Abstract Body

Background:

The study evaluated weight (wt) change and shift in BMI class in PWH on antiretroviral therapy (ART) over 3 years (3y) vs PWoH matched on baseline (BL) characteristics.

Methods:

Retrospective study using Trio Health HIV Network EMR data from federally qualified health centers in US. Eligibility: ≥18 yrs, in care between 1/1/2015-8/15/2023, with BL and 3y wt measures (all); PWH: treatment-experienced suppressed at BL and 3y or suppressed on 1st ART with BL≥ 6mo since suppression and ≥ 12mo since ART start. BL characteristics were compared (Χ-square, t-test). Propensity scores (PS) and odds of shifting up BMI class were calculated (logistic regression); groups were matched on site, BL year, gender, age, race, and PS to compare 3y wt and BMI changes, with additional adjustment for imbalanced covariates.

Results:

Of 68856 qualified individuals, 11888 (17%) were PWH suppressed, 902 (8%) suppressed on 1st ART. PWH and PWoH differed respectively in key BL characteristics (all listed p<.05): age (median 50 vs 54 years), gender (male 77 vs 40%), race (42 vs 48% Black), eGFR <60 mL/min/1.73m2 (11 vs 6%), obesity (BMI> 30 kg/m2; 29 vs 47%), hyperlipidemia (14 vs 4%), hypertension (24 vs 46%), diabetes (5 vs 2%), neuropsychiatric disorders (26 vs 14%), cardiovascular disease (28 vs 49%), sexually transmitted diseases (10 vs 1%), smoking (11 vs 2%), or substance use (9 vs 6%). PWoH had statistically higher mean [median] BL BMI (all: 30.8 [29.4] vs 28 [26.9], matched: 29.6 [27.5] vs 28.6 [26.9]). In unmatched analyses, PWH gained 0.6 kg (95% CI 0.3-1.0) more at 3y vs PWoH after adjusting for BL differences. After matching, PWH and PWoH did not differ in 3y wt change (0.1 kg CI -0.5-0.7), Figure 1. Despite small differences in wt change, unmatched PWH were 1.3 (CI 1.1-1.5) times more likely to shift up a BMI class (17% vs 10% PWoH); smaller, but similar differences were observed in matched cohorts (17% vs 13%; 1.3 [CI 1.1-1.7]).

Conclusions:

This is the largest study that matched and compared wt change in suppressed PWH vs PWoH, accounting for return to health, population differences, and geography. After adjusting for differences in BL characteristics, small (unmatched) and no difference (matched) in 3y mean wt change were observed in PWH vs PWoH. A greater proportion of PWH vs PWoH shifted up a BMI class after 3 y, although results may be influenced by higher prevalence of obesity at BL among PWoH. Further examination of drivers of outlier wt gain and the role of ARTs is ongoing.