A Latent Class Analysis of Chronic Health Conditions Among HIV Positive Transgender Women of Color

Background. Research on the health of transgender people has focused on risk for and the health consequences of HIV and other sexually transmitted infections or on the cardiovascular effects of long-term hormone use. Consequently, little is known about the broader medical profiles of transgender people requiring primary medical care. The current study sought to determine if medical profiles (i.e., multimorbidity) of male to female transgender persons could be developed based on the experiential pattern of a range of chronic health conditions (CHCs). We also wanted to determine if higher need for health care based upon medical profile was associated with factors such as age and psychological distress well as factors more specific to transgender persons that could influence health such as hormone treatment and the injection of illicit fillers such as silicone for body contouring.

Methods. Data for this study were obtained from 223 HIV+ transgender women of color (TWOC), 18 years of age or older, at two Chicago sites participating in a nine-site demonstration project funded from 2012 to 2017 by HRSA as a Special Project of National Significance (SPNS). Self-reported, ACASI interview data collected at baseline and six months were used to assess twenty-four lifetime CHCs derived from the WHO-CIDI questionnaire. These became the binary indicators for the latent class (LC) models. Covariate LC predictors included AIDS status (ever/never), silicone injection (never/not currently/currently), hormone use (never/not currently/currently), severe psychological distress (no/yes), and homelessness (never/< 1 year/> 1 year). Mplus 8.1 was used to estimate the LC model with a 3-step procedure to first estimate the best-fitting LC structure followed by multivariable regression of LC on the predictor set.

Results. The best-fitting LC model (BLRT=272.6. p < .001) had 2-classes designated as low and moderate/high probability contingent on the probability of having multiple CHCs with 26% of participants classified in the moderate/high probability LC. Participants in the moderate/high probability LC had higher probabilities of: skin problems, arthritis, hepatitis (all kinds), COPD, seasonal allergies, and pneumonia. Participants in both groups had high probabilities of STIs other than HIV, asthma, and dental problems. Among covariate predictors age (OR=13.8, p < .001), ever having AIDS (OR=4.0, p < .05) and severe psychological distress (OR = 5.1, p < .05) were associated with moderate/high probability class membership.

Conclusions. A substantial subset (26%) of HIV+ TWOC seen in primary care settings are medically complex and have a range of CHCs beyond those explicitly associated with HIV disease. The results suggest medical multimorbidity is common, corresponding to a need for routine screening for multiple conditions especially respiratory conditions (asthma, COPD, pneumonia), dental problems, and STIs other than HIV. Focusing on HIV-related care or hormonal treatment and its attendant cardiovascular issues could result in sub-optimal care for a population that has been disengaged from regular medical care owing to stigma, affordability, lack of training for medical personnel on transgender health, etc. Where possible, social workers working with transgender persons should encourage engagement in medical care while advocating for comprehensive as well as culturally competent care.