Body image, body satisfaction, and unsafe anal intercourse among men who have sex with men.

Introduction

The popular press and scientific community have shown increased interest in body image in men over the past several years. Pope, Phillips, and Olivardia (2000) estimated that there are more than 50 million "muscle-dissatisfied" men in the United States. Other studies suggest that exposure to media images of lean, muscular men may be contributing to this trend, causing some men to feel depressed, dissatisfied, and unable to achieve a muscular ideal (Agliata & Tantleff-Dunn, 2004; Leit, Gray, & Pope, 2002; Vartanian, Giant, & Passino, 2001).

Men who feel dissatisfied with their body may be more prone to dieting, anxiety, depression, obsessive-compulsive behavior, smoking, alcohol/drug use, and unhealthy weight loss practices such as severely restricting food consumption, binging, and excessive exercise (Barry, Grilo, & Masheb, 2002; Bohne, Keuthen, Wilhelm, Deckersbach, & Jenike, 2002; Carlat & Camargo, 1991; Carlat, Camargo, & Herzog, 1997; Phillips & Diaz, 1997; Robb & Dadson, 2002). Gay men, in particular, may feel more dissatisfied with their bodies than heterosexual men, likely due to the strong emphasis placed on appearance in the gay community (Beren, Hayden, Wilfley, & Grilo, 1996; Herzog, Newman, & Warshaw, 1991; Siever, 1994; Silberstein, Mishkind, Striegel-Moore, Timko, & Rodin, 1989; Yager, Kurtzman, Landsverk, & Wiesmeier, 1988). Depression and low self-valuation may then lead some gay men to use sex encounters as a coping strategy, without considering safer sex practices (Martin & Knox, 1997a, 1997b). A recent study of mood and sexuality in gay men by Bancroft and colleagues (2003) reported that 24% of the gay men surveyed reported increased sexual interest when they experience anxiety, while 14% of the men undergoing in-depth qualitative interviews reported reduced concern about sexual risk when depressed.

While much work has been done to identify and understand the roles of social and behavioral factors in the transmission of STIs and HIV among MSM, few studies have explored how body image may have an impact on the spread of these infections. Using data from the 1998 Men's Health and Sexuality Survey that was collected by the Minnesota AIDS Project, we conducted a secondary analysis to examine the associations of body image with high-risk sexual behavior in MSM and the role of body satisfaction as a mediator in this relationship. We focused on perceived body image in this analysis on the premise that how a person views his own body is more likely to influence his sexual behavior than actual body mass. We hypothesized that a person's body image perception combined with his degree of body satisfaction would influence sexual risk taking. Specifically, we wished to explore whether MSM who perceived themselves as underweight, overweight, or obese would be more likely to be dissatisfied with their bodies than average-weight men and, as a result, engage more frequently in UAI.

Method

Respondents

The data for this study were collected from a cross-sectional survey of gay men who attended the 2-day 1998 Twin Cities LGBT Pride Festival in Minneapolis. The Twin Cities LGBT Pride Festival typically is attended by more than 100,000 LGBT persons. The festival features social, political, and recreational activities, including child-oriented activities for the many families parented by gay men and lesbians who attend the festival.

To be eligible for inclusion in the study, men must have been 18 years of age or older, male by birth (i.e., not transgendered from female to male), and primarily residents of the Minneapolis/St. Paul area. We limited the sample to the Minneapolis/St. Paul area in an effort to maintain regional cultural homogeneity. For example, safe sex norms may vary by geographic region and ethnicity; we geographically restricted enrollment to control for this potential source of variability. Of the 2,075 men approached for this study, 1,132 men (54.7%) agreed to participate. Of these, 394 men were not eligible because they were younger than 18 years of age, did not reside primarily in Minneapolis/St. Paul, or reported no sex with men in the past 2 years. Of the 738 eligible men, 185 were excluded either because they reported no recent sex with a man (n = 122) or were missing data on sexual behavior (n = 63). An additional 18 men were excluded because they did not report their body image, resulting in a final study sample of 535 men (72.5% of the eligible men who agreed to participate).

Measures

Respondent characteristics and high-risk sexual behavior. We collected information about respondent age, gender (i.e., male by birth or transgendered), race/ethnicity, education, sexual orientation, types of partnerships/sexual relationships, HIV status, sex/ club drug use (marijuana, crystal methamphetamine, cocaine, special K, poppers, or ecstasy), UAI and sexual positioning during anal sex in the past month. UAI was defined as unprotected anal intercourse with a primary partner of serodiscordant or unknown status, or any secondary partner.

Assessment of body image and body satisfaction. Body image was assessed by asking respondents to choose a male body drawing that best resembled them from the Figure Rating Scale adapted from Stunkard, Sorensen, and Schulsinger (1983; see Figure 1). The Figure Rating Scale has been used extensively in body image research due to its ease of administration and high test retest reliability ([alpha] = 0.74, Stunkard, Sorensen, & Schulsinger, 1983) and continues to be used by researchers for these reasons (Bhuiyan, Gustat, Srinivasan, & Berenson, 2003; Perez, Voelz, Pettit, & Joiner, 2002). We categorized both Body Mass Index (BMI) and the Figure Rating Scale responses as underweight, average weight, overweight, or obese as adopted by the National Institutes of Health (1998) and World Health Organization (1995). Body satisfaction was assessed by a single question asking respondents to rate on a five-point scale how satisfied they were with their present weight, with response options that ranged from not satisfied to extremely satisfied.

Procedure

The survey instrument was developed in collaboration with the Minnesota AIDS Project (MAP), a statewide nongovernmental organization dedicated to providing education, counseling, and other resources for all persons with HIV.

Given the diversity and family orientation of the Twin Cities LGBT Pride Festival, sampling from this setting afforded the best opportunity of enrolling a group of men who were highly reflective of the general gay-acculturated MSM community in terms of their risk for HIV infection and other STIs. Men were selected by a process designed to reduce self-selection and other biased recruitment practices. As described elsewhere (Bauer & Welles, 2001), gay and lesbian volunteers were trained to use random number tables to determine the order in which men were approached for participation. Men who agreed to participate were then brought by the volunteers to a tent where informed consent was obtained and they could complete the survey privately. Most men were able to complete the survey within 45 60 minutes. Oral informed consent was obtained to preserve anonymity. Methods for recruitment and consent were approved by the Institutional Review Board of the University of Minnesota.

[FIGURE 1 OMITTED]

Analysis

We conducted analysis with SAS statistical software, version 8.02 (SAS Institute Inc., Cary, NC). Due to the small number of MSM self-reporting as obese, the overweight and obese categories were combined in the analysis. In answering questions about sexual activities in the past month, respondents were instructed to leave questions blank if an activity did not apply to them. We therefore treated all unanswered sexual activity questions as "not applicable" in the analysis. In addition, 87 respondents did not provide information on race or age; we coded and included these subjects in our analyses to verify that these missing data were not systematically associated with our outcome of UAI and to improve the precision of the estimates in the final model. Chi-square tests were used to assess associations between respondent characteristics and perceived body image. A kappa statistic with an accompanying 95% confidence interval was calculated to evaluate the level of agreement between perceived body image and actual body mass. Unadjusted odds ratios with accompanying 95% confidence intervals and p values were used to examine associations among body image, UAI, and other factors.

In assessing the role of body satisfaction as a mediator of the relationship between body image and UAI, we used the methods described by Baron and Kenny (1986). For mediation to be present, the following conditions must be met: (1) perceived body image predicts body satisfaction; (2) perceived body image predicts UAI; and (3) body satisfaction predicts UAI while controlling for perceived body image. A final multiple logistic regression model then was developed to simultaneously control for all factors and evaluate possible confounding by age, race/Hispanic ethnicity, education, sexual relationship type, sexual positioning, alcohol use, and sex/club drug use. For the multivariate model, only factors that had at least a 10% effect on the main association of body image and UAI were retained as potential confounders (Greenland & Rothman, 1998).

Results

Demographics

As shown in Table 1, the majority of men in this study were White, gay, and well educated; the median age was 34.1 years. Seventy-one percent of the men reported an average weight body image, 14% perceived themselves as underweight, and 16% perceived themselves as overweight/obese.

Compared with men who reported a body image of average weight (median age, 34.1) or overweight/obese (median age, 38.9), underweight identifying men tended to be younger (median age, 29.7; p < .0001, Table 1). Overweight/obese men reported the lowest level of body satisfaction; men identifying as underweight reported the highest (p < .0001). Men identifying as average weight had the highest percentage of men engaging in UAI compared with either the underweight or overweight/obese groups (p < .05). More men identifying as underweight or average weight, however, had been both an insertive and receptive partner during anal sex in the past month compared with the overweight/ obese group. Conversely, more average weight or overweight/obese identifying men reported always being the receptive partner during anal sex in the past month compared with the underweight group (p < .05, Table 1).

Perceived Body Image Compared to Actual Body Mass

There was poor agreement between perceived body image and actual body mass in our sample (K = 0.13; 95% CI = 0.13, 0.80, Table 2). The majority of men who perceived themselves as underweight were, in fact, average weight based on their BMI. Among those perceiving themselves as average weight, approximately one-third were overweight based on BMI, and among those reporting overweight body image, approximately one-half were obese. Only men who perceived themselves as obese were all classified as obese based on BMI (Table 2).

Factors Associated With Unsafe Anal Intercourse

As shown in Table 3, the factors most strongly associated with UAI were: being age 25-35 compared with 45 or older (p < .05); being of Black race or Hispanic ethnicity (p < .001) compared with White; being in a sexually nonexclusive relationship (p < .0001) or not partnered (p < .001) compared with being exclusively partnered; always being the insertive partner compared with being versatile (i.e., decreased risk of UAI, p < .01); having 12 or fewer years of education (p < .05) compared with being a college graduate; and currently using at least one sex/club drug (p = .05) compared with no sex/club drug use. There was no association between UAI and the missing age/race data. In the regression models assessing for the mediational effect of body satisfaction (results from models not shown), while perceived body image independently predicted body satisfaction (p < .0001) and UAI (p < .05), body satisfaction did not predict UAI while controlling for perceived body image (p = .44). Body satisfaction was therefore not a mediator of the relationship between body image and UAI in this sample.

Adjusted Associations Between Body Image and Unsafe Anal Intercourse

Table 4 shows the adjusted odds ratios and 95% confidence intervals for the final model, adjusting for confounding factors. Of the factors we evaluated, only age, race/ethnicity, sexual relationship type, and sexual positioning were found to have a significant confounding effect on the relationship between body image and UAI. In this final model, MSM with underweight body images had 67% lower odds than those who identified as average weight of engaging in UAI (p < .05). Blacks and Hispanics had nearly four times the odds of UAI (p < .01) compared with Whites, and, compared with exclusively partnered MSM, men in nonexclusive relationships also had greatly increased odds of UAI (p < .0001) as did men who were not partnered (p < .001). Compared with men who were versatile, men who reported that they were always the receptive partner during anal sex had three times the odds of UAI (p < .0001). While not statistically significant at the p < .05 level, our model also suggests that men age 25 35 may have greater odds of UAI compared with men 45 and older (p = .07).

Discussion

Men who perceived themselves as underweight reported the highest body satisfaction and were less likely to report UAI than MSM with average body images. Silberstein and colleagues (1989) reported that, compared with heterosexual men, homosexual men considered appearance more central to their sense of self and they were more motivated to exercise by a desire to improve attractiveness. More recently, Kaminski and colleagues (2005) found that homosexual men diet more and are more fearful of becoming fat than heterosexual men. These findings suggest that a gay culture that promotes thinness also may contribute to a positive body image among underweight men and perhaps empower them to refrain from UAI. This interpretation agrees with recent research of adolescent women that suggested that a positive body image may contribute to women feeling more confident in refraining from risky sexual behavior including limiting their number of partners or insisting on using a condom with their partner (Wild, Flisher, Bhana, & Lombard, 2004; Wingood, DiClemente, Harrington, & Davies, 2002). While the study subjects for these two previous studies were women, it is possible that a similar process might be occurring among underweight-identifying MSM.

Conversely, men reporting as overweight/obese reported being more dissatisfied with their bodies than MSM of average weight, yet, like the underweight men, then also were less likely to report UAI. While no recent studies have investigated predictors of body shame in men, a study of college women in Australia reported that increases in body shame may be attributable to internalization of a thin ideal and appearance in comparison with others (Markham, Thompson, & Bowling, 2005). If a similar dynamic exists among overweight/obese MSM, then body shame may help explain the reduced odds of UAI among these men. A recent study reported that MSM with a better body image were more likely than men with lower body image to have had anal sex, even after adjusting for differences in body weight and age (Kraft, Robinson, Nordstrom, Bockting, & Rosser, 2006); this result also suggests that the opposite is true and that men with a lower body image are less likely to have had anal sex. If overweight/obese MSM are dissatisfied or ashamed of their bodies, they may be reluctant to engage in any sexual activity that reveals their bodies. As a result, oral sex or mutual/ solo masturbation could be the sexual activities of choice for overweight/obese men, and our reduced odds of UAI in this case may reflect their not having any anal intercourse rather than simply not engaging in UAI.

Interestingly, our study also revealed that nearly all of the men in our sample perceived themselves as thinner than they actually were. This body dysmorphia, perceiving oneself to be thinner or more muscular than actual, has become an emerging theme in body image studies among men. Based on a comprehensive review of the literature from 1985 to 2005, Leone and colleagues (Leone, Sedory, & Gray, 2005) also noted that societal pressure on men to appear more muscular and lean has contributed to this trend. This "muscularity craze" has affected both gay and heterosexual men. For example, Duggan and McCreary (2004) found in their study of body image and muscularity among gay and heterosexual men that purchasing fitness magazines was positively correlated with levels of body dissatisfaction for both groups. In fact, gay and heterosexual men are likely very similar in both their body ideals and body-image distortion (Hausmann, Mangweth, Walch, Rupp, & Pope, 2004).

Based on 1998 data from the Centers for Disease Control and Prevention (CDC), 57.0% of the AIDS cases diagnosed in 1998 were among MSM aged 25-39 (1999); this is consistent with our finding of increased risk of UAI among the 25-35-year-old men in this study. The increased odds relative to Whites of UAI among Blacks and Hispanics also aligns well with results from previous studies on STI/HIV transmission. It is well established that individuals of Black race and Hispanic ethnicity are at higher risk for STI/HIV transmission, and recent reports from the CDC confirm that these groups remain at high risk (2005a, 2005b).

Our finding of increased odds of UAI among nonexclusively partnered or not partnered MSM may be attributable, in part, to the increased sexual freedom that often is present in these relationships. This is supported by a study conducted by Crawford and colleagues (2003) that reported MSM in nonexclusive relationships have greater levels of sensation seeking than MSM in exclusive relationships. Previous studies have shown sexually adventurous and sensation-seeking gay men to be more likely to have multiple partners and unsafe sex (Difranceisco, Ostrow, & Chmiel, 1996; Halkitis & Parsons, 2003; Kalichman, Cain, Zweben, & Swain, 2003; Kalichman et al., 1994). Sensation seeking has also been reported as a specific predictor of UAI among MSM (Chng & Geliga-Vargas, 2000).

Unfortunately, our initial hypothesis was not fully supported by our results. Body satisfaction was not found to be a mediator of the relationship between body image and UAI, and this could be attributable to our measurement of body satisfaction with a single question. A validated and detailed body satisfaction scale, asking several questions about satisfaction with specific areas of the body, such as the abdomen or arms, may be necessary to accurately assess the role of body satisfaction in relation to body image and UAI.

This study has some limitations. Because no information was collected from men who did not agree to participate, self-selection may limit the generalizability of the behavioral and psychosocial measures for this study. In addition, information was obtained from self-report and from only one person in a partnership or sexual relationship.

Since these data were originally collected in 1998, it is also important to consider how the generalizability of our results might be affected by the passage of time. While the rates of men self-reporting as underweight, average weight, or overweight/obese may change over time in response to fitness marketing and cultural factors, we do think that, given the consistency of our results with the more recent studies cited above, the associations with UAI we have identified in this study have not significantly changed. Furthermore, we believe that because the role of body image in sexual behavior among MSM has not been widely studied, our results provide an important reference point to compare with future studies.

Due to homophobia and a lack of widespread acceptance of gays and lesbians, many researchers of gay and lesbian health concerns cannot readily recruit study respondents from the general population and instead use gay-oriented sources such as gay bars, pornographic catalogue mailing lists, community events, gay and lesbian student groups, and local bathhouses, cafes, and retail outlets (Remafedi, 2002; Siever, 1994; Van de Ven et al., 1997; Van De Ven, Rawstorne, Crawford, & Kippax, 2002). Each of these sources is problematic in that no single one is necessarily representative of the gay and lesbian target population one wishes to study, potentially limiting the generalizability of the study results. This is also a limitation of this study because respondents were selected from a gay community event--the Twin Cities LGBT Pride Festival. While such festivals attract a very large number of people, gay men with heavily internalized homophobia and nongay identifying MSM were not well represented in this study. Despite these limitations, however, the method of random selection employed in recruiting respondents provided us with a study sample that is more diverse with respect to age and likely is more representative of the larger gay-acculturated MSM community than previous studies.

This study has potentially important implications for future HIV and STI prevention efforts among MSM. Further research is needed to identify the specific factors that contribute to decreased risk of UAI among MSM who perceive themselves as underweight and determine if these factors are desirable and replicable for individuals with other body image types. While body satisfaction in our study was not a mediator between body image and sexual risk taking, the true mediator(s) may be other factors that were mot measured in this study such as anxiety/depression, sensation seeking, self-esteem, social acceptance, and sexual power dynamics. Specific research on how these other factors are associated with body image and risk taking among MSM could reveal exactly how body image operates to inform partnerships and influence sexual behavior among MSM.

References

Agliata, D. & Tantleff-Dunn, S. (2004). The impact of media exposure on males' body image. J Soc Clin Psychol, 23(1), 7-22.

Bancroft, J., Janssen, E., Strong, D., & Vukadinovic, Z. (2003). The relation between mood and sexuality in gay men. Arch Sex Behav, 32(3), 231-242.

Baron, R. M. & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. J Pers Soc Psychol, 51(6), 1173-1182.

Barry, D. T., Grilo, C. M., & Masheb, R. M. (2002). Gender differences in patients with binge eating disorder. Int J Eat Disord, 31(1), 63-70.

Bauer, G. R. & Welles, S. L. (2001). Beyond assumptions of negligible risk: Sexually transmitted diseases and women who have sex with women. Am J Public Health, 91(8), 1282-1286.

Beren, S. E., Hayden, H. A., Wilfley, D. E., & Grilo, C. M. (1996). The influence of sexual orientation on body dissatisfaction in adult men and women, Int J Eat Disord, 20(2), 135-141.

Bhuiyan, A. R., Gustat, J., Srinivasan, S. R., & Berenson, G. S. (2003). Differences in body shape representations among young adults from a biracial (Black-White), semirural community: The Bogalusa Heart Study. Am J Epidemiol, 158(8), 792-797.

Bohne, A., Keuthen, N. J., Wilhelm, S., Deckersbach, T., & Jenike, M. A. (2002). Prevalence of symptoms of body dysmorphic disorder and its correlates: A cross-cultural comparison. Psychosomatics, 43(6), 486-490.

Carlat, D. J. & Camargo, C. A., Jr. (1991). Review of bulimia nervosa in males. Am J Psychiatry, 148(7), 831-843.

Carlat, D. J., Camargo, C. A., Jr., & Herzog, D. B. (1997). Eating disorders in males: A report on 135 patients. Am J Psychiatry, 154(8), 1127-1132.

Centers for Disease Control and Prevention (CDC). (1999). HIV surveillance supplemental report, 8.

Centers for Disease Control and Prevention (CDC). (2005a). HIV/AIDS surveillance report, 2004. 6-7.

Centers for Disease Control and Prevention (CDC). (2005b). Sexually transmitted disease surveillance, 63-69.

Chng, C. L. & Geliga-Vargas, J. (2000). Ethnic identity, gay identity, sexual sensation seeking and HIV risk taking among multiethnic men who have sex with men. AIDS Educ Prevent, 12(4), 326-339.

Crawford, I., Hammack, P. L., McKirnan, D. J., Ostrow, D., Zamboni, B. D., Robinson, B., et al. (2003). Sexual sensation seeking, reduced concern about HIV and sexual risk behaviour among gay men in primary relationships. AIDS Care, 15(4), 513-524.

Difranceisco, W., Ostrow, D. G., & Chmiel, J. S. (1996). Sexual adventurism, high-risk behavior, and human immunodeficiency virus-I seroconversion among the Chicago MACS-CCS cohort, 1984 to 1992. A case-control study. Sex Transm Dis, 23(6), 453-460.

Duggan, S. J. & McCreary, D. R. (2004). Body image, eating disorders, and the drive for muscularity in gay and heterosexual men: The influence of media images. J Homosex, 47(3-4), 45-58.

Greenland, S. & Rothman, K. J. (1998). Introduction to stratified analysis. In K. J. Rothman & S. Greenland (Eds.), Modern epidemiology, (2nd ed., pp. 253-279). Philadelphia: Lippincott, Williams and Wilkins.

Halkitis, P. N. & Parsons, J. T. (2003). Intentional unsafe sex (bare-backing) among HIV-positive gay men who seek sexual partners on the Internet. AIDS Care, 15(3), 367-378.

Hausmann, A., Mangweth, B., Walch, T., Rupp, C. I., & Pope, H. G. (2004). Body-image dissatisfaction in gay versus heterosexual men: Is there really a difference? J Clin Psychiatry, 65(11), 1555-1558.

Herzog, D. B., Newman, K. L., & Warshaw, M. (1991). Body image dissatisfaction in homosexual and heterosexual males. J Nerv Ment Dis, 179(6), 356-359.

Kalichman, S. C., Cain, D., Zweben, A., & Swain, G. (2003). Sensation seeking, alcohol use and sexual risk behaviors among men receiving services at a clinic for sexually transmitted infections. J Stud Alcohol, 64(4), 564-569.

Kalichman, S. C., Johnson, J. R., Adair, V., Rompa, D., Multhauf, K., & Kelly, J. A. (1994). Sexual sensation seeking: Scale development and predicting AIDS-risk behavior among homosexually active men. J Pers Assess, 62(3), 385-397.

Kaminski, P. L., Chapman, B. P., Haynes, S. D., & Own, L. (2005). Body image, eating behaviors, and attitudes toward exercise among gay and straight men. Eat Behav, 6(3), 179-187.

Kraft, C., Robinson, B. B., Nordstrom, D. L., Bockting, W. O., & Rosser, B. R. (2006). Obesity, body image, and unsafe sex in men who have sex with men. Arch Sex Behav, 35(5), 587-595.

Leit, R. A., Gray, J. J., & Pope, H. G., Jr. (2002). The media's representation of the ideal male body: A cause for muscle dysmorphia? Int J Eat Disord, 31(3), 334-338.

Leone, J. E., Sedory, E. J., & Gray, K. A. (2005). Recognition and treatment of muscle dysmorphia and related body image disorders. J Athl Train, 40(4), 352-359.

Markham, A., Thompson, T., & Bowling, A. (2005). Determinants of body-image shame. Pers Indiv Differ, 38(7), 1529-1541.

Martin, J. I. & Knox, J. (1997a). Loneliness and sexual risk behavior in gay men. Psychol Rep, 81(3 Pt 1), 815-825.

Martin, J. I. & Knox, J. (1997b). Self-esteem instability and its implications for HIV prevention among gay men. Health Soc Work, 22(4), 264-273.

National Institutes of Health. (1998). Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults--The evidence report. Obes Res, 6(Suppl 2), 51S-209S.

Perez, M., Voelz, Z. R., Pettit, J. W., & Joiner, T. E., Jr. (2002). The role of acculturative stress and body dissatisfaction in predicting bulimic symptomatology across ethnic groups. Int J Eat Disord, 31(4), 442-454.

Phillips, K. A. & Diaz, S. F. (1997). Gender differences in body dysmorphic disorder. J Nerv Ment Dis, 185(9), 570-577.

Pope, H. G., Jr., Phillips, K. A., & Olivardia, R. (2000). Secrets of the men at the Olympic gym. In The Adonis complex: The secret crisis of male body obsession (pp. 1-26). New York: Free Press.

Remafedi, G. (2002). Suicidality in a venue-based sample of young men who have sex with men. J Adolesc Health, 31(4), 305-310.

Robb, A. S. & Dadson, M. J. (2002). Eating disorders in males. Child Adolesc Psychiatr Clin N Am, 11(2), 399-418, xi.

Siever, M. D. (1994). Sexual orientation and gender as factors in socio-culturally acquired vulnerability to body dissatisfaction and eating disorders. J Consult Clin Psychol, 62(2), 252-260.

Silberstein, L. R., Mishkind, M. E., Striegel-Moore, R. H., Timko, C., & Rodin, J. (1989). Men and their bodies: A comparison of homosexual and heterosexual men. Psychosom Med, 51(3), 337-346.

Stunkard, A. J., Sorensen, T., & Schulsinger, F. (1983). Use of the Danish Adoption Register for the study of obesity and thinness. Res Publ Assoc Res Nerv Ment Dis, 60, 115-120.

Van de Ven, P., Kippax, S., Crawford, J., French, J., Prestage, G., Grulich, A., et al. (1997). Sexual practices in a broad cross-sectional sample of Sydney gay men. Aust N Z J Public Health, 21(7), 762-766.

Van de Ven, P., Rawstorne, P., Crawford, J., & Kippax, S. (2002). Increasing proportions of Australian gay and homosexually active men engage in unprotected anal intercourse with regular and with casual partners. AIDS Care, 14(3), 335-341.

Vartanian, L. R., Giant, C. L., & Passino, R. M. (2001). Ally McBeal vs. Arnold Schwarzenegger: Comparing mass media, interpersonal feedback and gender as predictors of satisfaction with body thinness and muscularity. Soc Behav Personal, 29, 711-723.

Wild, L. G., Flisher, A. J., Bhana, A., & Lombard, C. (2004). Associations among adolescent risk behaviours and self-esteem in six domains. J Child Psychol Psychiatry, 45(8), 1454-1467.

Wingood, G. M., DiClemente, R. J., Harrington, K., & Davies, S. L. (2002). Body image and African American females' sexual health. J Womens Health Gend Based Med, 11(5), 433-439.

World Health Organization. (1995). Physical status: The use and interpretation of anthropometry. Report of a WHO Expert Committee. Worm Health Organ Tech Rep Ser, 854, 1-452.

Yager, J., Kurtzman, F., Landsverk, J., & Wiesmeier, E. (1988). Behaviors and attitudes related to eating disorders in homosexual male college students. Am J Psychiatry, 145(4), 495-497.

Donald Allensworth-Davies

Data Coordinating Center, Boston University School of Public Health

Seth L. Welles

Department of Epidemiology and Biostatistics, Drexel University School of Public Health

Wendy L. Hellerstedt

Division of Epidemiology and Community Health, University of Minnesota School of Public Health

Michael W. Ross

Department of Behavioral Sciences, University of Texas School of Public Health

Correspondence should be addressed to Donald Allensworth-Davies, Data Coordinating Center, Boston University School of Public Health, 715 Albany Street, Crosstown Center-3, Boston, Massachusetts 02118, USA. E-mail: [e-mail].

Table 1. Demographics and Characteristics of Men Who Have Sex
With Men by Body Image (N = 535)

                                            Underweight
Characteristic                                (N = 73)

Age (p < .0001), N (%)
    18-25                                    15 (20.5)
    25-35                                    40 (54.8)
    35-45                                    10 (13.7)
    45+                                       4 (5.5)
    Missing                                   4 (5.5)
Race/ethnicity (p = .16), N (%)
    White                                    60 (82.2)
    Black/Hispanic                            2 (2.7)
    Asian                                     5 (6.8)
    Missing                                   6 (8.2)
Education (p = .83), N (%)
    Some high school/high school
      graduate                                7 (9.6)
    Trade or technical school/some
      college                                22 (30.1)
    College graduate                         27 (37.0)
    Graduate/professional school             17 (23.3)
Sexual orientation (p = .67), N (%)
    Gay                                      71 (97.3)
    Straight                                  0 (0.0)
    Bisexual                                  0 (0.0)
    No preference but have sex with men       2 (2.7)
    Missing                                   0 (0.0)
Body satisfaction (a) (p < .0001),
  mean [+ or -] SD                        3.8 [+ or -] 1.0
Sexual relationship type at time of
  survey (p = .21), N (%)
    Exclusively partnered with a man         31 (42.5)
    Partnered with man or woman, sex
      outside relationship
      (nonexclusive)                          9 (12.3)
    Not partnered, multiple and
      one-time partners                      33 (45.2)
Currently uses at least one sex or club
  drug (p = .34), N (%)                       8 (11.0)
Positioning during anal sex in past
  month (p = .02), N (%)
    Always the insertive partner             14 (19.2)
    The insertive and receptive partner      25 (34.2)
    Always the receptive partner              7 (9.6)
    Missing                                  27 (37.0)
Engaged in anal sex in the past month
  (p = .10), N (%)
    Yes                                      46 (63.0)
    No                                       15 (20.6)
    Missing                                  12 (16.4)
Engaged in oral sex in the past month
  (p = .89), N (%)
    Yes                                      70 (95.9)
    No                                        3 (4.1)
    Missing                                   0 (0.0)
Engaged in mutual masturbation in the
  past month (p = .44), N (%)
    Yes                                      65 (89.0)
    No                                        8 (11.0)
Average number of times had sex in
  past month (anal, oral, or mutual
  masturbation) (p = .39),
  mean [+ or -] SD                        7.2 [+ or -] 5.6
Average number of partners over the
  past month (p = .66), mean
  [+ or -] SD                             2.2 [+ or -] 2.0
Engaged in UAI with partner of
  serodiscordant/unknown status or
  with any secondary partner in
  past month (p = .01), N (%)                 6 (8.2)

                                            Underweight
Characteristic                                (N = 73)

Age (p < .0001), N (%)
    18-25                                    15 (20.5)
    25-35                                    40 (54.8)
    35-45                                    10 (13.7)
    45+                                       4 (5.5)
    Missing                                   4 (5.5)
Race/ethnicity (p = .16), N (%)
    White                                    60 (82.2)
    Black/Hispanic                            2 (2.7)
    Asian                                     5 (6.8)
    Missing                                   6 (8.2)
Education (p = .83), N (%)
    Some high school/high school
      graduate                                7 (9.6)
    Trade or technical school/some
      college                                22 (30.1)
    College graduate                         27 (37.0)
    Graduate/professional school             17 (23.3)
Sexual orientation (p = .67), N (%)
    Gay                                      71 (97.3)
    Straight                                  0 (0.0)
    Bisexual                                  0 (0.0)
    No preference but have sex with men       2 (2.7)
    Missing                                   0 (0.0)
Body satisfaction (a) (p < .0001),
  mean [+ or -] SD                        3.8 [+ or -] 1.0
Sexual relationship type at time of
  survey (p = .21), N (%)
    Exclusively partnered with a man         31 (42.5)
    Partnered with man or woman, sex
      outside relationship
      (nonexclusive)                          9 (12.3)
    Not partnered, multiple and
      one-time partners                      33 (45.2)
Currently uses at least one sex or club
  drug (p = .34), N (%)                       8 (11.0)
Positioning during anal sex in past
  month (p = .02), N (%)
    Always the insertive partner             14 (19.2)
    The insertive and receptive partner      25 (34.2)
    Always the receptive partner              7 (9.6)
    Missing                                  27 (37.0)
Engaged in anal sex in the past month
  (p = .10), N (%)
    Yes                                      46 (63.0)
    No                                       15 (20.6)
    Missing                                  12 (16.4)
Engaged in oral sex in the past month
  (p = .89), N (%)
    Yes                                      70 (95.9)
    No                                        3 (4.1)
    Missing                                   0 (0.0)
Engaged in mutual masturbation in the
  past month (p = .44), N (%)
    Yes                                      65 (89.0)
    No                                        8 (11.0)
Average number of times had sex in
  past month (anal, oral, or mutual
  masturbation) (p = .39),
  mean [+ or -] SD                        7.2 [+ or -] 5.6
Average number of partners over the
  past month (p = .66), mean
  [+ or -] SD                             2.2 [+ or -] 2.0
Engaged in UAI with partner of
  serodiscordant/unknown status or
  with any secondary partner in
  past month (p = .01), N (%)                 6 (8.2)

                                          Overweight/Obese
Characteristic                                (N = 83)

Age (p < .0001), N (%)
    18-25                                     6 (7.2)
    25-35                                    18 (21.7)
    35-45                                    25 (30.1)
    45+                                      26 (31.3)
    Missing                                   8 (9.7)
Race/ethnicity (p = .16), N (%)
    White                                    69 (83.1)
    Black/Hispanic                            5 (6.0)
    Asian                                     0 (0.0)
    Missing                                   9 (10.8)
Education (p = .83), N (%)
    Some high school/high school
      graduate                                5 (6.0)
    Trade or technical school/some
      college                                25 (30.1)
    College graduate                         28 (33.7)
    Graduate/professional school             25 (30.1)
Sexual orientation (p = .67), N (%)
    Gay                                      82 (98.8)
    Straight                                  0 (0.0)
    Bisexual                                  1 (1.2)
    No preference but have sex with men       0 (0.0)
    Missing                                   0 (0.0)
Body satisfaction (a) (p < .0001),
  mean [+ or -] SD                        2.1 [+ or -] 1.0
Sexual relationship type at time of
  survey (p = .21), N (%)
    Exclusively partnered with a man         26 (31.3)
    Partnered with man or woman, sex
      outside relationship
      (nonexclusive)                         15 (18.1)
    Not partnered, multiple and
      one-time partners                      42 (50.6)
Currently uses at least one sex or club
  drug (p = .34), N (%)                      16 (19.3)
Positioning during anal sex in past
  month (p = .02), N (%)
    Always the insertive partner             11 (13.2)
    The insertive and receptive partner      17 (20.5)
    Always the receptive partner             17 (20.5)
    Missing                                  38 (45.8)
Engaged in anal sex in the past month
  (p = .10), N (%)
    Yes                                      45 (54.2)
    No                                       23 (27.7)
    Missing                                  15 (18.1)
Engaged in oral sex in the past month
  (p = .89), N (%)
    Yes                                      78 (94.0)
    No                                        5 (6.0)
    Missing                                   0 (0.0)
Engaged in mutual masturbation in the
  past month (p = .44), N (%)
    Yes                                      68 (81.9)
    No                                       15 (18.1)
Average number of times had sex in
  past month (anal, oral, or mutual
  masturbation) (p = .39),
  mean [+ or -] SD                        6.9 [+ or -] 7.4
Average number of partners over the
  past month (p = .66), mean
  [+ or -] SD                             2.4 [+ or -] 3.4
Engaged in UAI with partner of
  serodiscordant/unknown status or
  with any secondary partner in
  past month (p = .01), N (%)                11 (13.3)

(a) 1 = Not at all satisfied; 3 = Not satisfied or unsatisfied;
5 = Extremely satisfied.

Table 2. Comparison of Perceived Body Image to Actual
Body Mass (N = 535).

                                Perceived Body Image

                 Underweight  Average Weight   Overweight    Obese
Actual Body       (N = 73)      (N = 379)      (N = 75)     (N = 8)
Mass                (%)            (%)            (%)         (%)

Underweight       7 (9.6)         2 (0.5)       0 (0.0)     0 (0.0)
Average weight   63 (86.3)      233 (61.5)      5 (6.7)     0 (0.0)
Overweight        3 (4.1)       128 (33.8)     30 (40.0)    0 (0.0)
Obese             0 (0.0)        16 (4.2)      40 (53.3)    8 (100.0)

Table 3. Factors Associated With Unsafe Anal Intercourse in
the Past Month Among Men Who Have Sex With Men
(N = 535)
                                   Unadjusted OR
Description                          (95% CI)          p

Body image
  Underweight                    0.32 (0.14, 0.77)   .008
  Average weight                 1.00
  Overweight/obese               0.55 (0.28, 1.09)   .08
Age
  18-25                          0.93 (0.35. 2.45)   .88
  25-35                          2.06 (1.01. 4.20)   .04
  35-45                          2.02 (0.96. 4.26)   .06
  45+                            1.00
  Missing                        0.71 (0.18. 2.72)   .62
Race/ethnicity
  White                          1.00
  Black/Hispanic                 4.07 (1.76, 9.44)   .0005
  Asian                          1.31 (0.36, 4.82)   .68
  Missing                        0.94 (0.44, 2.01)   .88
Education
  Some high school/high
    school graduate              2.28 (1.08, 4.77)   .02
  Trade or technical
    school/some college          1.00 (0.59, 1.69)   .99
  College graduate               1.00
  Graduate/professional
    school                       0.68 (0.36, 1.27)   .22
Sexual relationship type
  at time of survey
  Exclusively partnered          1.00
    with a man
  Partnered with man
    or woman, sex outside
    relationship
    (nonexclusive)               5.92 (3.13, 11.21)  <.0001
  Not partnered, multiple
    and one-time partners        2.72 (1.50, 4.94)   .0007
Positioning during anal
  sex in past month
  Always the insertive partner   0.35 (0.17, 0.73)   .004
  The insertive and
    receptive partner            1.00
  Always the receptive
    partner                      1.35 (0.80, 2.30)   0.26
  Missing (a)                    --                  --
Currently uses at least one
  sex or club drug               1.70 (0.99, 2.90)   0.05

(a) An odds ratio could not be estimated since all of the
respondents with missing data in this category reported
having safe anal intercourse.

Table 4. Adjusted Associations Between Body Image and
Unsafe Anal Intercourse Among Men Who Have Sex With Men
(N = 535)
                                     Adjusted OR
Independent Variables                  (95% CI)          p

Body image
  Underweight                     0.33 (0.13, 0.85)     .02
  Average weight                  1.00
  Overweight/obese                0.57 (0.27, 1.19)     .13
Age
  18-25                           1.19 (0.42, 3.35)     .74
  25-35                           2.03 (0.94, 4.39)     .07
  35-45                           1.94 (0.88, 4.25)     .10
  45+                             1.00
  Missing                         0.66 (0.16, 2.66)     .56
Race/ethnicity
  White                           1.00
  Black/Hispanic                  3.98 (1.50, 10.55)    .006
  Asian                           1.03 (0.52, 4.27)     .97
  Missing                         0.69 (0.30, 1.58)     .38
Sexual relationship
  type at time of survey
  Exclusively partnered           1.00
  Partnered with man or           5.78 (2.96, 11.29)   <.0001
    woman, sex outside
    relationship (nonexclusive)
  Not partnered, multiple         3.20 (1.72, 5.93)     .0002
    and one-time partners
Positioning during anal
  sex in past month
  Always the insertive            0.80 (0.37, 1.70)    0.56
    partner
  The insertive and               1.00
    receptive partner
  Always the receptive            3.33 (1.87, 5.93)    <.0001
    partner