Medical Research & Innovations

For decades, doctors assumed men’s libido follows testosterone, peaking in youth and declining steadily. A study of 67,000 people just showed that assumption is wrong

For decades, doctors assumed men’s libido follows testosterone, peaking in youth and declining steadily. A study of 67,000 people just showed that assumption is wrong

Testosterone is the standard story of male sexual desire. The hormone rises sharply through adolescence, peaks in the late teens and early 20s, and then begins a slow but steady decline across adulthood. Medical textbooks, clinical guidelines, and popular health advice have all treated desire as a function of this hormonal curve. If a man reports lower libido in his 40s or 50s, the assumption is almost automatic: testosterone is falling, and desire follows.

A large population study just published in Scientific Reports has tested that assumption directly, using one of the biggest samples ever assembled for this kind of research. The findings do not support the testosterone narrative.

Researchers at the University of Tartu in Estonia analyzed data from 67,334 adults drawn from the Estonian Biobank, a national biological and lifestyle repository representing roughly 7% of Estonia’s adult population. Participants ranged in age from 18 to 89. Each was asked to rate their general level of sexual desire using a validated two-item scale measuring sexual thoughts and urges. The researchers then mapped those ratings against age, gender, sexual orientation, relationship status, relationship satisfaction, parenthood, education, and occupational background.

For men, the pattern that emerged bore little resemblance to the testosterone curve.

Male desire did not peak in early adulthood. It rose steadily through the 20s and 30s, reached its highest point in the late 30s and early 40s, and then declined gradually rather than sharply. And at the far end of the age range, the results produced perhaps the most counterintuitive finding of all: men in their 60s reported levels of sexual desire comparable to men in their 20s.

“One thing that surprised us was that men’s sexual desire appeared to peak around their late 30s or early 40s rather than simply decline steadily with age,” said lead researcher Toivo Aavik of the University of Tartu. “That pattern does not map neatly onto testosterone trajectories, so it suggests that relational or social factors may also play a substantial role in sustaining desire.”

Why the testosterone explanation breaks down

The mismatch between the hormonal model and the actual data points toward a problem that researchers in sexual medicine have been circling for years. Testosterone does influence sexual desire, but the relationship is not as clean or direct as the standard model implies.

Studies measuring testosterone levels and self-reported desire simultaneously have found that the correlation is weaker than expected, particularly in the normal range. Men with clinically low testosterone reliably report reduced desire, and testosterone replacement tends to restore it. But among men whose testosterone is within the normal range, variation in hormone levels predicts relatively little about who reports high or low desire on any given day or across years.

What the Estonian data adds is a population-level picture showing that desire and testosterone diverge not just across individuals but across the lifespan. If testosterone were the primary driver, the age curve for desire should roughly mirror the hormone curve: a peak in youth, a gradual slope downward. The observed curve, rising through midlife before declining, is structurally different.

Aavik and colleagues suggest the divergence points toward psychosocial factors, including relationship stability, emotional intimacy, reduced performance anxiety with age and experience, and the influence of long-term partnership on desire. Stable relationships have been associated in prior research with both increased sexual activity and sustained emotional investment. Men in their 30s and early 40s are more likely to be in established partnerships than men in their 20s, and this relational context may support desire in ways that partially compensate for any hormonal decline.

The gender difference, and what it reveals

The study also mapped desire trajectories for women, and the patterns diverged substantially from men in ways that add further texture to the findings.

Women’s desire peaked earlier, during the 20s and early 30s, and declined more steeply, with a pronounced drop after age 50 that likely reflects the hormonal shifts of menopause. The gender gap in desire was large across most of the adult lifespan, with men reporting substantially higher desire than women on average, a difference larger than what prior smaller studies had documented.

But the researchers were careful to note that reported desire in women may be influenced by social and cultural factors beyond biology. In many contexts, men’s sexual desire is more socially acceptable to express and acknowledge, while women may underreport desire due to internalized norms, uncertainty about what counts as desire, or concerns about stigma.

The study also found that sexual orientation shaped desire patterns in notable ways. Bisexual and pansexual individuals reported stronger sexual desire than heterosexual individuals across the sample, while asexual individuals reported lower desire, as expected. These differences held after controlling for age, gender, and relationship factors, suggesting orientation has an independent relationship with desire intensity that warrants further investigation.

Parenthood, relationships, and the factors that shape desire

One of the more intriguing subsidiary findings concerned the relationship between parenthood and desire. Participants who had welcomed a child within the past year reported slightly higher sexual desire than those who had not, though the researchers cautioned that this likely reflects age rather than a causal effect of new parenthood, since people having children are typically younger and therefore closer to their desire peak.

Over the longer term, having more children was associated with lower desire in women but not in men. Relationship satisfaction showed the expected positive correlation with desire across both genders: people who reported greater satisfaction with their partnerships also reported stronger desire.

Education and occupational background showed smaller associations, but the overall picture from the data was one in which relational and social factors account for meaningful variation in desire alongside age and gender. Demographic variables combined accounted for about 28% of the variation in desire scores, suggesting that the majority of individual differences in sexual desire remain unexplained by the factors the study measured.

What this changes for how desire is understood

The practical implication for men is direct. The assumption that desire peaks in youth and declines from there is not supported by data from 67,000 people across seven decades of adult life. A man in his late 30s or early 40s reporting strong sexual desire is not anomalous. He may be at his statistical peak. And a man in his 60s reporting desire comparable to his younger self is not contradicting the biology but rather behaving entirely consistently with what the population data actually shows.

For clinicians, the findings suggest that conversations about desire and its changes with age may need updating. The testosterone framework, while clinically relevant at the extremes, does not capture the midlife arc that this data reveals. Relational factors, life circumstances, and the complex interplay of social and psychological variables appear to shape desire across the lifespan in ways that a hormone-only model misses.

“The mid-life peak in men suggests that factors beyond biological ageing, such as relational dynamics, may play a more significant role than initially anticipated,” the researchers wrote in the paper.

The study “Age, gender, sexual orientation, relationship, and parental status as predictors of sexual desire: a population-based study” was authored by Toivo Aavik and colleagues at the University of Tartu, Estonia, and published in Scientific Reports in January 2026.

Source: University of Tartu. DOI: 10.1038/s41598-025-99408-4