Longitudinal Studies on Aging

Longitudinal studies on aging involve repeated observations of the same variables over long periods of time, usually at regular intervals (two to three years), and they study trends, life events, and general life experiences across the life span. Although they do not have the strength of experimental designs in [Page 805]terms of building a case for causality (they observe; they do not manipulate situations to study the effect), they are stronger than cross-sectional studies as, in following the same people over time, they may track incidence as well as prevalence of events; rely on current reported rather than recalled information; offer more opportunity to control for, or at least identify where there may be, unobserved individual differences; and consider a time dimension to observations of how differences in people's resources (financial and personal), lives, and experiences translate into different outcomes in older age. A particular focus of Many longitudinal studies on aging is identifying the predictors of both onset of disease and types of disease progression. There is also an opportunity to understand if phenomena (e.g., impact of life events) are of short-or long-term duration.

The type of information gathered in longitudinal studies, therefore, is of interest to human services, often helping identify the most salient modifiable factors in disease and disadvantage that human services may then target, including those that are most relevant to diverse populations.

Longitudinal studies are usually either cohort studies or panel studies. Cohort studies sample a cohort, defined as a group experiencing some event (e.g., persons who have experienced a heart attack or a particular age group) within a particular time band and then gather related data (e.g., variables potentially associated with onset of additional heart attacks or success in return to normal functioning) at intervals over time. Panel studies sample a cross-section of a population (a random sample, a particular geographic or sociocultural group, or a professional group such as nurses) and then have members of the panel (who present a range of circumstances and experiences) complete questionnaires and other measures at regular intervals. The target population for a longitudinal study may be individuals or may be households. There are also retrospective longitudinal studies, for example, studies using medical records that look back in time trying to understand trends or to identify common markers in people's histories that may explain outcomes, particularly medical outcomes, in their later lives.

Examples of such studies include the oldest longitudinal study on aging (http://www.blsa.nih.gov), focused on understanding changes due to normal aging and those due to disease or other causes, and the Health and Retirement study (HRS, hrsonline.isr.umich.edu). In HRS, data on physical and mental health and health care utilization, on living situations, and on financial and social support resources are available and are being collected over regular intervals. Other studies target more specific health concerns; for example, the Framingham Heart study (http://www.framinghamheartstudy.org) is focused primarily on cardiovascular disease and the Nun study on Alzheimer's disease (http://www.mc.uky.edu/nunnet); or particular populations, for example, the Nurses’ Health study looking at women's health over time (http://www.channing.harvard.edu/nhs/History/index.shtml). Several studies have included the collection of biomarker material and informed the development of subsequent longitudinal studies in the United States and other countries. In particular, there has been a more prominent featuring of anthropometric measurement (e.g., http://www.ifs.org.uk/elsa), collection of genetic materials (e.g., http://www.tilda.ie), and the inclusion of unique populations such as in P. McCallion's Intellectual Disability Supplement to the Irish Longitudinal study of Aging.

...