Original Contribution - February 10, 1999
Sexual Dysfunction in the United States
Prevalence and PredictorsEdward O. Laumann, PhD
Anthony Paik, MA
Raymond C. Rosen, PhD
[Fulltext, Feb 10 JAMA.1999;281:537-544]
Context While recent pharmacological advances have generated increased public interest and demand for clinical services regarding erectile dysfunction, epidemiologic data on sexual dysfunction are relatively scant for both women and men.
Objective To assess the prevalence and risk of experiencing sexual dysfunction across various social groups and examine the determinants and health consequences of these disorders.
Design Analysis of data from the National Health and Social Life Survey, a probability sample study of sexual behavior in a demographically representative, 1992 cohort of US adults.
Participants A national probability sample of 1749 women and 1410 men aged 18 to 59 years at the time of the survey.
Main Outcome Measures Risk of experiencing sexual dysfunction as well as negative concomitant outcomes.
Results Sexual dysfunction is more prevalent for women (43%) than men (31%) and is associated with various demographic characteristics, including age and educational attainment. Women of different racial groups demonstrate different patterns of sexual dysfunction. Differences among men are not as marked but generally consistent with women. Experience of sexual dysfunction is more likely among women and men with poor physical and emotional health. Moreover, sexual dysfunction is highly associated with negative experiences in sexual relationships and overall well-being.
Conclusions The results indicate that sexual dysfunction is an important public health concern, and emotional problems likely contribute to the experience of these problems.
Sexual dysfunctions are characterized by disturbances in sexual desire and in the psychophysiological changes associated with the sexual response cycle in men and women. Despite increasing demand for clinical services and the potential impact of these disorders on interpersonal relationships and quality of life,[2,3] epidemiologic data are relatively scant. Based on the few available community studies, it appears that sexual dysfunctions are highly prevalent in both sexes, ranging from 10% to 52% of men and 25% to 63% of women.[4-6] Data from the Massachusetts Male Aging Study (MMAS) showed that 34.8% of men aged 40 to 70 years had moderate to complete erectile dysfunction, which was strongly related to age, health status, and emotional function. Erectile dysfunction has been described as an important public health problem by a National Institutes of Health Consensus Panel, which identified an urgent need for population-based data concerning the prevalence, determinants, and consequences of this disorder. Even less is known about the epidemiology of female sexual dysfunction.
Professional and public interest in sexual dysfunction has recently been sparked by developments in several areas. First, major advances have occurred in our understanding of the neurovascular mechanisms of sexual response in men and women.[9-11] Several new classes of drugs have been identified that offer significant therapeutic potential for the treatment of male erectile disorder,[12-14] while other agents have been proposed for sexual desire and orgasm disorders.[15,16] Availability of these drugs could increase dramatically the number of patients seeking professional help for these problems. Epidemiologic data would be of obvious value in developing appropriate service delivery and resource allocation models. Additionally, changing cultural attitudes and demographic shifts in the population have highlighted the pervasiveness of sexual concerns in all ethnic and age groups.
The present study addresses these issues by analyzing data on sexual dysfunction from the National Health and Social Life Survey (NHSLS), a study of adult sexual behavior in the United States. Sampling, data collection, and response analysis were all conducted under highly controlled conditions. This unique data source provides extensive information on key aspects of sexual behavior, including sexual problems and dysfunction, health and lifestyle variables, and sociocultural predictors. Prior analyses of sexual dysfunction, using NHSLS data, are limited, presenting basic prevalence rates across demographic characteristics and indicators of overall health and well-being.[17(pp368-374)] The present study, in contrast, uses multivariate techniques to estimate relative risk (RR) of sexual dysfunction for each demographic characteristic as well as for key risk factors.
The NHSLS, conducted in 1992, is a national probability sample of 1410 men and 1749 women between the ages of 18 and 59 years living in households throughout the United States. It accounts for about 97% of the population in this age rangeroughly 150 million Americans. It excludes people living in group quarters such as barracks, college dormitories, and prisons, as well as those who do not know English well enough to be interviewed. The sample completion rate was greater than 79%. Checks with other high-quality samples (eg, US Census Bureau's Current Population Survey) indicated that the NHSLS succeeded in producing a truly representative sample of the population. Each respondent was surveyed in person by experienced interviewers, who matched respondents on various social attributes, for an interview averaging 90 minutes. Extensive discussion of the sampling design and evaluations of sample and data quality are found in the book by Laumann et al.[17(pp35-73,549-605)]
Sexual dysfunction was indexed in this study according to 7 dichotomous response items, each measuring presence of a critical symptom or problem during the past 12 months.[17(p660)] Response items included: (1) lacking desire for sex; (2) arousal difficulties (ie, erection problems in men, lubrication difficulties in women); (3) inability achieving climax or ejaculation; (4) anxiety about sexual performance; (5) climaxing or ejaculating too rapidly; (6) physical pain during intercourse; and (7) not finding sex pleasurable. The last 3 items were asked only of respondents who were sexually active during the prior 12-month period. Taken together, these items cover the major problem areas addressed in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition classification of sexual dysfunction. Self-reports about sexual dysfunctions, especially in face-to-face interviews, are subject to underreporting biases arising from personal concerns about social stigmatization. Moreover, there may be systematic biases in underreporting related to particular attributes of the respondents. For example, older or less educated women or younger Hispanic men might be more reluctant to report sexual problems. Lack of privacy during interviews could also result in underreporting. However, analyses (not reported herein) indicate that reporting biases due to lack of privacy are negligible in NHSLS data.[17(pp564-570)]
A latent class analysis (LCA) was used to evaluate the syndromal clustering of individual sexual symptoms. Latent class analysis is a statistical method well suited for grouping categorical data into latent classes[18,19] and has a number of medical applications, such as evaluation of diagnostic systems[20-23] and generation of epidemiologic estimates using symptom data.[24,25] Latent class analysis tests whether a latent variable, specified as a set of mutually exclusive classes, accounts for observed covariation among manifest, categorical variables. A more detailed discussion of this method is available on request from the authors. Since diagnostic criteria for disorders of sexual dysfunction involve a complex of symptoms, we used LCA for grouping symptoms into categories. These categories, then, represent a typology of disorders for sexual dysfunction found in the US population, indicating both prevalence and types of symptoms.
We analyzed only those respondents reporting at least 1 partner in the prior 12-month period. Respondents who were sexually inactive during this period were excluded. This procedure may limit our results because excluded respondents may have avoided sex because of sexual problems. However, this procedure was necessary to ensure that each respondent answered all the symptom items since 3 items were asked only of sexually active respondents. A total of 139 men and 238 women were excluded on this basis. Excluded men were more likely to be single and have lower levels of education. We expect that this will bias our estimates of prevalence of sexual dysfunction downward since sexually inactive men generally reported higher rates of symptoms. Excluded women tended to be older and single. The exclusion of these women is likely to bias our estimates of the prevalence of sexual dysfunction upward given that these women tended to report lower rates.
Analyses performed in this study were made by use of logistic and multinomial logistic regression. For assessing the prevalence of symptoms across demographic characteristics, we performed logistic regressions for each symptom. This approach produced adjusted odds ratios (ORs), which indicate the odds that members of a given social group (eg, never married) reported the symptom relative to a reference group (eg, currently married), while controlling for other demographic characteristics. Demographic characteristics included respondent's age, marital status, educational attainment level, and race and ethnicity. Next, while controlling for these characteristics, we estimated adjusted ORs using multinomial logistic regressions for 3 sets of risk factors, each modeled separately in a nonnested manner. Risk factors associated with health and lifestyle included alcohol consumption, prior contraction of sexually transmitted diseases (STDs), presence of urinary tract symptoms, circumcision, health status, and experience of emotional or stress-related problems. Social status variables included change in income level and normative orientation, indexed by how liberal or conservative respondents' attitudes were toward sex. Risk factors associated with sexual experience included the number of lifetime sex partners, frequency of sex, how often respondents think about sex, frequency of masturbation, same sex contact, and experience in potentially traumatic events such as adult-child contact, forced sexual contact, sexual harassment, and abortion. Finally, we conducted a set of logistic regressions that used the categories of sexual dysfunction as predictor variables. These models measured the association between experience of dysfunction categories and quality-of-life concomitants, which included being satisfied personally and in relationships. We stress that concomitant outcomes cannot be causally linked as an outcome of sexual dysfunction. Latent class analyses were performed using maximum likelihood latent structure analysis. All logistic regressions used STATA version 5.0. Information regarding variable construction, LCA methods, and data quality are available from the authors.
Prevalence of Sexual Problems
Use of NHSLS data allows for calculating national prevalence estimates of sexual problems for adult women and men. While NHSLS data on critical symptoms do not connote a clinical definition of sexual dysfunction, their prevalence does provide important information about their extent and differential distribution among the US population. Table 1 and Table 2 analyze the prevalencing sexual problems across selected demographic characteristics. For women, the prevalence of sexual problems tends to decrease with increasing age except for those who report trouble lubricating. Increasing age for men is positively associated with experience of erection problems and lacking desire for sex. The oldest cohort of men (ages 50-59 years) is more than 3 times as likely to experience erection problems (95% confidence interval [CI], 1.8-7.0) and to report low sexual desire (95% CI, 1.6-5.4) in comparison to men aged 18 to 29 years. The prevalence of sexual problems also varies significantly across marital status. Premarital and postmarital (divorced, widowed, or separated) statuses are associated with elevated risk of experiencing sexual problems. Nonmarried women are roughly 1½ times more likely to have climax problems (95% CI, 1.0-2.1 and 1.2-2.3, respectively) and sexual anxiety (95% CI, 1.0-2.4 and 1.1-2.4, respectively) than married women. Similarly, nonmarried men report significantly higher rates for most symptoms of sexual dysfunction than married men. Thus, married women and men are clearly at lower risk of experiencing sexual symptoms than their nonmarried counterparts.
High educational attainment is negatively associated with experience of sexual problems for both sexes. These differences are especially marked between women who do not have high school diplomas and those who have college degrees. Controlling for other demographic characteristics, women who have graduated from college are roughly half as likely to experience low sexual desire (95% CI, 0.3-0.8), problems achieving orgasm (95% CI, 0.3-0.7), sexual pain (95% CI, 0.3-1.0), and sexual anxiety (95% CI, 0.3-1.0) as women who have not graduated from high school. Male college graduates are only two thirds (95% CI, 0.4-1.0) as likely to report climaxing too early and half as likely to report nonpleasurable sex (95% CI, 0.2-0.9) and sexual anxiety (95% CI, 0.3-0.8) than men who do not have high school diplomas. Overall, women and men with lower educational attainment report less pleasurable sexual experience and raised levels of sexual anxiety.
The association between race and ethnicity and sexual problems is more variable. Black women tend to have higher rates of low sexual desire and experience less pleasure compared with white women, who are more likely to have sexual pain than black women. Hispanic women, in contrast, consistently report lower rates of sexual problems. Differences between men are not as marked but are generally consistent with what women experience. Indeed, although the effects of race and ethnicity are fairly modest among both sexes, blacks appear more likely to have sexual problems while Hispanics are less likely to have sexual problems, across the categories of sexual dysfunction.
Latent Class Analysis
The results of LCA allow for analyzing risk factors and quality-of-life concomitants in relation to categories of sexual dysfunction, rather than individual symptoms. Analyses presented in Table 3, Table 4, and Table 5 use the results of LCA instead of individual symptoms. These results indicate that the clustering of symptoms according to syndrome can be represented by 4 categories for women as well as for men. Latent class analysis also estimates the size of each class as a proportion of the total sample, a result corresponding to prevalence of categories of sexual dysfunction in the US population. Finally, LCA identifies each class's symptoms, indicating the likelihood that respondents in that class will exhibit a given symptom, thus providing researchers with information about what elements characterize each category. Although not equivalent to clinical diagnosis, this approach offers a statistical representation of sexual dysfunction.
For women, the 4 categories identified by LCA roughly correspond to major disorders of sexual dysfunction as outlined by the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition. These include an unaffected group (58% prevalence), a low sexual desire category (22% prevalence), a category for arousal problems (14% prevalence), and a group with sexual pain (7% prevalence). Similarly, a large proportion of men (70% prevalence) constitutes an unaffected population. The remaining categories consist of premature ejaculation (21% prevalence), erectile dysfunction (5% prevalence), and low sexual desire (5% prevalence). Overall, the results of LCA show that the total prevalence of sexual dysfunction is higher for women than men (43% vs 31%).
Tables 3 and 4 present multinomial logistic regressions on categories of sexual dysfunction. Adjusted ORs indicate the relative risk of experiencing a given category of sexual dysfunction vs reporting no problems for each risk factor, while controlling for other characteristics. With regard to health and lifestyle risk factors, those who experience emotional or stress-related problems are more likely to experience sexual dysfunctions defined in each of the categories. In contrast, health problems affect women and men differently. Men with poor health have elevated risk for all categories of sexual dysfunction, whereas this factor is only associated with sexual pain for women. The presence of urinary tract symptoms appears to impact sexual function only (eg, arousal and pain disorders for women or erectile dysfunction for men). Finally, having had an STD, moderate to high alcohol consumption, and circumcision generally do not result in increased odds of experiencing sexual dysfunction.
Social status variables, which measure an individual's socioeconomic and normative position relative to other persons, assess how sociocultural position affects sexual function. Deterioration in economic position, indexed by falling household income, is generally associated with a modest increase in risk for all categories of sexual dysfunction for women but only erectile dysfunction for men. Normative orientation does not appear to have any impact on sexual dysfunction for women; men with liberal attitudes about sex, in contrast, are approximately 1¾ times more likely to experience premature ejaculation (95% CI, 1.2-2.5).
Finally, various aspects of sexual experience result in an increased risk of sexual dysfunction. Sexual history, indicated by having more than 5 lifetime partners and by masturbation practices, does not increase relative risk for either women or men. Women with low sexual activity or interests, however, have elevated risk for low sexual desire and arousal disorders. Men do not exhibit similar associations. The impact of potentially traumatic sexual events is markedly different for women and men. Women respondents reporting any same-sex activity are not at higher risk for sexual dysfunction, while men are. Men reporting any same-sex activity are more than twice as likely to experience premature ejaculation (95% CI, 1.2-3.9) and low sexual desire (95% CI, 1.1-5.7) than men who have not. Arousal disorder appears to be highly associated in women who have experienced sexual victimization through adult-child contact or forced sexual contact. Similarly, male victims of adult-child contact are 3 times as likely to experience erectile dysfunction (95% CI, 1.5-6.6) and approximately 2 times as likely to experience premature ejaculation (95% CI, 1.2-2.9) and low sexual desire (95% CI, 1.1-4.6) than those who have not been victims of adult-child contact. Finally, men who have sexually assaulted women are 3½ times as likely to report erectile dysfunction (95% CI, 1.0-12.0). Indeed, traumatic sexual acts continue to exert profound effects on sexual functioning, some effects lasting many years beyond the occurrence of the original event.
The experience of sexual dysfunction is highly associated with a number of unsatisfying personal experiences and relationships. Table 5 highlights the associations of categories of sexual dysfunction with emotional and physical satisfaction with sexual partners and with feelings of general happiness. However, no causal order should be assumed since quality-of-life indicators are concomitant outcomes of sexual dysfunction. For women, all categories of sexual dysfunctionlow sexual desire, arousal disorder, sexual painhave strong positive associations with low feelings of physical and emotional satisfaction and low feelings of happiness. Similar to women, men with erectile dysfunction and low sexual desire experience diminished quality of life, but those with premature ejaculation are not affected. In brief, experience of sexual dysfunction is generally associated with poor quality of life; however, these negative outcomes appear to be more extensive and possibly more severe for women than men. In an examination of help-seeking behavior (analysis not shown here, but available on request), we found that roughly 10% and 20% of these afflicted men and women, respectively, sought medical consultation for their sexual problems.
Demographic factors such as age are strongly predictive of sexual difficulties, particularly erectile dysfunction. Sexual problems are most common among young women and older men. Several factors may explain these differential rates. Since young women are more likely to be single, their sexual activities involve higher rates of partner turnover as well as periodic spells of sexual inactivity. This instability, coupled with inexperience, generates stressful sexual encounters, providing the basis for sexual pain and anxiety. Young men are not similarly affected. Older men are more likely to have trouble maintaining or achieving an erection as well as to lack an interest in sex. Low sexual interest and erection problems are age-dependent disorders, possibly resulting from physiological changes associated with the aging process. Indeed, our results are consistent with those generated by the MMAS, which determined that 9.6% of its sample experienced complete impotence and showed a strong age association increasing from 5% to 15% between subject ages of 40 and 70 years.
Other factors such as poor health and lifestyle are differentially predictive across demographic groups. While nonmarital status is associated with lower overall well-being, part of the higher risk of sexual dysfunction probably stems from differences in sexual lifestyles. Similarly, elevated risks associated with low educational attainment and minority status attest to the fact that better-educated individuals are healthier and have lifestyles that are physically and emotionally less stressful. To understand the factors that predispose individuals to sexual dysfunction, we should analyze risk factors.
The NHSLS data indicate that emotional and stress-related problems among women and men generate elevated risk of experiencing sexual difficulties in all phases of the sexual response cycle. While we caution that the causal order of this relationship is uncertain, these results suggest that psychosocial disturbances affect sexual functioning. This does not imply that the impact of poor health is negligible; in fact, the opposite is demonstrated since age, health problems, and urinary tract infections result in elevated risk of experiencing sexual dysfunction. Rather, both physiological and psychological statuses are independent factors that affect sexual functioning.
Given the salience of emotional distress on sexual dysfunction, we examine 2 underlying sources of psychosocial stress: social status and sexual trauma. The NHSLS data clearly suggest that deteriorating social position negatively affects sexual functioning. Deterioration in economic position induces higher levels of stress, which in turn affects sexual functioning, a result more pervasive among women than men. Future research should be directed toward mapping the social distribution of emotional distress.
With respect to possibly traumatic sexual experiences, our findings are complex and show distinct differences between sexes but clearly provide evidence that these experiences are sources of psychosocial stress. First, we found that the impact of same-sex activity is relevant for men but not women. The source of this difference may be rooted in the subjective meaning of these sex acts, because many male-to-male encounters have involved adult-child contact. We should note that these results assess the impact of historical instances of same-sex activity, not the relationship between homosexuality and sexual problems. Similarly, indicators for sexual victimization show strong effects in persons of both sexes. For women, adult-child contact or forced sex, both generally perpetrated by men, results in increased risk of experiencing arousal disorders. These results support the view that sexual traumas induce lasting psychosocial disturbances, which ultimately affect sexual functioning. Similarly, men who were touched sexually before puberty also are more likely to experience all categories of sexual dysfunction. In short, both female and male victims of unwanted sexual contact exhibit long-term changes in sexual functioning.
While the causal relationship between quality-of-life concomitants and sexual dysfunction also remains to be investigated, the strong associations observed in NHSLS data suggest that sexual dysfunction is a largely uninvestigated yet significant public health problem. Recent advances in therapy for erectile dysfunction may increase quality of life for some men. However, since low well-being is strongly associated with female sexual problems, researchers should focus on identifying the consequences of these problems as well as developing appropriate therapies. With the affected population rarely receiving medical therapy for sexual dysfunction, service delivery efforts should be augmented to target high-risk populations.
This report provides the first population-based assessment of sexual dysfunction in the half-century since Kinsey et al.[29,30] The results from the NHSLS indicate that sexual problems are widespread in society and are influenced by both health-related and psychosocial factors. The role of the latter implies that stress-inducing events, due to either individual or social sources, can affect sexual functioning in both men and women. Differing patterns of sexual dysfunction were observed across sex, age, and demographic groups, highlighting the need for further research on etiologic mechanisms. With the strong association between sexual dysfunction and impaired quality of life, this problem warrants recognition as a significant public health concern.
Author Affiliations: Department of Sociology, University of Chicago, Chicago, Ill (Dr Laumann and Mr Paik); and Department of Psychiatry, University of Medicine and Denistry of New JerseyRobert Wood Johnson Medical School, Piscataway (Dr Rosen).
Funding/Support: This study was supported in part by grant NICHD RO1-HD28356 from the National Institutes of Health, Bethesda, Md, and by grant 940-1417 from the Ford Foundation, New York, NY.
Acknowledgment: We also would like to express our appreciation for the helpful comments of George Rhodes, MD.
Corresponding Author and Reprints: Edward O. Laumann, PhD, Department of Sociology, University of Chicago, 5848 S University Ave, Chicago, IL 60637 (e-mail: firstname.lastname@example.org).
1. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition. Washington, DC: American Psychiatric Association; 1994:493-522.
2. Morokoff PJ, Gillilland R. Stress, sexual functioning, and marital satisfaction. J Sex Res. 1993;30:43-53.
3. Fugl-Meyer AR, Lodnert G, Branholm IB, Fugl-Meyer KS. On life satisfaction in male erectile dysfunction. Int J Impot Res. 1997;9:141-148.
4. Frank E, Anderson C, Rubinstein D. Frequency of sexual dysfunction in "normal" couples. N Engl J Med. 1978;299:111-115.
5. Rosen RC, Taylor JF, Leiblum SR, Bachmann GA. Prevalence of sexual dysfunction in women: results of a survey study of 329 women in an outpatient gynecological clinic. J Sex Marital Ther. 1993;19:171-188.
6. Spector IP, Carey MP. Incidence and prevalence of the sexual dysfunctions: a critical review of the empirical literature. Arch Sex Behav. 1990;19:389-408.
7. Feldman HA, Goldstein I, Hatzichristou DG, Krane RJ, McKinlay JB. Impotence and its medical and psychosocial correlates: results of the Massachusetts Male Aging Study. J Urol. 1994;151:54-61.
8. NIH Consensus Development Panel on Impotence. Impotence. JAMA. 1993;270:83-90.
9. Rajfer J, Aronson WJ, Bush PA, Dorey FJ, Ignarro LJ. Nitric oxide as a mediator of relaxation of the corpus cavernosum in response to nonadrenergic, noncholinergic neurotransmission. N Engl J Med. 1992;326:90-94.
10. Burnett AL. The role of nitric oxide in the physiology of erection. Biol Reprod. 1995;52:485-489.
11. Park K, Goldstein I, Andry C, Siroky MB, Krane RJ, Azadzoi KM. Vasculogenic female sexual dysfunction: the hemodynamic basis for vaginal engorgement insufficiency and clitoral erectile insufficiency. Int J Impot Res. 1997;9:27-37.
12. Boolell M, Gepi-Attee S, Gingell JC, Allen MJ. Sildenafil: a novel effective oral therapy for male erectile dysfunction. Br J Urol. 1996;78:257-261.
13. Heaton JP, Morales A, Adams MA, Johnston B, el-Rashidy R. Recovery of erectile function by the oral administration of apomorphine. Urology. 1995;45:200-206.
14. Morales A, Heaton JP, Johnston B, Adams M. Oral and topical treatment of erectile dysfunction: present and future. Urol Clin North Am. 1995;22:879-886.
15. Rosen RC, Ashton AK. Prosexual drugs: empirical status of the "new aphrodisiacs." Arch Sex Behav. 1993;22:521-543.
16. Segraves RT, Saran A, Segraves K, Maguire E. Clomipramine versus placebo in the treatment of premature ejaculation: a pilot study. J Sex Marital Ther. 1993;19:198-200.
17. Laumann EO, Gagnon JH, Michael RT, Michaels S. The Social Organization of Sexuality: Sexual Practices in the United States. Chicago, Ill: University of Chicago Press; 1994.
18. Clogg CC. Latent class models. In: Arminger G, Clogg CC, Sobel ME, eds. Handbook of Statistical Modeling for the Social and Behavioral Sciences. New York, NY: Plenum Press; 1995:311-359.
19. McCutcheon AL. Latent Class Analysis. Newbury Park, Calif: Sage Publications; 1987.
20. Uebersax JS, Grove WM. Latent class analysis of diagnostic agreement. Stat Med. 1990;9:559-572.
21. Young MA. Evaluating diagnostic criteria: a latent class paradigm. J Psychiatr Res. 1983;17:285-296.
22. Young MA, Tanner MA, Meltzer HY. Operational definitions of schizophrenia: what do they identify? J Nerv Ment Dis. 1982;170:443-447.
23. Rindskopf D, Rindskopf W. The value of latent class analysis in medical diagnosis. Stat Med. 1986;5:21-27.
24. Eaton WW, McCutcheon AL, Dryman A, Sorenson A. Latent class analysis of anxiety and depression. Sociol Methods Res. 1989;18:104-125.
25. Kohlman T, Formann AK. Using latent class models to analyze response patterns in epidemiologic mail surveys. In: Rost J, Langeheine R, eds. Applications of Latent Trait and Latent Class Models in the Social Sciences. New York, NY: Waxmann Munster; 1997:345-351.
26. Clogg CC. Unrestricted and Restricted Maximum Likelihood Latent Structure Analysis: A Manual for Users. University Park, Pa: Population Issues Research Center, Pennsylvania State University; 1977. MLLSA Working Paper 1977-09.
27. STATA Release 5. College Station, Tex: Stata Press; 1997.
28. Browning C, Laumann EO. Sexual contact between children and adults: a lifecourse perspective. Am Sociol Rev. 1997;62:540-560.
29. Kinsey AC, Pomeroy WB, Martin CE. Sexual Behavior in the Human Female. Philadelphia, Pa: WB Saunders Co; 1948.
30. Kinsey AC, Pomeroy WB, Martin CE, et al. Sexual Behavior in the Human Female. Philadelphia, Pa: WB Saunders Co; 1953.
© 1995-1999 American Medical Association. All rights reserved.