model and the apparent success of the behaviour. The observer must also feel confident that the behaviour can be performed.

Consequences also shape what happens next. Direct reinforcement is a process in which a consequence makes the individual’s own behaviour more likely to occur again. Vicarious reinforcement is a process in which seeing another person rewarded makes the observer more likely to perform similar behaviour. By contrast, watching someone receive punishment or an unwanted outcome can reduce imitation.

Applying the theory to a health problem

The explanation needs to form a connected chain. Begin by identifying the behaviour relevant to the chosen health problem. Next, explain who models it and what the observer notices. Then consider which consequences are witnessed or personally experienced, and why they make repetition more or less likely.

Social learning may support healthier behaviour too. A person might watch a credible model carry out a protective action, notice desirable consequences and become more willing to reproduce that action. The theory can therefore explain the maintenance of a health problem as well as attempts to reduce it.

The theory is particularly useful for explaining why behaviour may cluster within social groups, even when group members haven’t received formal instruction. Observation alone, however, doesn’t provide a complete explanation. Biological vulnerability and individual thought processes may influence the health problem, as may material circumstances and direct experience. Evidence that two people associate with each other doesn’t prove by itself that one learned the behaviour from the other. People may choose companions whose behaviour already resembles their own.

Understanding stress

Stress is a psychological and physiological process that occurs when someone sees demands as taxing or beyond their available coping resources. A stressor is an event, condition or demand that can trigger this process. The stressor is the possible source; stress is the response to it.

Part of the stress response depends on appraisal: a cognitive evaluation of what a situation means and whether the person has enough resources to manage it. As a result, two people may face a similar demand yet experience different levels of stress. They may judge its threat, predictability or controllability differently.

Short-lived stress can mobilize energy and attention, so it isn’t automatically harmful. Risk becomes more likely when stress is severe, repeated or prolonged, particularly if there are few opportunities to recover. Duration, intensity and perceived control all matter.

Direct pathways to a health problem

Stress can affect a health problem through direct physiological pathways. Activation prepares the body to meet a demand, changing functions such as arousal, cardiovascular activity and energy availability. When this activation is repeatedly prolonged, the accumulated strain may disrupt regulation and interfere with recovery, increasing vulnerability to a health problem.

Immune and inflammatory activity may also change under stress. The exact pathway differs between health problems. A strong explanation must therefore link the stress response to the particular problem rather than simply claim that stress is harmful.

Indirect pathways

Stress may affect health indirectly through behaviour. Someone under stress might sleep less effectively, withdraw from supportive relationships, neglect treatment or turn to behaviours that bring immediate relief but increase longer-term risk. Here, stress changes behaviour, which then contributes to the health problem.

An existing problem can also become worse because of stress. Its symptoms may bring further demands, uncertainty or disruption, creating additional stress. The relationship may therefore be bidirectional rather than a simple one-way chain.

Explaining variation and causality

People don’t all respond to stress in the same way. Its impact can vary with appraisal, coping resources, social support, previous experience, current health and the duration of exposure. These factors may explain why one person develops a health problem, or experiences a worsening one, while someone exposed to a similar stressor does not.

An association between reported stress and a health problem isn’t enough on its own to show that stress caused the problem. The health problem might increase stress, or another condition could contribute to both. Memory and current mood can also affect self-report measures. A convincing account specifies a direct or indirect mechanism, considers the direction of influence and avoids treating stress as either necessary or sufficient on its own.