Incidence measures the rate of new cases arising in a population over a defined time, while prevalence measures the proportion of a population with a condition at a given point. Incidence rate requires a denominator of person-time at risk; prevalence is a snapshot reflecting both incidence and disease duration. The relationship between the two—prevalence ≈ incidence × mean duration—helps public health practitioners understand whether changes in a condition reflect new exposure, improved survival, or both. Crude rates are often standardized by age or sex to enable valid comparisons across populations.
Work through numerical examples contrasting point prevalence and cumulative incidence. Calculate age-standardized rates for two populations with different age structures to see why standardization matters before comparing disease burden across regions.
When epidemiologists study a disease in a population, they need two fundamentally different measurements: how often new cases arise, and how much disease exists at any given moment. Incidence answers the first question; prevalence answers the second. Understanding the distinction is not merely definitional — it shapes how you interpret trends and design public health responses.
Incidence measures the occurrence of new cases over time. The most precise form, incidence density (also called the incidence rate), uses person-time as the denominator: each participant contributes observation time to the denominator only while they are at risk. If 10 new cases arise in a cohort contributing 500 person-months of observation, the incidence rate is 10/500 = 0.02 per person-month. Cumulative incidence, by contrast, is the proportion of an at-risk population that develops the disease over a fixed interval — it answers "what is the probability of getting the disease in the next year?" Person-time and fixed-interval denominators answer different questions and cannot be used interchangeably, as the misconceptions section notes.
Prevalence is a snapshot. Point prevalence is the proportion of the population carrying the disease at a single moment in time. It has no time unit in its denominator — it is a proportion, not a rate — which is why calling it a "prevalence rate" is technically wrong, even though you will encounter this usage in practice. The formula that connects the two measures is the steady-state approximation: prevalence ≈ incidence rate × mean disease duration. At equilibrium, the pool of prevalent cases grows when incidence rises or duration lengthens, and shrinks when either falls.
That formula carries a critical interpretive implication. Prevalence can rise for two completely different reasons: more new cases (rising incidence) or longer survival with the disease (rising duration). When effective treatments for chronic conditions emerge, patients who would have died quickly now live for years — prevalence rises even if incidence is flat or declining. Misreading this as "more disease" leads to misguided policy. The first question to ask when prevalence trends change is always: has incidence changed, or has survival changed?
Finally, comparing prevalence across populations with different age structures requires standardization. Older populations have more prevalent chronic disease simply because of age, not because of anything specific to their environment. Age-standardization applies each population's age-specific rates to a common standard population, enabling fair comparisons by holding the confounding variable constant. This is the same logic underlying covariate adjustment in the statistical methods you have seen: isolate the relationship of interest by removing variation attributable to other factors.