This was a cross-sectional study in which we conducted a single visit that included a face-to-face interview, anthropometric measurements and laboratory investigations. In addition, previous clinical information was extracted from the participants’ files. In this study we determined the period prevalence of excessive weight gain in ALHIV after initiating DTG within the last 5 years.

The study was conducted at the Kampala Capital City Authority (KCCA) health centres in Kampala, the capital city of Uganda between February and May 2022. There are six health centres that are public health facilities supported by the Infectious Diseases Institute through the U.S President’s Emergency Plan for AIDS Relief to provide comprehensive HIV prevention, care and treatment services to PLHIV, including adolescents in the Kampala metropolitan area. Approximately, 1,300 adolescents are enrolled in care at these facilities and have a designated day for the adolescent clinic where they are provided with a range of services including; comprehensive HIV prevention, care and treatment, sexually transmitted infections management, family planning and counselling. Clinic visits are scheduled every 3–6 months, and on these visits, anthropometric measurements are taken including, weight and height. Viral load measurements are done every six months from the National Public Health Laboratory as per the Ministry of Health (MOH) guidelines and CD4 cell count measurement is no longer recommended. Adherence to ART at each visit is documented. Following the release of the consolidated guidelines for the prevention and treatment of HIV and AIDS in Uganda by the MOH in 2020, there is an on-going nationwide DTG rollout in all health facilities among PLHIV including adolescents with recommendation to initiate eligible treatment naïve patients to DTG-based ART as well as switch treatment experienced patients with virologic suppression (defined as < 1,000 copies/ml within the last 6 months) from EFV or protease inhibitor (PI)-based regimens to DTG. Abacavir, Zidovudine, Lamivudine, and TDF are the recommended companion drugs to DTG in our setting [27].

We included 165 ALHIV aged 10–19 years on DTG-based ART for at least one year attending the ART clinic in KCCA health centres. Adolescents 18 years or older and emancipated minors provided written informed consent. Written parental consent and assent was obtained for those aged 10–17 years who were not emancipated minors. Adolescents with missing baseline height and weight at DTG initiation, unknown DTG start date, those on antipsychotic medications, steroids or with underlying chronic medical conditions such as diabetes mellitus, renal, thyroid and cardiac disease and those that were pregnant while on DTG-based ART were excluded. In addition, ALHIV who declined to participate, were unreachable when contacted or had date of birth discrepancies were excluded from the study.

The sample size of 165 was estimated using the Modified Kish Leslie formula for cross-sectional studies, based on the prevalence of overweight in youth living with perinatal HIV infection on antiretroviral treatment in Thailand (11%) [28] and 10% non-response (missing information).

Participants were recruited from three public health facilities that routinely collect both the weight and height anthropometric measurements during clinic visits. A list of ALHIV aged 10–19 years who were on DTG for at least one year was obtained from each of these health facilities. We consecutively reviewed and screened clinic files of ALHIV to identify those eligible for inclusion in the study. Eligible participants and/or parents/caretakers where contacted via phone and invited to participate in the study at a convenient date. Participants were encouraged to fast (have last taken a meal, food or drink except for water at least 8 h prior) on the day of the study visit. Prior to enrolment, the research assistants provided details about the study to the participants and their parents/caretakers and eligibility was reassessed. Written informed consent, parental consent and assent were obtained accordingly. A structured questionnaire was used to collect socio-demographic and clinical data. We obtained information about pre-existing medical conditions, history of side effects after DTG start, smoking, alcohol intake and other substance use, concomitant medications, hormonal contraceptive use, family history of diabetes mellitus and hypertension and self report of involvement in physical activities that included exercise and sports such as running, aerobics, skipping ropes, football, etc. with in the past one month. In addition, we extracted data from the participant’s clinic file that included, date of ART and DTG-based regimen start, previous opportunistic infections, ART regimens, weight and height measurements and viral loads. We used viral load as a surrogate marker for medication adherence.

During the enrolment, anthropometric measurements including weight (kg) and height (cm) were taken as per WHO guidelines [29]. Weight was measured to the nearest 0.1 kg and height to the nearest millimetre with participants wearing light clothing and no shoes using a well-calibrated scale and stadiometer (Seca, Germany). BMI was defined as the weight in kilograms divided by the square of the height in meters (kg/m2) and classified according to WHO BMI-for-age z scores (5–19 years) as underweight (z-score <-2SD), normal weight (z-score ≥-2 and ≤ + 1SD), overweight (z-score > + 1SD (equivalent to BMI 25 kg/m2 at 19 years) and obesity (z-score > + 2SD (equivalent to BMI ≥ 30 kg/m2 at 19 years) [30]. A single blood pressure measurement was undertaken for each participant at rest using a calibrated OMRON M2 automatic upper arm blood pressure monitor and was graded as normal blood pressure < 120/80 mmHg and elevated blood pressure ≥ 120/80 mmHg for adolescents 13 years and older and normal if < 90th percentile and elevated if ≥ 90th percentile based on age, sex and height for adolescents 10–12 years [31].

Laboratory investigations: Fasting blood glucose (FBG) was measured using a calibrated On Call Plus blood glucose meter. FBG < 5.6 mmol/L was normal, pre-diabetic 5.6–6.9 mmol/L and high/diabetic if ≥ 7.0 or random blood glucose (RBG) > 11.1 mmol/L [32, 33]. Lipid profile was done for all participants and glycated haemoglobin (HbA1c) was done only for those with FBG ≥ 5.6 mmol/L. For HbA1c levels, a venous blood sample of 3 mL was drawn into an ethylenediamine tetraacetic acid blood collection tube from the study participant. The blood was quantitatively tested for HbA1c levels using the Roche COBAS Integra 400 plus automated analyser. HbA1c was categorised normal ≤ 5.6%, pre-diabetes 5.7–6.4% and diabetes ≥ 6.5% [33]. Similarly, for lipid profile, a venous blood sample of 4 mL was drawn into a plain (Red top) blood collection tube from each study participant. Blood was centrifuged at 3000 revolutions per minute (rpm) for ten minutes, and serum was collected in a Sarstedt 2 mL cryovial and quantitatively tested for lipid profile tests using the Roche COBAS Integra 400 plus automated analyser and categorised as follows; Triglycerides - acceptable < 90 mg/dl, borderline 90–129 mg/dl, high ≥ 130 mg/dl; Low- density lipoprotein cholesterol acceptable < 110 mg/dl, borderline 110–129 mg/dl, high ≥ 130 mg/dl; High-density lipoprotein cholesterol acceptable > 45 mg/dl, borderline 40–45 mg/dl, Low < 40 mg/dl; Total cholesterol acceptable < 170 mg/dl, borderline 170–199 mg/dl, high ≥ 200 mg/dl [34].

Data was analysed using STATA 14.0 (Stata Corp., College Station, TX, USA). Continuous variables were summarized with means and standard deviations for normally distributed data and with medians and interquartile ranges for data that is not normally distributed. Categorical variables were described using frequencies and, percentages.BMI and BMI-for-age z scores were calculated using WHO Anthro software [35]. . The primary outcome variable excessive weight gain was defined as an ALHIV who becomes overweight or obese at least one year after DTG initiation. Overweight or obesity were treated as binary outcomes. Chi-square or Fisher’s exact tests were used to compare participant characteristics between BMI categories (normal weight vs. overweight and obesity). Logistic regression was used to identify the factors associated with excessive weight gainafter DTG start. Bivariate analysis assessed the crude odds ratios of exposure variables. Variables included in the analysis were sex, ART experience, pre-switch anchor drug, current viral load, and duration of DTG treatment, lipid profile, blood pressure and glucose. Stepwise multiple logistic regression models were employed to control for confounding and interaction effects. Significant variables at the bivariate level (p ≤ 0.20) and those with clinical significance from literature were included in the multivariate model. Variables with a p-value of < 0.05 (two-sided) were considered statistically significant. Missing data were assessed and imputed using standard techniques.