As with all surveys, one has to consider what factors influenced the selection (and self-selection) of the responders. In general, participants in these studies tend to be better educated and more liberal in their attitudes than their contemporaries; and important groups, such as people with chronic illness, may be under (or over) represented, depending on the setting of the survey.

Attempts to assess the effects of ageing on sexuality by surveying different age-groups at a single point in time are vulnerable to cohort effects—the sexual experience of those brought up before the Second World War is different from that of people whose childhood was in the 1950s, and need have nothing to do with age. Longitudinal studies are demanding and costly, but they reveal the effects of ageing more clearly, and they also allow individual patterns, stable over time, to be identified. For example, the series of studies conducted at Duke University showed that although the prevalence of sexual interest and activity in both men and women decreases with age, individual patterns of sexuality tend to be stable until some event (such as illness or loss of a partner) disrupts the pattern.⁽³⁾

There is more to sex than actual intercourse, especially in old age. A consistent finding is that in both sexes interest in sex is more resistant to ageing than is sexual activity, even when non-coital activity is included. For example, in a cross-sectional study of volunteer respondents aged 80 or over, living in a residential home in the United States,⁽⁴⁾ 88 per cent of the men and 71 per cent of the women enjoyed daydreams and fantasies about sex, 82 per cent of the men and 61 per cent of the women engaged in touching and caressing, while 63 per cent of the men and 30 per cent of the women had sexual intercourse ‘at least sometimes’. Janus and Janus⁽⁵⁾ surveyed 2765 subjects in the United States by questionnaire, supplemented by 125 interviews. They found less reduction of activity in the older groups than had been found in earlier surveys, probably because they did not confine the question to coitus. In every age group men thought that they were more active than 3 years previously, but women, particularly those over 50, noted a decline. Unlike earlier studies foreplay was discussed, and the authors concluded that older men gain greater pleasure and experience more intimacy and warmth after coitus than younger men.

The commonest reasons given by people for a decline in their accustomed level of sexual activity were illness, and the loss of their partner.⁽³⁾ Interestingly, women (but not men) also gave illness of their partner as a reason. This may reflect the traditional male role in initiating sexual activity, and also perhaps the fact that husbands tend to be older (therefore more at risk of illness) than their wives. For example, in a Swedish community study of 85-year olds,⁽⁶⁾ participation in sexual intercourse was reported by 10 per cent of married women but only 1 per cent of the unmarried women, and by 22 per cent of married men compared to 13 per cent of the unmarried men. The rates for sexual interest (as opposed to activity) were higher: 46 and 37 per cent for married and unmarried men respectively; 24 and 15 per cent for married and unmarried women.

Attitudes towards sexuality in old age—as revealed in responses to systematically varied vignettes—have become generally more positive over the last half-century, among both younger and older people.^(2,7) Probably this is true also of professional attitudes, and where this matters most is among the staff caring for older people in institutional settings. The evidence suggests that care staff who are older, better educated and have had vocational training, and who have more experience of caring, are likely to be more open to the sexual needs of their residents. However as Bouman⁽²⁾ points out, the same cannot be said of health policies and strategy. Government guidance on the care of older people has ignored their sexuality, and in most official surveys and policies on sexual issues, people aged 60 or more are excluded from consideration.

With increasing life-expectancy many marriages continue well into the 9th decade of one or both partners. Divorce has now overtaken death as the main cause for the ending of a marriage, though this does not necessarily mean increasing numbers of people left single in old age. Many older people experience second or third marriages (or cohabitations), and the age gap between the partners in these new relationships tends to be larger than that between partners still in their first marriage. The complex effects on family structure of these different social trends are analysed by Harper.⁽⁸⁾ She shows that married older people have higher levels of health, social participation, and life satisfaction than those not married, and they live longer; while divorced men (compared to women, and to widowers) are the most disadvantaged in those respects. To this we can add (based on the survey data mentioned above) that in old age married people also enjoy greater opportunities than the unmarried for sexual expression.