Abstract
Introduction
Obesity is increasing across the US, with people with HIV experiencing greater risk of obesity-related adverse health outcomes including metabolic diseases. Weight gain has been shown during the widespread shutdowns during the COVID-19 pandemic. We examined weight trajectories before, during and after the COVID-19 pandemic among people with and without HIV.
Setting:
Participants in the Multicenter AIDS Cohort Study-Women’s Interagency HIV Study Combined Cohort Study (MWCCS)
Methods
Study time periods were: 1) Pre-pandemic period, May 15, 2018 through March 15, 2020; 2) Pandemic period, March 15, 2020 through September 30, 2021; and 3) Post-pandemic period, October 1, 2021 through September 30, 2024. A piecewise linear mixed effects regression model adjusted for baseline age was fitted with a random intercept for individual. Interaction terms examined differences by sociodemographic characteristics.
Result
Among 1,586 participants, 66.5% were living with HIV. From the pre- to during-pandemic period, there was a statistically significant 0.14 kg/m2/visit increase in BMI (95% CI:0.07. 0.22). There was a 0.3 kg/m2 reduction in mean BMI in the 36 months from pandemic [32 kg/m2 (SD: 8.6)] to post-pandemic [31.7 kg/m2 (SD: 8.5)] periods. Similar trajectories were noted among sociodemographically vulnerable subgroups.
Discussion
Contrary to our hypothesis, we observed a downward BMI trajectory back to baseline with the exception of those with residential instability in the post-COVID-19 pandemic period following a statistically significant weight gain during the pandemic. Understanding factors associated with decreasing BMI trajectories in the post-pandemic period is important in continuing to address the obesity epidemic in the U.S.
Keywords: HIV, BMI, COVID-19 Pandemic
INTRODUCTION
Obesity rates continue to increase across the US, including among people with HIV (PWH).[1–3] This has been attributed to a variety of factors, including an obesogenic environment (the systemic factors that contribute to weight gain), as well as possible direct effects of certain ART regimens.[4–6] Obesity is associated with numerous negative health sequelae, including increased rates of metabolic disease and mortality, with greater risk of adverse health outcomes occurring with weight gain among PWH compared with people without HIV.[7, 8]
Several studies have reported weight gain among US adults associated with the widespread shutdowns during the COVID-19 pandemic, including among PWH.[9–15] While most of these studies utilized self-reported data, some studies using electronic health data also documented small but significant weight increases during the pandemic.[12, 15, 16] Importantly, several studies showed higher rates of weight gain among vulnerable populations including those of lower socioeconomic status or minority race or ethnicity, thus widening health inequities among marginalized groups.[9, 15]
To our knowledge, there have been limited assessments of weight trajectories in the post COVID-19 pandemic timeframe compared with before or during the pandemic. One might hypothesize that the lifting of physical activity restrictions and reduced stress following the acute pandemic period would result in returns to pre-pandemic weight levels.[17] In addition, studies have shown that social stressors such as food or economic insecurity that occurred during the pandemic were associated with COVID-19 lockdown-associated weight gain.[18] Similarly, studies have shown associations between residential instability and obesity; major stressors such as the COVID-19 pandemic may have exacerbated this association.[19] Of note, studies have shown that weight gained in relatively short periods of time (such as holidays) is often not reversed thus resulting in increased body weight over the long term.[20, 21] Therefore, several authors have postulated that additional weight gained during the COVID-19 pandemic would likely persist, thus exacerbating the obesity epidemic and its associated negative health consequences.[13, 22]
In this study, we utilized data from the Multicenter AIDS Cohort Study-Women’s Interagency HIV Study Combined Cohort Study (MWCCS), a multi-center cohort study of men and women living with and without HIV, with the scientific aim of advancing clinical, behavioral and epidemiologic characteristics of HIV.[23] Our aim was to describe BMI trajectories before, during, and after the COVID-19 pandemic, stratified by key sociodemographic characteristics.
METHODS
Study Population
Participants enrolled in either the Multicenter AIDS Cohort Study (MACS) or Women’s Interagency Health Study (WIHS) that were carried over into the MACS/WIHS Combined Cohort Study (MWCCS) had study visits that were conducted approximately every 6 months and contributed sociodemographic, risk behavior, laboratory, physical exam, and clinical event data. The historic MACS study targeted enrollment of men who have sex with men from 4 cities across the United States between 1984 and 2017, ultimately achieving a 1:1 ratio of PWH and PWOH. The historic WIHS study enrolled women with and without HIV at a 3:1 ratio between 1994 and 2012 initially across 6 cities and finally from 9 cities across the U.S.[23] The cohorts were approved by relevant institutional review boards and all participants provided written informed consent.[23]
For this analysis, we divided our study into three time periods aligning with study visit windows and the COVID-19 pandemic: 1) Pre-pandemic period, May 15, 2018 through March 14, 2020, a 20-month period prior to the widespread pandemic-associated shutdowns in the US; 2) Pandemic period, March 15, 2020 through September 30, 2021; and 3) Post-pandemic period, October 1, 2021 through September 30, 2024. Although several facets of the pandemic continued well past Fall 2021, we selected this date as when most activity restrictions were lifted and schools re-opened and thus most individuals would be able to return to pre-pandemic physical/leisure activity levels.[24–26] To be eligible for this study population, MWCCS enrollees had to have at least two Body Mass Index (BMI) measurements in the pre-pandemic period, at least one BMI measurement in the pandemic period, and at least two BMI measurements in the post-pandemic period. For participants with more than two BMI’s during the pre-pandemic or post-pandemic periods, the analytic sample was restricted to the two BMI measurements taken closest to the pandemic. All participants had only one BMI during the pandemic period.
Covariates
Our outcome of interest was BMI, calculated as body weight (kilograms) divided by height (meters) squared and was categorized as “Underweight/Normal” (BMI <25.0) vs. “Overweight/Obese” (BMI ≥25.0).
Explanatory variables were selected a priori, based on covariates shown in the literature to be associated with weight gain or obesity. All covariates were assessed at study baseline (first pre-pandemic visit) and treated as time invariant. HIV serostatus at baseline was categorized as “positive” vs. “negative”; there were two participants that seroconverted during the study time period and were excluded from this analysis. Age was measured as a continuous value, and self-reported sex assigned at birth was dichotomized as “male” vs. “female”. Self-identified racial categories and ethnicity categories were combined and collapsed as “Black/African-American, not Hispanic”, “White/Caucasian, not Hispanic”, “Other, not Hispanic”, and “Hispanic”. Highest level of educational attainment was categorized as “High School or Less” vs. “Some College or More”. Employment status was dichotomized as “Not employed” vs. “Employed/Student/Retired/Disabled”. Annual individual/household income at baseline was dichotomized as “≤$30,000” vs “>$30,000”. Health insurance status was categorized as “Yes” vs. “No”. Enrollment sites were categorized by geographic region as follows: West (San Francisco and Los Angeles, CA), Northeast (Brooklyn and Bronx, NY) Mid-Atlantic (Washington D.C. and Baltimore, MD), South (Birmingham AL, Miami FL, Atlanta GA, Jackson MS, and Chapell Hill NC), and Midwest (Columbus OH, Pittsburg PA, and two sites in Chicago IL).
To explore the role of social stressors associated with pandemic-related changes in weight, social hardship variables were collected within the pandemic time-period. As a measure of food insecurity, participants indicated whether they were “worried that food would run out before getting money for more” as “Often”, “Sometimes”, or “Never” true; these were dichotomized as “Often/Sometimes” vs. “Never” true. Residential stability was measured as those who reported residing in their own house or apartment vs. living at another individual’s home/in a shelter/on the street. Economic insecurity was measured as having difficulty paying for basic needs such as food, shelter, and/or heat was dichotomized as “Somewhat/Very hard” vs. “Not very hard”.
Analytic Approach
Preliminary analyses were conducted to assess data quality, including missingness. Participant characteristics were summarized in terms of means or medians for continuous variables and proportions for categorical variables. Missingness was negligible, with only two items missing greater than 1% of responses: income (3%) and food insecurity (5.5%). A piecewise linear mixed effects regression model adjusted for baseline age was fitted with a random intercept for individual to examine BMI trajectories over time. Time was characterized by study visit at which BMI was collected, with the first pre-pandemic BMI considered visit one (baseline), the pandemic BMI considered visit three, and the last post-pandemic BMI considered visit five. A spline term was inserted to account for non-linear BMI trajectories, with a knot at the pandemic visit to allow for comparisons of trajectories between 1) the pre- and during-pandemic, as well as 2) the during- and post-pandemic time periods. Models were repeated, introducing individual covariates as a main effect and as interactions with time and the spline term, to describe the independent effects of sex, pre-pandemic BMI category, HIV serostatus, income, education, and employment. Participants missing income and employment were excluded from the related covariate models.
Given the specific and relatively stringent requirements of needing to complete two study visits in the pre- and post-pandemic periods to be included in the study population, we created a comparison population of MWCCS enrollees who completed at least one study visit in the pre-pandemic period and then one study visit in either the during or post-pandemic periods. The comparison population was compared with the study population on core sociodemographic characteristics using Chi-Square tests.
Statistical significance was defined as P < 0.05 and all analyses were performed using SAS 9.4 (SAS Institute, Cary, NC).
RESULT
Our study population included 1,586 participants, the majority (n=1,152, 72.6%) of whom were female (Table 1). Approximately one-third (n=613, 38.7%) were 50–59 years old, followed by 29% (n=457) aged ≥60 years and 22.9% (n=153) aged 40–49 years. The vast majority (n=1,287, 81.2%) were overweight/obese at baseline. Over two-thirds (n=1,054, 66.5%) were living with HIV, 56.4% (n=894) were employed, and 93.8% (n=1,487) had health insurance. Approximately half (n=807, 50.9%) had some college education or more, and 34.7% (n=550) reported an annual individual/household income of >$30,000. In terms of social stressors, approximately 15% (n=234) reported food insecurity, 12.4% (n=197) reported residential instability, and 34% (n=537) reported economic insecurity.
Table 1:
Characteristics of the Study Population Overall and Stratified by HIV Serostatus and the Comparison Population by Sociodemographic Characteristics, MWCCS, 2018 – 2024
| Analytic Population n=1,586 |
Comparison Population n=2,876 |
PWH n=1054 (66.5%) |
PWOH n=532 (33.5%) |
|
|---|---|---|---|---|
| Age | ||||
| <40 years | 153 (9.7%) | 278 (9.7%) | 84 (8%) | 69 (13%) |
| 40 - <50 years | 363 (22.9%) | 592 (20.6%) | 255 (24.2%) | 108 (20.3%) |
| 50 - <60 years | 613 (38.7%) | 1056 (36.7%) | 448 (42.5%) | 165 (31%) |
| ≥60 years | 456 (28.8%) | 950 (33.0%) | 267 (25.3%) | 190 (35.7%) |
| BMI Category | ||||
| Underweight/Normal | 299 (18.9%) | 645 (22.4%) | 193 (18.3%) | 106 (19.9%) |
| Overweight/Obese | 1,287 (81.2%) | 2,231 (77.6%) | 861 (81.7%) | 426 (80.1%) |
| HIV Serostatus | ||||
| PWH | 1,054 (66.5%) | 1,043 (36.3%) | 1,054 (100%) | |
| PWOH | 532 (33.5%) | 1,833 (63.7%) | 532 (100%) | |
| Sex at Birth | ||||
| Female | 1,152 (72.6%) | 1,725 (59.9%) | 826 (78.4%) | 326 (61.3%) |
| Male | 434 (27.4%) | 1,151 (40.0%) | 228 (21.6%) | 206 (38.7%) |
| Race and Ethnicity | ||||
| Black/African-American, not Hispanic | 992 (62.6%) | 1,556 (54.1%) | 703 (66.7%) | 289 (54.3%) |
| White/Caucasian, not Hispanic | 378 (23.8%) | 842 (29.3%) | 199 (18.9%) | 179 (33.7%) |
| Hispanic | 163 (10.3%) | 383 (13.3%) | 114 (10.8%) | 49 (9.2%) |
| Other, not Hispanic | 53 (3.3%) | 95 (3.3%) | 38 (3.6%) | 15 (2.8%) |
| Region | ||||
| West | 155 (9.8%) | 525 (18.3%) | 103 (9.8%) | 52 (9.8%) |
| Northeast | 371 (23.4%) | 512 (17.8%) | 261 (24.8%) | 110 (20.7%) |
| Mid-Atlantic | 286 (18.0%) | 504 (17.5%) | 180 (17.1%) | 106 (19.9%) |
| South | 388 (24.5%) | 611 (21.2%) | 284 (26.9%) | 104 (19.6%) |
| Midwest | 386 (24.3%) | 724 (25.2%) | 226 (21.4%) | 160 (30.1%) |
| Highest Educational Attainment | ||||
| High School or less | 779 (49.1%) | 1,296 (45.1%) | 559 (53%) | 220 (41.4%) |
| Some College or more | 807 (50.9%) | 1,580 (54.9%) | 495 (47%) | 312 (58.7%) |
| Employment | ||||
| Not employed | 686 (43.3%) | 1,093 (38.0%) | 507 (48.1%) | 179 (33.7%) |
| Employed/Student/Retired/Disability | 894 (56.4%) | 1,607 (55.9%) | 544 (51.6%) | 350 (65.8%) |
| Missing | 6 (0.4%) | 176 (6.1%) | 3 (0.3%) | 3 (0.6%) |
| Individual/Household Income | ||||
| ≤$30,000 | 989 (62.4%) | 1,593 (55.4%) | 705 (66.9%) | 284 (53.4%) |
| >$30,000 | 550 (34.7%) | 993 (34.5%) | 324 (30.7%) | 226 (42.5%) |
| Missing | 47 (3.0%) | 290 (10.1%) | 25 (2.4%) | 22 (4.1%) |
| Health Insurance Status | ||||
| Yes | 1,487 (93.8%) | 2,678 (93.1%) | 1000 (94.9%) | 487 (91.5%) |
| No | 93 (5.9%) | 180 (6.3%) | 51 (4.8%) | 42 (7.9%) |
| Missing | 6 (0.4%) | 18 (0.6%) | 3 (0.3%) | 3 (0.6%) |
| Worried food would run out before getting money for more | ||||
| Often/Sometimes true | 234 (14.8%) | 382 (13.3%) | 161 (15.3%) | 73 (13.7%) |
| Never true | 1265 (79.8%) | 2217 (77.1%) | 836 (79.3%) | 429 (80.6%) |
| Missing | 87 (5.5%) | 277 (9.6%) | 57 (5.4%) | 30 (5.6%) |
| Reside in own house/apartment | ||||
| Yes | 1381 (87.1%) | 2257 (78.5%) | 919 (87.2%) | 462 (86.8%) |
| No | 197 (12.4%) | 348 (12.1%) | 130 (12.3%) | 67 (12.6%) |
| Missing | 8 (0.5%) | 271 (9.4%) | 5 (0.5%) | 3 (0.6%) |
| Difficulty paying for basics needs (i.e., food/shelter/heat) | ||||
| Somewhat/Very hard | 537 (33.9%) | 879 (30.6%) | 373 (35.5%) | 163 (30.7%) |
| Not very hard | 1041 (65.6%) | 1722 (59.9%) | 675 (64%) | 366 (68.8%) |
| Missing | 8 (0.5%) | 275 (9.6%) | 5 (0.5%) | 3 (0.6%) |
| On ART (PWH only) | ||||
| Not on ART | 40 (2.2%) | 25 (2.4%) | ||
| On ART at baseline | 1716 (93.6%) | 983 (93.3%) | ||
| Started ART during study period | 77 (4.2%) | 46 (4.4%) | ||
| Months between first pre-pandemic and during-pandemic BMI measures (Median (IQR)) | 28 (25 – 30) | 28 (25 – 30) | 28 (26 – 30) |
PWH: People with HIV; PWOH: People without HIV; ART: Anti-Retroviral Therapy
Enrollment sites associated with each region: West (California), Northeast (New York), Mid-Atlantic (Washington, D.C., and Maryland), South (Alabama, Florida, Georgia, Mississippi, North Carolina), Midwest (Illinois, Ohio, Pennsylvania).
When stratified by HIV serostatus, 66.5% (n=1,054) were living with HIV (PWH) while 33.5% (532) were HIV seronegative (PWOH). Approximately 80% of both populations were overweight/obese at baseline. The two groups did differ on basic demographic characteristics, including 78% of PWH being female vs. 61.3% of PWOH (p<0.01), 66.7% of PWH vs. 54.3% of PWOH identifying as Black or African American, non-Hispanic (p<0.01), and 26.9% of PWH vs. 19.6% of PWOH being from the South. In terms of socioeconomic status, PWH were more likely to be unemployed (48.1% vs. 33.7%), have a lower household income (66.9% vs. 53.4%), and report having health insurance (100% vs. 91.5%). Reports of food insecurity, residential instability and economic insecurity were similar between the two groups. Over 97% of PWH were on ART, with 93% already being on ART at study baseline.
The comparison population included 2,876 participants who had similar age group and health insurance status distributions as the study populations (Table 1). However, they were otherwise statistically significantly different from the study population in several important characteristics. The comparison population was more likely to be underweight/normal BMI (36.3% vs. 33.5%), white/Caucasian (29.3% vs. 22.8%), some college or more education (54.9% vs. 50.9%), be employed (55.9% vs. 51.1%), and have an annual household income >$30,000 (34.5% vs. 31.8%). The comparison population was slightly more likely to come from the West (18.3% vs. 9.8%) while the study population was more likely to come from the Northeast (17.8% vs. 23.5%). The remaining regions were equally represented in the two populations. Reports of food insecurity, residential instability and economic insecurity were similar between the comparison and study populations.
In the study population, the mean BMI in the pre-pandemic period was 31.8 kg/m2 (SD: 8.2), and the mean BMI in the during-pandemic period was 32 kg/m2 (SD: 8.6) (Figure 1). There was a notable downward trajectory in BMI in the post-pandemic period that differs from what might have been predicted had the pre-pandemic to pandemic trajectory continued. From the pre-pandemic to the during-pandemic period, there was a statistically significant 0.14 kg/m2/visit increase in BMI (95% Confidence Interval (CI):0.07, 0.22) in the overall population (Table 2). In the interaction models, a similar increase was seen among females (0.212, 95% CI: 0.128, 0.295), baseline BMI category overweight/obese (0.093, 95% CI: 0.014, 0.171), PWH (0.160, 95% CI: 0.073, 0.247), Black participants (0.169, 95% CI: 0.079, 0.259), those living in the South (0.188, 95% CI: 0.044, 0.331), those with lower income (0.229, 95% CI: 0.138, 0.319), those with lower education (0.188, 95% CI: 0.864, 0.289), and participants not employed (0.266, 95% CI: 0.158, 0.374) (Figure 2). We also noted increased BMI among those with (0.359, 95% CI: 0.175, 0.543) and without (0.109, 95% CI: 0.029, 0.188) food insecurity, those with (0.128, 95% CI: 0.053, 0.205) and without (0.229, 95% CI: 0.027, 0.431) residential instability, and those with (0.255, 95% CI: 0.133, 0.377) economic insecurity. In contrast, no statistically significant increase in BMI was noted among males, people without HIV, white individuals, employed individuals, individuals reporting moderate-high income, or those without economic insecurity (Figure 2).
Figure 1: Trajectories of BMI through the Study Period, Compared to the Counterfactual Trajectory of BMI had the Trend Not Changed in the Post-Pandemic Period.
Dotted line: Projected BMI based on pre- and during-pandemic data; Solid line: Observed post-pandemic BMI data.
Table 2:
BMI Trajectories Before, During and Post Pandemic, Modified by Sociodemographic Characteristics, MWCCS, 2018–2024
| BMI Change (Pre-Pandemic through During Pandemic) | BMI Change (During Pandemic through Post-Pandemic) | |
|---|---|---|
| Overall | 0.140 (0.073, 0.216) | −0.210 (−0.284, −0.142) |
| BMI Category | ||
| Underweight/Normal | 0.367 (0.204, 0.531) | 0.006 (−0.157, 0.170) |
| Overweight/Obese | 0.093 (0.014, 0.171) | −0.264 (−0.343, −0.186) |
| HIV Serostatus | ||
| PWH | 0.160 (0.072, 0.247) | −0.216 (−0.304, −0.129) |
| PWOH | 0.114 (−0.009, 0.237) | −0.20. (−0.330, −0.084) |
| Sex at Birth | ||
| Female | 0.212 (0.128, 0.295) | −0.295 (−0.379, −0.212) |
| Male | −0.034 (−0.169, 0.102) | 0.005 (−0.131, 0.141) |
| Race | ||
| Black/African-American | 0.169 (0.079, 0.259) | −0.231 (−0.321, −0.140) |
| White/Caucasian | −0.027 (−0.173, 0.120) | −0.011 (−0.158, 0.135) |
| Region | ||
| South | 0.188 (0.044, 0.331) | −0.29 (−0.434, −0.146) |
| Mid-Atlantic/Midwest/ Northeast/West | 0.131 (0.049, 0.212) | −0.189 (−0.27, −0.107) |
| Highest Educational Attainment | ||
| High school or less | 0.188 (0.086, 0.289) | −0.267 (−0.368, −0.165) |
| Some college or more | 0.103 (0.003, 0.202) | −0.162 (−0.261, −0.062) |
| Employment | ||
| Not employed | 0.266 (0.158, 0.374) | −0.459 (−0.567, −0.351) |
| Employed/Student/Retired/Disability | 0.051 (−0.044, 0.146) | −0.025 (−0.120, 0.069) |
| Individual/Household Income | ||
| ≤$30,000 | 0.229 (0.138, 0.319) | −0.348 (−0.438, −0.257) |
| >$30,000 | −0.012 (−0.133, 0.110) | 0.010 (−0.111, 0.132) |
| Worried food would run out before getting money for more | ||
| Often/Sometimes true | 0.359 (0.175, 0.543) | −0.239 (−0.424, −0.055) |
| Never true | 0.109 (0.029, 0.188) | −0.215 (−0.294, −0.136) |
| Reside in own house/apartment | ||
| Yes | 0.128 (0.053, 0.205) | −0.223 (−0.299, −0.147) |
| No | 0.229 (0.027, 0.431) | −0.144 (−0.346, 0.058) |
| Difficulty paying for basics needs (i.e., food/shelter/heat) | ||
| Somewhat/Very hard | 0.255 (0.133, 0.377) | −0.297 (−0.4192, −0.1748) |
| Not very hard | 0.083 (−0.005, 0.17) | −0.1696 (−0.2574, −0.08181) |
PWH: People with HIV; PWOH: People without HIV
Figure 2: Comparisons of BMI Trajectories Pre-Pandemic through During-Pandemic and Pandemic through Post-Pandemic, Modified by Sociodemographic Characteristics, MWCCS, 2018–2024.
PWH: People with HIV; PWOH: People without HIV; Moderate/High Income: >$30,000; Low income: ≤$30,000, HS: High School, “basics” denote basic needs (i.e., food, shelter, heat)
Notably, there was a reduction in mean BMI from 32 (SD: 8.6) in the during-pandemic period to 31.7 (Standard Deviation, SD: 8.5) in the post-pandemic period in the study population (Figure 1). During this time period, there was a statistically significant 0.21 kg/m2/visit decrease (95% CI: 0.14 – 0.28) (Table 2). A similar reduction was noted in the subpopulations where increases had been seen prior; there were statistically significant per visit decreases in BMI among females (−0.295, 95% CI: −0.379, −0.212), baseline BMI category of overweight/obese (−0.264, 95% CI: −0.343, −0.186), Black participants (−0.231, 95% CI: −0.321, −0.140), those with lower income (−0.348, 95% CI: −0.438, −0.257), those with lower education (−0.267, 95% CI: −0.368, −0.165), and those unemployed (−0.459, 95% CI: −0.567, −0.351) (Figure 2). There were reductions among those with (−0.239, 95% CI: −0.424, −0.055) and without (−0.215, 95% CI: −0.294, −0.136) food insecurity as well as those with (−0.297, 95% CI: −0.419, −0.175) and without (−0.170, 95% CI: −0.257, −0.082) economic insecurity. There was also a reduction among those reporting having a stable residence (−0.233, 95% CI: −0.299, −0.147) but, importantly, not among those reporting residential instability (−0.144, 95% CI: −0.346, 0.058). There was a statistically significant per-visit decrease in BMI among both PWH (−0.216, 95% CI: −0.304, −0.129) and PWOH (−0.207, 95% CI: −0.330, −0,084) with no significant difference by serostatus. Given the overall sample was comprised of over 70% females, we repeated all stratified models among females only, and the results were similar (data not shown).
DISCUSSION
We showed a statistically significant weight gain from the pre-pandemic period to the during-COVID-19 pandemic period among participants enrolled in the MWCCS. Also similar to prior literature, this weight gain occurred specifically among vulnerable subpopulations (namely, individuals reporting lower income, lower education, unemployment, living with HIV, and self-identifying as Black race), thereby emphasizing health disparities among these participants. Notably, our study showed an increase in BMI in the pre-pandemic to the during-pandemic period among PWH. Other studies have shown that PWH with underlying obesity were protected from excess weight gain during the pandemic period; disentangling influences on weight trajectories among PWH on ART with underlying obesity during major life stressors such as the COVID-19 pandemic will be important in future studies.[27] While the amount of weight gain shown in this study is small, studies have shown significant health and healthcare cost increases associated with as little as a one-unit change in BMI.[28, 29]
However, contrary to what we had hypothesized, there was a statistically significant reduction in the BMI trajectory between the during-pandemic and post-pandemic periods. Importantly, this reduction was seen among all vulnerable subpopulations in our cohort with the exception of those reporting residential instability; this group did not have a statistically significant reduction in BMI from the during- to post-pandemic period. Stratified by HIV serostatus, the reduction in BMI occurred among both PWH and PWOH, with no statistically significant difference between the two groups.
If these findings are replicated in other populations, it will be important to investigate what individual-level characteristics are associated with loss of the weight that was gained during the pandemic period. The impact on social stressor characteristics such as residential instability not only on weight gained during major life-altering events but also on subsequent weight loss after normal routines resume should be further explored. Our study suggests that contrary to individuals belonging to other social and demographically vulnerable subgroups, individuals reporting residential instability did not lose the weight initially gained during the shutdowns associated with the COVID-19 pandemic. If this finding is corroborated in additional studies, it underscores the importance of considering housing policies as critical factors in mitigating adverse consequences of major life events.
It is possible that our strict study inclusion criteria two study visits with BMI measurement in the pre-pandemic and post-pandemic periods as well as one visit in the during-pandemic period resulted in a skewed subset of less vulnerable participants included in our study. To at least partially explore this, we compared our study population to a comparison population of participants who only needed to complete one visit in the pre-pandemic period and one visit in the during or post-pandemic period. Contrary to what might have been expected, the study population was actually sociodemographically more vulnerable than the comparison population. It does not appear that our study findings are a result of the study population being a more privileged subgroup of participants that were better able to lose the weight gained during the pandemic.
Our study has several limitations. First, due to time and data limitations, we only have two BMI data points in our pre-pandemic and post-pandemic time periods. In addition, due to the MWCCS visit structure, we could not explore BMI changes at a per-month level which is more commonly seen in the literature. It will be important for researchers to continue monitoring weight trajectories for longer periods of time in order to fully understand weight trajectories in this time after the COVID-19 pandemic. Second, as this was meant to be descriptive paper, we did not include detailed information about medication use, clinical characteristics, or other significant lifestyle changes that may have occurred for many individuals during the COVID-19 pandemic. A more detailed exploration of sociocultural and clinical factors that changed during the pandemic will be important in understanding possible mechanisms contributing to weight changes that may have occurred during this time. Third, our study population was nearly three-quarters women. Future studies would benefit from having a broader representation of sex and gender identity. Finally, of PWH in our study population, over 80% were categorized as overweight/obese and over 90% were on ART at baseline. Both of these characteristics have been shown to influence weight gain/loss among PWH.[30] Future studies should explore the influences ART on BMI trajectories associated with major life stressors such as the COVID-19 pandemic among PWH.
To our knowledge, this study is among the first to document a decrease in BMI trajectories in the post-pandemic period compared with the during-pandemic period among populations that experienced weight gain during the COVID-19 pandemic, including among PWH. While unexpected, this could be positive news that at least some of the weight gain observed in many vulnerable populations during the COVID-19 pandemic have been reversed. In addition, there may be particular subgroups, such as those living in unstable housing, that require focused intervention. Understanding what characteristics or factors are associated with successful vs. unsuccessful decreasing BMI trajectories in the post-pandemic period will be important in continuing to address the obesity epidemic in the U.S.
Conflicts of Interest and Sources of Funding:
Authors have no additional conflicts of interest to report.
Data in this manuscript were collected by the Multicenter AIDS Cohort Study (MACS) and the Women’s Interagency HIV Study (WIHS), and the MACS/WIHS Combined Cohort Study (MWCCS). The contents of this publication are solely the responsibility of the authors and do not represent the official views of the National Institutes of Health (NIH). MWCCS (Principal Investigators): Atlanta CRS (Ighovwerha Ofotokun, Anandi Sheth, and Gina Wingood), U01-HL146241; Baltimore CRS (Todd Brown and Joseph Margolick), U01-HL146201; Bronx CRS (Kathryn Anastos, David Hanna, and Anjali Sharma), U01-HL146204; Brooklyn CRS (Deborah Gustafson and Tracey Wilson), U01-HL146202; Data Analysis and Coordination Center (Gypsyamber D’Souza, Stephen Gange and Elizabeth Topper), U01-HL146193; Chicago-Cook County CRS (Mardge Cohen, Audrey French, and Ryan Ross), U01-HL146245; Chicago-Northwestern CRS (Steven Wolinsky, Frank Palella, and Valentina Stosor), U01-HL146240; Northern California CRS (Bradley Aouizerat, Jennifer Price, and Phyllis Tien), U01-HL146242; Los Angeles CRS (Roger Detels and Matthew Mimiaga), U01-HL146333; Metropolitan Washington CRS (Seble Kassaye and Daniel Merenstein), U01-HL146205; Miami CRS (Maria Alcaide, Margaret Fischl, and Deborah Jones), U01-HL146203; Pittsburgh CRS (Jeremy Martinson and Charles Rinaldo), U01-HL146208; UAB-MS CRS (Mirjam-Colette Kempf, James B. Brock, Emily Levitan, and Deborah Konkle-Parker), U01-HL146192; UNC CRS (M. Bradley Drummond and Michelle Floris-Moore), U01-HL146194. The MWCCS is funded primarily by the National Heart, Lung, and Blood Institute (NHLBI), with additional co-funding from the Eunice Kennedy Shriver National Institute of Child Health & Human Development (NICHD), National Institute on Aging (NIA), National Institute of Dental & Craniofacial Research (NIDCR), National Institute of Allergy and Infectious Diseases (NIAID), National Institute of Neurological Disorders and Stroke (NINDS), National Institute of Mental Health (NIMH), National Institute on Drug Abuse (NIDA), National Institute of Nursing Research (NINR), National Cancer Institute (NCI), National Institute on Alcohol Abuse and Alcoholism (NIAAA), National Institute on Deafness and Other Communication Disorders (NIDCD), National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), National Institute on Minority Health and Health Disparities (NIMHD), and in coordination and alignment with the research priorities of the National Institutes of Health, Office of AIDS Research (OAR). MWCCS data collection is also supported by UL1-TR000004 (UCSF CTSA), UL1-TR003098 (JHU ICTR), UL1-TR001881 (UCLA CTSI), P30-AI-050409 (Atlanta CFAR), P30-AI-073961 (Miami CFAR), P30-AI-050410 (UNC CFAR), P30-AI-027767 (UAB CFAR), P30-AI-124414 (ERC-CFAR), P30-MH-116867 (Miami CHARM), UL1-TR001409 (DC CTSA), KL2-TR001432 (DC CTSA), and TL1-TR001431 (DC CTSA).
The authors gratefully acknowledge the contributions of the study participants and dedication of the staff at the MWCCS sites.
Contributor Information
Aruna Chandran, Johns Hopkins Bloomberg School of Public Health.
Sarah Olson, Johns Hopkins Bloomberg School of Public Health.
Andrew Edmonds, UNC Gillings School of Global Public Health.
Caitlin A. Moran, Emory School of Medicine.
Jordan E. Lake, McGovern Medical School.
Phyllis Tien, University of California San Francisco.
Ernesto Marques, University of Pittsburgh.
Anjali Sharma, Albert Einstein College of Medicine.
Maria Alcaide, Miller School of Medicine.
Todd Brown, Johns Hopkins School of Medicine.
Deborah Gustafson, SUNY Downstate Health Sciences.
Frank Palella, Northwestern Feinberg School of Medicine.
Michael Plankey, Georgetown University.
Shivanjali Shankaran, Rush University Medical Center.
Jenni Wise, University of Alabama at Birmingham.
References
- 1.Tate T, et al. , HIV infection and obesity: where did all the wasting go? Antivir Ther, 2012. 17(7): p. 1281–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Koethe JR, et al. , Rising Obesity Prevalence and Weight Gain Among Adults Starting Antiretroviral Therapy in the United States and Canada. AIDS Res Hum Retroviruses, 2016. 32(1): p. 50–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Bailin SS, et al. , Obesity and Weight Gain in Persons with HIV. Curr HIV/AIDS Rep, 2020. 17(2): p. 138–150. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Willig AL, et al. , Physical activity trends and metabolic health outcomes in people living with HIV in the US, 2008–2015. Prog Cardiovasc Dis, 2020. 63(2): p. 170–177. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kanters S, et al. , Evidence synthesis evaluating body weight gain among people treating HIV with antiretroviral therapy - a systematic literature review and network meta-analysis. EClinicalMedicine, 2022. 48: p. 101412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ruderman SA, et al. , Brief Report: Weight Gain Following ART Initiation in ART-Naive People Living With HIV in the Current Treatment Era. J Acquir Immune Defic Syndr, 2021. 86(3): p. 339–343. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Herrin M, et al. , Weight Gain and Incident Diabetes Among HIV-Infected Veterans Initiating Antiretroviral Therapy Compared With Uninfected Individuals. J Acquir Immune Defic Syndr, 2016. 73(2): p. 228–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Alvarez JA, et al. , Sex Differences in Metabolic Disorders of Aging and Obesity in People with HIV. Curr HIV/AIDS Rep, 2024. 22(1): p. 3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Khubchandani J, Price JH, and Sharma S, COVID-19 Pandemic and Weight Gain in American Adults: A Nationwide Population-Based Study. Diabetes and Metabolic Syndrome: Clinical Research & Reviews, 2022. 16(1): p. 1–6. [Google Scholar]
- 10.Almandoz JP, et al. , Changes in body weight, health behaviors, and mental health in adults with obesity during the COVID-19 pandemic. Obesity (Silver Spring), 2022. 30(9): p. 1875–1886. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Sauceda JA, et al. , Brief Report: The Influence of the COVID-19 Pandemic on the Physical, Social, and Mental Health of Black and Latinx Young People With HIV in the United States. J Acquir Immune Defic Syndr, 2023. 93(3): p. 187–190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Chang KM, et al. , Body Mass Index Change in People Living with HIV during the COVID-19 Pandemic in New York. Open Forum Infectious Diseases, 2022. 9(Supp 2): p. S599–600. [Google Scholar]
- 13.Bhutani S, vanDellen MR, and Cooper JA, Longitudinal Weight Gain and Related Risk Behaviors during the COVID-19 Pandemic in Adults in the US. Nutrients, 2021. 13(2). [Google Scholar]
- 14.Chew HSJ and Lopez V, Global Impact of COVID-19 on Weight and Weight-Related Behaviors in the Adult Population: A Scoping Review. Int J Environ Res Public Health, 2021. 18(4). [Google Scholar]
- 15.Goitia J, et al. , Factors associated with weight gain during the COVID-19 pandemic. Obes Res Clin Pract, 2022. 16(2): p. 174–176. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Freedman DS, et al. , Weight gain among US adults during the COVID-19 pandemic through May 2021. Obesity (Silver Spring), 2022. 30(10): p. 2064–2070. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Caldwell AE, et al. , Improving lifestyle obesity treatment during the COVID-19 pandemic and beyond: New challenges for weight management. Obes Sci Pract, 2022. 8(1): p. 32–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Mulugeta W, Desalegn H, and Solomon S, Impact of the COVID-19 pandemic lockdown on weight status and factors associated with weight gain among adults in Massachusetts. Clin Obes, 2021. 11(4): p. e12453. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Clennin M, et al. , Association of Housing Instability with Obesity Status Among Insured Adults. Am J Prev Med, 2024. 67(3): p. 417–422. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Yanovski JA, et al. , A prospective study of holiday weight gain. N Engl J Med, 2000. 342(12): p. 861–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Cooper JA and Tokar T, A prospective study on vacation weight gain in adults. Physiol Behav, 2016. 156: p. 43–7. [DOI] [PubMed] [Google Scholar]
- 22.Bhutani S and Cooper JA, COVID-19-Related Home Confinement in Adults: Weight Gain Risks and Opportunities. Obesity (Silver Spring), 2020. 28(9): p. 1576–1577. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.D'Souza G, et al. , Characteristics of the MACS/WIHS Combined Cohort Study: Opportunities for Research on Aging With HIV in the Longest US Observational Study of HIV. Am J Epidemiol, 2021. 190(8): p. 1457–1475. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Institute of Education Sciences, U.S. Education in the Time of COVID. 2022, National Center for Education Statistics (NCES). [Google Scholar]
- 25.National Academy for State Health Policy. 2021 COVID-19 State Restrictions, Re-Openings, and Mask Requirements. State Tracker 2022 January 11, 2022. [cited 2025 September 11]; Available from: https://nashp.org/state-tracker/2021-covid-19-state-restrictions-re-openings-and-mask-requirements/. [Google Scholar]
- 26.Ruhm CJ, US State Restrictions and Excess COVID-19 Pandemic Deaths. JAMA Health Forum, 2024. 5(7): p. e242006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Taramasso L, et al. , Weight Gain in Overweight and Obese People with HIV-The OBHIV Cohort. J Clin Med, 2024. 13(5). [Google Scholar]
- 28.Wang F, et al. , Association of healthcare costs with per unit body mass index increase. J Occup Environ Med, 2006. 48(7): p. 668–74. [DOI] [PubMed] [Google Scholar]
- 29.Who Cvd Risk Chart Working Group, World Health Organization cardiovascular disease risk charts: revised models to estimate risk in 21 global regions. Lancet Glob Health, 2019. 7(10): p. e1332–e1345. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Bansi-Matharu L, et al. , Contemporary antiretrovirals and body-mass index: a prospective study of the RESPOND cohort consortium. Lancet HIV, 2021. 8(11): p. e711–e722. [DOI] [PubMed] [Google Scholar]