Abstracts
James D. Weinrich, Ph.D.

Weinrich, James D.; Susan Little, Richard Haubrich, Jimmy Hwang, J. Allen McCutchan, and the California Collaborative Treatment Group (CCTG) (March 1997) Personality and Psychological Predictors of Adherence: Patient Self-Reports Versus Physician Provider Estimates. Poster and oral presentation at the 14th annual investigators' meeting of the Universitywide AIDS Research Program, San Francisco.

Objective of Project: California Collaborative Treatment Group (CCTG) protocol 570 is designed to study the effects of giving patients and providers feedback on HIV-1 RNA level and CD4 count during antiretroviral therapy. The present substudy, CCTG 570A (for Adherence), is a multidisciplinary, psychosocial/clinical assessment of adherence to the drug regimens prescribed for the patients by their medical care providers.

To date, we have studied just over 100 women and men using questionnaires assessing their self-reported, recalled adherence to their drug regimens; their estimate of the likelihood that they will complete the 570 protocol; recreational drug use; several psychosocial or personality factors hypothesized to affect the probability of adherence; and other variables such as CD4 count and viral load. We also asked their providers to estimate how likely each of these patients would be to complete the study.

Significant Findings: Sample size ranges from 93 to 110, depending upon the number of responses to different questions. A cluster analysis of items comprising a personality and coping strategies questionnaire yielded four subscales, with the following names and sample items:

• High patient scores on MMPI L-F-K (odd answers) were significantly associated with the providers' rating their patients' likelihood to complete the protocol as less than "Excellent."
• High patient scores on Entitlement were associated at a trend level (p < 0.07) with a provider rating of less than "Excellent."
• High scores on Procrastination were associated with higher levels of self-admitted failure to take medications regularly, and with higher levels of self-admitted use of recreational drugs.
• High scorers on Sensible problem-solving were more likely to state that they found it "very easy" to keep regularly scheduled doctor's appointments, as well as study appointments.

Conclusions: Personality and Coping-style questions show promise in predicting adherence to complicated clinical trials such as CCTG 570, and to other variables of interest in HIV research (e.g., drug use). Moreover, physician estimates of the probability of patient compliance may reflect an informal but well-founded assessment of underlying personality styles in the patients they treat.

Weinrich, James D. (21 March 1996) Childhood personality and nonsexual abuse predict adult sexual role preferences. Poster and oral presentation at the 13th annual investigators' meeting of the Universitywide AIDS Research Program, San Francisco.

Objective of Project: We are conducting a precedent-setting 3-year study of motivations underlying preferences for sexual acts potentially involved in HIV transmission. In the first phase of AIDS prevention research, models typically presumed that heterosexual (HT) and homosexual (HM) men's erotic preferences are completely different, and then treated each kind of man as a uniform "type." We strongly disagree, and will report at the meeting in March, for the first time anywhere, strong empirical evidence that HM men, at least, can be separated into erotic subgroups in a way which is meaningful for AIDS educational and intervention efforts. Moreover, these subgroups are probably of lasting significance, since they correlate with certain childhood events and childhood personality.

Soon, we will try to identify some of the deeper reasons why certain risk group members have resisted the safer-sex interventions to which others have responded. We are not yet ready to present these data.

To date, we have studied 119 men using questionnaires and a 90-minute clinically-oriented interview to ascertain each man's lovemap: his image of an ideal, optimally arousing sexual partner and a maximally arousing interaction with that partner. We will soon add data on personality, coping skills, depression, and disease staging from a larger study in which the men took part.

Significant Findings: The following results pertain to HM men; sample size does not yet permit a similar analysis of HT men.

Sexual orientation per se is not predicted by parental factors. However, childhood physical abuse, neglect, and paternal alcoholism are associated with erotically submissive lovemaps in adulthood. Childhood sexual abuse is not. The same factors are associated with a preference for receptive anal sex, but not with a preference for or against insertive anal sex.

Replicating a result we have previously obtained, there was a significant correlation between childhood core gender identity nonconformity and adult preference for receptive anal intercourse (RAI). In the present dataset, this correlation is exceptionally strong between adult core gender identity and preference for RAI. We used multiple logistic regression to demonstrate for the first time that this personality effect (itself highly correlated with the childhood personality trait) and the child abuse/neglect effect were independently and additively associated with preferences for a receptive-anal role in adulthood.

Conclusions: We need to appreciate men's erotic variability, understand how life history parameters trigger erotic value in sexual scenarios, and use that understanding to personalize messages for change. Even when preferences for particular sexual acts seem fixed, interventions may increase effectiveness by taking account of the deeper reasons why men prefer the acts they do.

Weinrich, James D., Atkinson, J. Hampton, McCutchan, J. Allen, Grant, Igor, and the HNRC Group (9 July 1996) Buried Pleasures: Childhood environment predicts preferences for certain safe and unsafe sexual acts. Poster #Tu.D.2714 presented at the XI International AIDS Conference, Vancouver BC, Canada, abstracts book page 387.

Issue: Too little attention has been given to erotic variables in AIDS research. Models typically presume that heterosexual (HT) and homosexual (HM) men's erotic preferences are completely different, and then treat each kind of man as a uniform "type." We strongly disagree, and present new data and a new model to understand such preferences.

Objectives: We sought: (1) to identify childhood and other antecedents for both HM and HT men associated with sexual identification; (2) to ascertain erotic subgroups of these men relevant to HIV transmission; and (3) to suggest strategies based on this erotic typology which would help fight transmission (e.g., by devising appeals to particular subgroups).

Methods: We studied 123 men using questionnaires and a 90-minute clinically-oriented interview to ascertain each man's lovemap: his image of an ideal, optimally arousing sexual partner and a maximally arousing interaction with that partner. We added data on personality, and coping skills from a larger study in which the men took part.

Results: One typical result: childhood physical abuse, neglect, and parental alcoholism (but not sexual abuse) are associated with submissive sexual lovemaps in adulthood for HM and HT men. Sexual orientation per se is not predicted by parental factors. Among HM men, such fantasies (as well as childhood gender nonconformity) are associated with receptive anal sex.

Conclusions: Even when preferences for particular sexual acts seem fixed, interventions may increase effectiveness by taking account of the deeper reasons why men prefer the acts they do. We need to appreciate men's erotic variability, understand how deeply buried pleasures trigger erotic value in sexual scenarios, and use that understanding to personalize messages for change.

NOTE: Above text has been updated slightly from abstract published in the official program.

Weinrich, James D. (1995b) Biological research on sexual orientation: A critique of the critics. In Sex, cells, and same-sex desire: The biology of sexual preference, John P. De Cecco and David Allen Parker, Eds., Haworth Press, pp. 197-213.

Evolutionary biologists are tired of being accused of being too biologically deterministic, by critics who have little understanding of what biological or evolutionary theories actually imply. Misunderstandings came about because social-science disciplines often do not share evolutionary biology's tendency to build into their models multiple "normal" paths of development. Sociobiologists first explained homosexuality adaptively because they first try to explain everything adaptively. Most nonbiologists are unaware of this very strong evolutionary tradition.

It is now fashionable to discount scientific objectivity, but there are many examples of where such an attack is unwarranted. Kinsey produced a nontypological theory of sexual orientation in spite of his history as a taxonomist. Sociobiologists produced a nonpathological explanation of nonreproductive homosexuality in spite of the centrality of reproductive success in their models.

In judging whether a discipline is particularly likely to be misused in social debates, one must perform the appropriate intellectual "controls." One must examine appropriate uses as well as misuses, and one must examine other disciplines to see whether there are differences in the relative likelihood of abuse. Indeed, many social-science theories have been even more clearly abused than biological ones.

Little attention has been given to dispositional variables in relapse into unsafe sex. In two previous papers, we showed (1) that recurrent adult depression in the gay/bisexual men in our sample is often associated with high gender nonconformity, especially core gender dysphoria, in childhood, (2) that coping strategies and personality scores are associated with changes from 1979 to 1989 in unsafe sex -- defined as receiving semen rectally without a condom (RSR), and (3) that sexual identity conflicts are related to a general style of escape-avoidant coping, (4) which is in turn associated with the MMPI profiles previously shown to correlate with unsafe sex. Here we break down these analyses to see if these patterns are related to HIV status.

We studied over 500 gay/bisexual men who contributed SCID interviews, Freund Feminine Gender Identity scale scores, MMPI-2 profiles, Ways of Coping (Revised) scores, Profile of Mood State (POMS) scores, and sex histories (ns varied by analysis).

Concerning the relationship between FGI and depression (1), we found that the correlations persisted in the HIV positive subsample, but vanished to insignificance among the HIV negative controls. A similar pattern emerged in the correlations pertaining to MMPI scores and unsafe sex (2), as well as in the association between escape-avoidant coping and sexual identity conflict (3). The analyses correlating the MMPI scores with escape-avoidant coping were, in contrast, equally strongly associated when broken down by HIV status.

A series of sample selection biases could account for these results. It would be more parsimonious, however, to hypothesize a deeper rationale. We suggest more investigation of the possibility that psychological and sexological variables may have more intriguing effects on sexual behaviors or the dispositional factors (perhaps even immunologic ones) leading to HIV exposure than has been recognized to date.

To evaluate the validity of the Quality of Well-Being Scale (QWB) for studies of patients with human immunodeficiency virus (HIV) disease, 514 men were studied who were divided among four categories: Centers for Disease Control and Prevention (CDC) Group A (N = 272), CDC-B (N = 81), CDC-C (N = 47), and uninfected male controls (N = 114). The QWB and a variety of medical, neuropsychological, and biochemical measures were administered to all participants. When QWB scores were broken down by HIV group, the CDC-C group was significantly lower (.614) than the CDC-B (.679), CDC-A (.754), or control group (.801). The difference between Groups CDC-C and CDC-A was about .14 units of well-being, which suggests that individuals lose 1/7 equivalents of 1 well year of life for each year they are in Group CDC-C in comparison to the asymptomatic group (Group CDC-A). In comparison to the controls, this would equal a 1-year of life loss for each seven infected individuals. The QWB was shown to be significantly associated with CD4+ lymphocytes (p < .001), clinician ratings of neuropsychological impairment (p < .04), neurologists ratings of dysfunction (p < .001), and all subscales of the Profile of Mood States. Baseline QWB scores were significant prospective predictors of death over a median follow-up time of 30 months. Multivariate models demonstrated high covariation between predictors of QWB. It was concluded that the QWB is a significant correlate of biological, neuropsychological, neurological, psychiatric, and mortality outcomes for male HIV-infected patients.

Few if any studies before the AIDS epidemic suggested that male homosexuals may on average have higher levels of depression than male heterosexuals. However, several samples of homosexual and bisexual men in HIV studies suggest that depression and anxiety are high in these populations, and that this psychiatric morbidity began before the AIDS epidemic. We tested the hypothesis that high childhood gender nonconformity (CGN) is associated with depression and anxiety, and so might account for differences in these variables among samples of homosexuals. A total of 254 homosexual or bisexual male subjects were assessed for depression, anxiety, and associated symptoms using various self-report and interview measures, as well as for CGN (using the Freund Feminine Gender Identity scale, FGI). For comparison purposes only, we also evaluated the subjects for the DSM-III diagnosis of Ego-Dystonic Homosexuality. Highly gender nonconforming men (high FGI scores) were more likely to have current symptoms of anxiety and depression by self-report, and to have had a lifetime history of depression by clinical interview. This association was more often due to FGI items dealing with childhood than adulthood. When the FGI was broken into subscales by a prior factor analysis, stepwise regression suggested that the subscale measuring core gender identity nonconformity (so-called "gender dysphoria") was more reliably associated with depression and anxiety than were the factors measuring nonconformity in the areas of masculine and feminine gender roles, or genitoerotic (sexual) roles. This subscale was also the only FGI measure correlating with Ego-Dystonic Homosexuality. AIDS (CDC stage and HIV serostatus) and age did not account for these findings. We conclude that the often-reported higher levels of depression, anxiety, and associated symptoms among homosexual and bisexual men in AIDS studies are more common in the subgroup of such men who are gender dysphoric. Theoretical and clinical implications of these data are discussed.

Self-identified homosexual (n = 30), bisexual (n = 29), and heterosexual (n = 31) men were compared on measures of gender-typical behavior, sex role, ego strength, and lipid levels. Homosexual men differed significantly from the heterosexual men on the gender-typical behavior and feminine sex-role measure (both in adulthood and in childhood), and several trends and significant differences were found on the biochemical measures of lipid levels (especially when 7 obese men were removed from the analyses). As a rule, the bisexual men were different from the heterosexual men on the above measures, but were indistinguishable from the homosexual men. Bisexuals differed from both of the other two groups, however, by scoring lower on the ego strength scale and by reporting themselves to be more often troubled, lonely, and depressed. We caution that the lipid analyses were made on single blood samples and require an extended replication; however, we report the data because of their possible theoretical interest and because they replicate work of 20 years ago.

OBJECTIVE: Previous studies examining the relationship between substance use and unsafe sexual behavior have found only weak correlations. These studies assume these attitudes and behaviors are normally distributed. We speculated that specific sub-groups might be associated with increased risk of substance-related unsafe sex. METHOD: Subjects were HIV+ (N = 149) and HIV- (n = 53) men participating in a cohort study. We developed a questionnaire which measures attitudes toward safer sexual practices, drug, and alcohol use (positive scores reflect heavy alcohol/drug use and unsafe sexual behavior, a negative score reflects little substance use and safer sex). Scores were normalized, and cluster analysis was performed. RESULTS: Four groups were identified, representing Low (Group 1), Moderate (Group 2) and High (Groups 3 & 4) risk taking behavior. The two high risk-taking groups (3 & 4) were elevated on alcohol and unsafe sex scores but diverged on attitudes towards other drugs. TABULAR DATA, SEE ABSTRACT VOLUME. CONCLUSION: Distinct subgroups reflecting attitudes toward sexual behavior, alcohol, and drugs appear to exist. In the moderate subgroup sexual behavior may be most highly influenced by alcohol/drug intake. Selective intervention may be needed to successfully intervene in moderate and high risk-taking populations.

Previous research has suggested increased psychopathology in prenatally diethylstilbestrol (DES)-exposed persons. The current study compares the psychiatric histories and social functioning of 27 men with a history of high-dose prenatal DES exposure and their unexposed brothers. We expected DES subjects to show greater lifetime psychopathology and poorer social functioning than controls. Both groups showed high rates of lifetime depression, lifetime alcoholism, and current psychiatric symptoms in excess of community norms. The only diagnosis on which DES subjects exceeded their unexposed brothers was Major Depressive Disorder (MDD). DES-exposed men had almost twice the prevalence of at least one episode of MDD and had significantly more recurrent episodes. The relatively small number of subjects with concomitant lack of statistical power may have contributed to the difficulty obtaining significant effects.

Many researchers interested in sexual orientation can be separated into two camps: The "lumpers," who try to reduce sexual classifications to as small a number of categories as possible, and the "splitters," who try to show differences among groups and individuals that make classification schemes increasingly difficult and/or intricate. We report factor analyses of the Klein Grid (a questionnaire with 21 sexual orientation items) to see how many factors emerge in two samples of strikingly different origins. In both samples, the first factor to emerge loaded substantially on all of the Klein Grid's 21 items. This factor accounted for a majority of the variance. In both samples, a second, correlated factor emerged which indexed a separation between most of the items and those having to do with social and/or emotional preferences. In both samples, a third correlated factor also emerged, but this factor differed between the two populations: one refined the social/emotional distinction and the other distinguished ideal behavior from past and current behavior. We conclude on the basis of our analysis that both the lumpers and the splitters are correct.

In a prospective study to determine the incidence of clinical dementia in patients with AIDS and ARC, 29 men and 3 women, 19 with ARC and 13 with AIDS, were examined neurologically and neuropsychologically every 6 months for 2 years during a placebo-controlled zidovudine (AZT) licensing trial. Most received two MRI brain scans. Although no patient was clinically demented at baseline, 9 (28%) developed dementia during the 2 years. Progression to dementia was associated with neuropsychological deterioration and with worsening on MRI during a preceding 6-month period, but not with baseline treatment group assignment. The results suggest that patients at CDC Stage IV who do not receive antiretroviral treatment earlier in their illness may develop clinical dementia at an annual rate of about 14%.

OBJECTIVE: To examine the lifetime prevalence and clinical correlates of suicidal ideation and attempts in HIV+ and HIV- homosexual men and community controls. METHODS: HIV+ and HIV- ambulatory homosexual men (N = 105) in a longitudinal cohort study were examined for suicidality using the Diagnostic Interview Schedule. Community controls were sociodemographically matched with heterosexual men obtained from the Epidemiologic Catchment Area (ECA) study (N = 294). RESULTS: Lifetime proportion of suicide ideas or attempts were: TABULAR DATA, SEE ABSTRACT VOLUME. Lifetime presence of a major psychiatric disorder (depression, substance abuse) was elevated in high risk HIV+ and HIV- men compared to controls and contributed strongly to their higher rates of suicidal ideas/attempts. CONCLUSIONS: HIV+ men, and men at high risk for HIV infection, may be at increased risk for suicide because of psychiatric syndromes rather than HIV+ alone.

OBJECTIVE: AIDS is thought to increase risk of suicide, but risks and correlates of suicidal behaviors in earlier stages of HIV infection are not well studied. We examined relationships among suicidality, coping, social support, and neurocognitive impairment in men at various stages of illness. METHODS: Homosexual men in CDC IV (N = 22) and CDC II/III (N = 48) participating in a longitudinal cohort study were evaluated with the Lazarus Ways of Coping Checklist, Sarason Social Support Questionnaire, and an extended Halstead-Reitan neuropsychological (NP) battery. Lifetime suicidality (defined as 2 weeks or more duration of preoccupation with death, or wanting to die, or suicidal ideation or attempt) was determined by the Diagnostic Interview Schedule (DIS). Coping, social support, and NP impairment were compared in men with and without lifetime suicidality. RESULTS: Compared to men without suicidality, subjects with lifetime episodes of suicidality had (1) coping strategies using more avoidance (p less than .02) and less seeking social support (p less than .05), and (2) tended to have a smaller size of social network. NP impairment at baseline was not associated with lifetime suicidality. On one year followup (N = 42), NP deterioration tended to relate to increased coping by use of self blame, but was not associated with annual incidence of suicidality. CONCLUSIONS: Avoidant coping and isolation are associated with increased suicidal behaviors. NP decline may increase maladaptive coping. Targeted training in coping strategies might help reduce risk of suicide in the context of NP impairment.

OBJECTIVES: To encourage enrollment of research volunteers. To maintain satisfactory levels of participation of volunteers. To provide researchers with guidelines for demonstrating a high level of concern for the individual's well being of volunteers and to provide a formal mechanism for volunteers' input to researchers with regard to their participation as study subjects. METHODS: A four-month Participant's Advisory Board pilot was designed and consisted of study subjects from an existing HIV longitudinal study. Individuals were selected by research staff recommendations. Participants reflected the full range of HIV disease CDC status and included controls. This pilot Board identified multiple areas of concern with regard to themselves as study subjects. RESULTS: A formal report was submitted to the HIV Center Director and Council of Investigators with numerous recommendations which are being considered or implemented such as: 1) development of a Participant's Research Handbook; 2) a videotape for volunteers which explains the scope of the research and the procedures they will undergo. Investigators have utilized recommendations in developing the Center's Policies and Procedures Manual. CONCLUSION: Researchers found the pilot Board as representing them and empowering them as study subjects whose concerns are heard and dealt with in an organized fashion. The establishment of an ongoing Board for the HIV Center has been approved. The Board's current functioning reflects the Center's high priority for humane and considerate care of its volunteers.

OBJECTIVE: To determine the association between isolation of HIV from CSF and the development of AIDS-related CNS disease. METHODS: Seventy homosexual men without clinical AIDS-dementia (49 HIV+, 36 CDC II/III and 13 CDC IV; and 21 HIV-) were evaluated for CNS disease with an extensive battery of neuropsychological tests (NPT) and had 1 ml of CSF cultured in PHA-stimulated peripheral blood mononuclear cells. RESULTS: Of 27 HIV+ pts with only 1 evaluation, 52% had HIV cultured from CSF (3/8 CDC IV and 11/19 CDC II/III) vs 0/12 controls. Twenty-two HIV+ pts had 2 evaluations 1 yr apart with at least 1 of 2 CSF HIV cult pos in 73% (2/5 CDC IV and 14/17 CDC II/III) vs 0/9 controls. Of 13 pts with moderate impairment on NPT at study entry, 7 (58%) had CSF HIV cult pos. However, of 3 pts with 2 CSFs HIV cult pos, 1 had impairment at entry and 2 had worsening NPT on repeat testing. HIV isolation from CSF correlated with a CSF pleocytosis (wbc/mm3 mean +/- S.D.: pos: 13.6+/-13.7 vs neg: 5.4+/-7.8; p=.003). No CSF specimen was HIV p24 antigen pos. Peripheral blood T4 counts in HIV+ pts did not correlate with CSF cult pos for HIV (mean 545/mm3 in cult pos vs 480/mm3 in cult neg). CONCLUSION: Most infected pts have HIV in their CSF and the ability to culture HIV is related to the number of CSF mononuclear cells. Also, although repeated isolation of HIV from CSF may be associated with development of CNS disease, pts frequently have virus present in the CSF without dementia, and some pts with HIV-related CNS disease may not have HIV cultured from their CSF.

OBJECTIVE: The goals of this study are to demonstrate the prevalence of clinically observable neurologic findings in a selected population and to examine their correlation with immune system function. METHODS: A detailed neurologic examination was performed on a cohort of 95 gay men being followed prospectively with neuropsychological studies. 67 are HIV+ and 28 were seronegative. Subjects were categorized into three groups: A) HIV+, less than 200 T4 cells, B) HIV+, greater than 200 T4 cells, C) HIV negative. There are no significant differences in age or education between the three groups. Chi square analysis was employed to compare findings among the subjects. RESULTS: On initial examination, several symptoms distinguished the HIV+ and seronegative subjects. These abnormalities were reported in the areas of employment, concentration, reading, memory, gait, involuntary movement, mood, social withdrawal, emotional lability, alertness, and numbness. These symptoms found in only 0-21% of the seronegative group but were present in 18.1-93% of the HIV+ subjects. Additionally, 10-25.3% demonstrated altered affect, lower extremity weakness, or impaired gait on examination. Overall, 32.4%-38.4% of HIV+ subjects demonstrated clinical evidence of neurocognitive impairment. CONCLUSION: Our results demonstrate that neurologic signs and symptoms are common in HIV+ subjects with greater than 200 T4 cells. The importance of these findings will be clarified by longitudinal studies. A review of follow-up examination results will also be presented.

OBJECTIVE: To examine the prevalence of psychiatric disorder in various stages of HIV infection. We recently reported elevated lifetime prevalences of psychiatric disorder in HIV illness, and noted that onset of selected psychiatric disorders often preceded the AIDS pandemic, suggesting that some individuals might be at increased vulnerability for later disorder. Here we report on a larger cohort and include preliminary one-year follow up. METHODS: Subjects were 135 homosexual (high risk) and 22 heterosexual (low risk) men who were examined with the Diagnostic Interview Schedule (DIS). RESULTS: The table gives the most prevalent lifetime diagnoses. TABULAR DATA, SEE ABSTRACT VOLUME. Six month rates did not differ across high risk groups but exceeded low risk rates. Alcohol use disorders and over 20% of affective disorders preceded documentation of serostatus. One year follow up (N = 26 to date) incidence of major depression was 8%; gen. anxiety 23%; and alcohol disorder 4%. CONCLUSION: Careful assessment of high risk men, regardless of serostatus, may be warranted for selected psychiatric disorders.

OBJECTIVE: Neuropsychological (NP) deficit has been reported in 30% to 87% of patients with AIDS and in 5% to 44% of persons in CDC II and III. However, few longitudinal NP studies are available. We report on patients in CDC classes II, III and IV re-examined after 6 to 12 months as a part of a longitudinal study. METHODS: Gay men in the following groups were re-examined at 6 to 12 months: CDC IV - N=6; CDC II/III - N=18; HIV- controls - N=5. NP testing included extended Halstead-Reitan battery. Ratings of change were performed by clinicians blinded to group status. MRI brain scans, using a 1.5T GE Signa system were also obtained. RESULTS: At followup 3 of 6 AIDS patients (50%), 4 of 18 CDC II/III (22%), and 0 of 5 controls (0%) were rated as worse. Worsening was not associated with initial neuropsychological rating, nor interim drug use/mood disorder or CD4+ count. Computer assisted analysis of baseline MRI indicated that those who subsequently worsened on NP had increased cortical fluid and subcortical hyperintensities. CONCLUSION: This limited series suggests that neuropsychological decline can occur in AIDS patients even after 6 months and in some CDC II/III persons after 12 months. It is possible that subtle structural brain changes may presage neuropsychological decline, as indicated by computer assisted analysis of MR images.

OBJECTIVE: To understand the relationship between depression and impaired neuropsychological (NP) performance in HIV illness. METHODS: Subjects were ambulatory homosexual men classed as CDC IV (N = 37) and CDC II-III (N = 58); and HIV negative homosexual (N = 34) and heterosexual (N = 27) controls, matched for age and education. Measures were a comprehensive NP battery and various measures of mood, including the Profile of Mood States (POMS). Recursive partitioning was used to classify subjects according to a decision tree approach using quantitative measures of cognitive impairment (NP) and depression (D). RESULTS: On the basis of optimally chosen yes-no questions regarding NP and D covariates, a test of speeded information processing (PASAT) most rigorously discriminated D from NP impairment. The decision tree yielded three subgroups, in which NP+ indicated abnormal PASAT and D+ was depressed based on POMS-D greater than 15: In Group 1 40/47 CDC IV were NP+/D-; In Group 2 52/61 controls were NP-/D-; and Group 3 consisted of CDC II-III men: 44/58 NP-/D-, 11/58 NP-/D+, 3/58 NP+/D-. Measures of verbal and visual learning and forgetting displayed less discrimination. CONCLUSION: Recursive techniques suggest depressed mood can be separated from HP impairment in HIV. Application of advanced statistical approaches to classification hold promise for further exploration of cognitive function and in mood disorder in HIV illness.

OBJECTIVE: Since depression can affect neuropsychological (NP) functioning, we assessed whether observed NP deficits in men with HIV illness might be due to depressed mood. METHODS: Subjects were 72 homosexual men (CDC IV, N=9; CDC II/III, N=41; HIV-controls, N=22) participating in a longitudinal study of NP functioning in HIV. Measures were an expanded Halstead-Reitan Battery (HRB) and four indices of severity of depressed mood: Beck Depression Inventory (BDI); Hamilton Rating Scale for Depression (HRSD); the Profile of Mood States (POMS) and Symptom Checklist -90R (SCL-90R) depression scales. RESULTS: Significant group differences were obtained in prevalence of impairment on three of the eight ability areas covered by the expanded HRB (abstraction, learning, motor skills), principally due to deficits in CDC IV men. Of 36 correlations computed between the eight NP ability areas and the four measures of depression there was only one significant (p less than .05) result, which correlated greater motor impairment and higher HRSD scores. Subjects (N=12) with a lifetime history of major depression by DSM-III criteria did not show significantly greater NP impairment. CONCLUSIONS: Results from this series suggest that NP impairment in ambulatory HIV infected men is not accounted for by depressed mood.

OBJECTIVE: To study the association between use of cocaine, marijuana, alcohol and HIV infection in homosexual men. METHODS: Drug, alcohol, and medical histories were compared in 50 HIV infected (HIV+) homosexual men and 44 HIV uninfected (HIV-) homosexual men, all without a history of intravenous drug use. Sexual histories of 11 HIV+ cases and 11 HIV- controls with were compared for the 7-year period (1978-1984) during which infection occurred and before they were told their HIV status. RESULTS: Odds ratios and multiple logistic regression analysis revealed a significant association between number of years of cocaine use and HIV seropositivity (p =.013). Odds of being seropositive increased with more prolonged exposure to cocaine (OR=3.7). HIV was weakly associated with levels of marijuana use, but not with levels of alcohol consumption. Analysis of the sexual histories of the subsample suggested an association between high-risk sex and levels of cocaine usage. TABULAR DATA, SEE ABSTRACT VOLUME. CONCLUSIONS: Cocaine, in contrast to alcohol and marijuana, may contribute to risk of HIV infection by promoting risky sexual behavior.

OBJECTIVE: To determine whether childhood gender nonconformity (CGN, an important variable in gender identity research) is of value in AIDS research. The central hypothesis is that homosexual men with high CGN become disproportionately seropositive for HIV-1, largely due to high levels of receptive anal intercourse (figure above). TABULAR DATA, SEE ABSTRACT VOLUME. METHODS: N = 72 homosexual men have been recruited (23 seronegative and 49 seropositive), and completed the Clarke Feminine Gender Identity Scale (FGI). A one-way ANOVA correlated seropositivity with FGI scores. Additional analyses used scoring variations. RESULTS: FGI scores differ significantly by seropositivity status (figure below). (F = 6.65, df = 71, p less than 0.012; Fisher PLSD = 2.41, Scheffe F-test = 6.65, p less than 0.05). All except one of the men with the highest FGI scores were seropositive. CONCLUSIONS: The association between FGI and seropositivity is so strong at the high end of the FGI distribution that it might not arise merely by way of the hypothesized intermediate variable. Personality variables as well as behavioral ones may be required. Additional analyses will be reported at the meeting that suggest that FGI may be a good surrogate for receptive anal intercourse experience, and might to some extent mitigate the underreporting of such experience in direct questions. Implications for medicine and sexology will be discussed.

jweinrich@ucsd.edu

Telephone: (619) 543-5025

Copyright ©1996-7 James D. Weinrich.
Version of July 27, 1997.