Risk Factors for Weight Gain Following ART Initiation

Study authors assessed various demographic, HIV-specific, and antiretroviral-therapy-specific risk factors associated with weight gain following initiation of antiretroviral therapy in patients with HIV.

Demographics, HIV-specific factors, and antiretroviral therapy (ART)-specific factors are associated with weight increase after ART initiation in people living with HIV (PLWH), according to study research published in Clinical Infectious Diseases.

In the current pooled analysis, study authors explored demographic, HIV-specific, and ART-specific factors associated with weight gain in several randomized comparative clinical trials of ART initiation. Prevalence of excess weight and obesity has increased in PLWH who have initiated certain regimens or drug classes of ART, the researchers noted. Potential weight gaining mechanisms include a return-to-health phenomenon, accelerated catabolism, and the resolution of opportunistic infections and gastrointestinal dysfunction that results in better appetite and nutrient absorption.

In all, 8 randomized Phase 3 clinical trials including treatment-naïve patients with HIV who initiated ART from 2003 to 2015 were pooled and analyzed, totaling 5680 patients and 10,000 person-years of follow-up.

At ART initiation, the median BMI was 24.8 kg/m2. A total of 924 (16.3%) patients were obese (BMI ≥30 kg/m2), 1785 (31.4%) patients were overweight (BMI 25-29.9 kg/m2), and 2829 (52.4%) of patients were of normal weight (18.5-24.9 kg/m2) or underweight (<18.5 kg/m2). The 96-week median weight gain was 2.0 kg (interquartile range [IQR], -0.9-5.9); the greatest rate of weight gain occurred during the first 48 weeks. At 96 weeks, a 3%, 5%, and 10% increase from baseline weight was observed in 48.6%, 36.6%, and 17.3% of patients, respectively, and 30.2% of patients lost weight.

Pooled analysis suggest that demographics, HIV specific-factors, and ART-specific factors are associated with weight increase, including:

  • Black race: 95% CI, 0.87-1.11; P <.001
  • BMI: overweight vs normal; 95% CI, -0.36 to -0.13; P <.001
  • CD4 cell count: less than 200 vs 200 cells/µL or more; 95% CI, 2.81-3.13; P <.001
  • HIV RNA: greater than 100,000 vs 100,000 or less copies/mL; 95% CI, 0.84-1.08; P <.001
  • HIV/AIDS: symptomatic vs asymptomatic; 95% CI, 0.36-0.65; P <.001
  • Injection drug use: use vs without use; 95% CI, 0.97-1.85; P <.001

Additional factors included higher BMI (obese vs normal; 95% CI, 0.06-0.36; P =.005), being a woman (95% CI, 0.07-0.4; P =.006), and age (<50 vs ≥50 years; 95% CI, 0.07-0.37; P =.004). More weight gain was observed with specific ART regimens or the use of newer ART regimens. Integrase strand transfer inhibitor use was associated with more weight gain when compared with protease inhibitors or nonnucleoside reverse transcriptase inhibitors (NNRTIs). Among NNRTIs, rilpivirine was associated with more weight gain than efavirenz. The nucleoside/nucleotide reverse transcriptase inhibitor, tenofovir alafenamide, was associated with more weight gain when compared with tenofovir disoproxil fumarate, abacavir, and zidovudine. During 48 weeks, 12.8% of participants gained at least 10% of their body weight. These patients were more likely to have a lower baseline weight or BMI (P <.001), be a woman(P <.001), be Black (P =.003), have higher baseline HIV-1 RNA (P <.001), and have a lower baseline CD4 cell count (P <.001).

Study authors concluded that, “modern ART regimens with improved tolerability and potency may lead to weight gain in some PLWH, necessitating increased clinical attention to the maintenance of healthy body weight, lifestyle modification, and exercise at ART initiation.”

Disclosure: Several study authors declared affiliations with the pharmaceutical industry. Please see the original reference for a full list of authors’ disclosures.


Sax PE, Erlandson KM, Lake JE, et al. Weight gain following initiation of antiretroviral therapy: risk factors in randomized comparative clinical trials. Clin Infect Dis. 2020;71(6):1380-1389. doi:10.1093/cid/ciz999.