How do surveys define "sexual activity" and how does that affect reported prevalence in people over 80?
Executive summary
Surveys of older adults use widely different definitions of “sexual activity”—ranging from strict intercourse-only measures to broad lists that include touching, kissing, masturbation and emotional intimacy—which systematically shifts reported prevalence, especially in people over 80. Studies that count any sexual contact or solo activities show much higher rates among the oldest groups than those that count only partnered intercourse or require activity within a narrow time window (for example, past 12 months), and methodological choices about sampling, question wording and partner availability further magnify these differences [1] [2] [3].
1. The definitional spectrum: what researchers mean by “sexual activity”
Some large, influential surveys define sexual activity expansively as “any mutually voluntary activity with another person that involves sexual contact, whether or not intercourse or orgasm occurs,” and typically then apply a time window such as “in the previous 12 months” to mark current activity [1]; other instruments list discrete behaviors—sexual intercourse, kissing, petting/fondling, and masturbation—and score respondents as sexually active if they report any of those behaviors over a specified period [2] [3]. Smaller qualitative studies and some participant-led accounts add emotional and physical intimacy (holding, caressing, non-sexual closeness) to what respondents consider “having sex,” particularly in older cohorts [4]. By contrast, some analyses reported in media or single-item polls use narrower metrics—frequency of intercourse or presence of a regular sexual partner—to mark sexual activity, which produces lower estimates for the oldest age brackets [5] [6].
2. Time windows, solo vs partnered acts, and their downstream effects on prevalence estimates
Using a 12‑month reference period (common in nationally representative work) will count fewer people as sexually active in very old samples than a lifetime or 5‑year window; NSF-style definitions that require “sex with at least one partner in the previous 12 months” yield sharp declines after age 75 in many datasets [1]. Including masturbation and non-intercourse activities raises prevalence among those 80+ because touching, caressing and solo sex are reported as comparatively common activities in the 80–102 age studies—often more frequent than intercourse in that demographic [7] [2]. Systematic reviews show that heterogeneity in what counts as “sex” explains a substantial share of the wide range in published prevalence figures for older adults [3].
3. Partner availability, gender ratios and the “opportunity” effect
Demographic realities—skewed sex ratios, high widowhood among women and lower marriage rates among very old women—meaningfully depress partnered-sex prevalence for people over 80 even if desire and solitary sexual practices remain [8] [9]. Studies that measure only partnered intercourse therefore undercount sexual expression among older women because they ignore solo behaviors and non‑penetrative intimacy, which are more commonly reported in the oldest age groups [7] [9].
4. Health, sampling and social desirability: who gets counted as “sexually active”?
Sexual activity in surveys tracks closely with self‑reported health and cohabitation status—people reporting good health and live‑in partners are much more likely to report sexual activity—so samples skewed toward healthier, wealthier or residential populations will show higher prevalence among the oldest cohorts [10] [11] [7]. Methodological choices also affect honesty: anonymous self‑administered questionnaires tend to elicit higher reporting of non‑normative or intimate behaviors than face‑to‑face interviews, and response rate and panel composition (web panels, clinic samples, retirement‑facility residents) introduce further bias [2] [1].
5. What the differences mean in practice: interpreting “31% of men 80+” versus “many remain sexual”
Headline percentages that report low intercourse rates in people over 80 (for example, single‑digit or low‑30s for intercourse) are accurate for that narrow behavior and timeframe, but they obscure substantial sexual expression captured by broader definitions—touching, kissing, masturbation, emotional intimacy—which appear with meaningful frequency even in nonagenarians in some studies [4] [7] [3]. Therefore, policymakers, clinicians and the public should read prevalence numbers alongside the survey’s operational definition of “sexual activity,” the timeframe used, and sample characteristics before concluding that sexuality wanes entirely after a given age [2] [12].