Abstract
Background: Integrase inhibitor (INSTI) use has been associated with greater weight gain (WG) among people living with HIV (PLWH), but it is unclear how this effect compares in magnitude to traditional risk factors for WG. We assessed the population attributable fractions (PAFs) of modifiable lifestyle factors and INSTI regimens in PLWH who experienced a ≥5% WG over follow-up.
Methods: In an observational cohort study from 2007 to 2019 at Modena HIV Metabolic Clinic, Italy, ART-experienced but INSTI-naive PLWH were grouped as INSTI-switchers vs non-INSTI. Groups were matched for sex, age, baseline BMI and follow-up duration. Significant WG was defined as an increase of ≥5% from 1st visit weight over follow-up. PAFs and 95% CIs were estimated to quantify the proportion of the outcome that could be avoided if the risk factors were not present.
Results: 118 PLWH switched to INSTI and 163 remained on current ART. Of 281 PLWH (74.3% males), mean follow-up was 4.2 years, age 50.3 years, median time since HIV diagnosis 17.8 years, CD4 cell count 630 cells/µL at baseline. PAF for weight gain was the greatest for high BMI (45%, 95% CI: 27–59, p < 0.001), followed by high CD4/CD8 ratio (41%, 21–57, p < 0.001) and lower physical activity (32%, 95% CI 5–52, p = 0.03). PAF was not significant for daily caloric intake (−1%, −9-13, p = 0.45), smoking cessation during follow-up (5%, 0–12, p = 0.10), INSTI switch (11%, −19-36; p = 0.34).
Conclusions: WG in PLWH on ART is mostly influenced by pre-existing weight and low physical activity, rather than switch to INSTI.
Background
Obesity in HIV depicts the irony of success of HIV care: initially a fatal disease associated with loss of both lean and fat mass (‘wasting syndrome’), HIV is now a chronic condition in which aging and progressive weight gain (WG) describes the contemporary HIV disease phenotype. The progressive increase in body weight raises the concern regarding the possibility of an obesity epidemic affecting people living with HIV (PLWH) and the consequences of co-morbid obesity.Citation1 In an AIDS Clinical Trial Group (ACTG) study of ART initiation in resource-diverse settings (A5175), more than 25% of participants were classified as overweight or obese at entry, and approximately 40% of participants were overweight or obese by week 144.Citation2
Integrase strand transfer inhibitor (INSTI) use has been associated with greater weight gain among people living with HIV (PLWH) in some studies, but it is unclear how this effect compares in magnitude to traditional risk factors for weight gain. In the ADVANCE trial, African men and women living with HIV were randomized to ART initiation with an INSTI-based regimen versus efavirenz-based regimen. Less than 10% of men, but up to 40% of women experienced obesity by week 48,Citation3 suggesting that men and women may display different patterns of WG. Other factors may contribute to WG, such as lifestyles, with particular regards to greater caloric consumption or limited physical activity be of paramount importance, and was seldom collected in larger cohorts or randomized clinical trials. In PLWH, higher intensity of physical exercise is needed to obtain physical function improvements than the general population,Citation4 and little is known about whether physical activity needs for body weight maintenance differ by HIV serostatus. Lower CD4 cell count and low CD4/CD8 ratio have been associated with WG in ART-naive individuals initiating ART.Citation5
A well-known epidemiological approach is the population attributable fraction (PAF) tool to quantify the proportion of a specific health condition that could be eliminated if particular risk factors were not present.Citation6 The aim of this study was to assess the population attributable fractions (PAFs) of modifiable lifestyles and INSTI regimens in ART experienced PLWH who experienced at least a 5% weight gain over 4 years follow-up.
Methods
Study design, inclusion and exclusion criteria
This was an observational retrospective longitudinal study that included ART-experienced PLWH attending Modena HIV Metabolic Clinic (MHMC) from January 2007 to July 2019. MHMC is a tertiary level referral centre established in 2004 where PLWH are screened for non-infectious co-morbidities (NICM), immuno-metabolic disorders, geriatric syndromes and frailty.
PLWH with at least two available visits, weight assessment and data on physical activity and diet were included. At the beginning of the observation period, all participants were INSTI-naïve. During follow-up, participants either switched (INSTI-s) to an INSTI-based regimen or remained on suppressive, non-INSTI ART (INSTI-n), with no switch to INSTI regimens. PLWH in INSTI-s could have switched to another INSTI, while PLWH in INSTI-n could have switch to other non-INSTI based regimens (e.g. from protease inhibitors to non-nucleoside reverse transcriptase inhibitors or vice versa). At baseline, groups were matched for sex, age, 1st visit body mass index (BMI) and follow-up duration. In the INSTI-s group the baseline visit was prior to switch, with maximum 12 months difference between BMI assessment and INSTI initiation. In INSTI-n, baseline was chosen as the closest visit to a matched visit in INSTI-s (within a 6-month tolerance). A follow-up visit is the last available observation in both groups that met inclusion criteria.
Switch to TAF was not included in this study in consideration that at baseline the drug was not available. Patients were eligible for analyses if lifestyles including physical activity and diet data were available at last observation only, as the collection of these data has begun in 2016.
Outcome measure
Significant weight gain was defined as an increase of ≥5% from 1st visit weight over a mean 4 years follow-up.
Covariates
Demographic, anthropometric, HIV-related and immune-metabolic variables were collected at baseline and follow-up. Moreover, body weight was accurately assessed with total body DEXA, which provides a standardized reliable measure.Citation7
Multimorbidity was defined as the presence of at least two comorbidities.
Physical activity was assessed with International Physical Activity Questionnaire (IPAQ) as metabolic equivalent of task (MET). A MET is the ratio of the rate of energy expended during an activity to the rate of energy expended at rest (1 MET is the rate of energy expenditure while at rest). Level of physical activity was further categorized as low (MET score <600), moderate (601 < MET score > 1500) and intense (MET score >1500).Citation8
Daily caloric intake (DCI) was evaluated with a standardized three-day food diary, and evaluated by dietician using a pictorial food atlas for portion size. Total mean caloric intake was compared with intensity of physical activity. The 2015–2020 Dietary Guidelines Estimated Calorie Needs were used to identify cut-off for high caloric intake in men and women in different age groups according to three different levels of physical activity: sedentary, moderate and intense.Citation9 For example, in the most represented age group 51–55 years, high caloric intake in a moderate active men and women were respectively higher than 2400 and 1800 calories.
Cigarette smoking history was collected at baseline as current or former exposure using the pack-year definition. In consideration of the potential weight change associated with smoking cessation, quitting/never smoker vs continuing smoking during follow up was chosen as dichotomous variable.
HIV variables were collected from electronic patient charts and included, time since HIV diagnosis, CD4 nadir and current value, CD4/CD8 ratio, HIV-1 RNA, current and past exposure to ARVs. Non-infectious comorbidities (NICM) were defined using the European AIDS Clinical Society (EACS) guidelines, as previously described.Citation10
Statistical analysis and PAF calculations
Results were expressed as mean and standard deviation (± SD), or median and interquartile range (IQR) for continuous variables based on the normality of distribution, and as frequencies and percentages for categorical variables. Student’s t-test and ANOVA were applied to identify statistical difference for the continuous variables with normal distribution assessed at baseline and at follow up, while the Mann-Whitney and Walls-Kruskal test was used for those without normal distribution. The χ2 test was performed to assess the frequency of the categorical variables.
In addition to the prevalence of the risk factor, PAF is influenced by the risk of the outcome associated with each preventable or modifiable risk factor. We calculated the crude hazard ratio allowing some risk factors (nadir CD4 cell count, CD4/CD8 ratio, high daily caloric intake, low physical activity, smoke pack year) to vary with time. PAFs and 95% CIs were estimated to quantify the proportion of outcomes that could be avoided if the risk factor was prevented, using the following dichotomic variables: BMI >25 kg/m2 vs <25 kg/m2, DCI above cut-off indicated by 2015–2020 Dietary Guidelines Estimated Calorie Needs, IPAQ MET <600 vs MET > 600, never smokers or quitting vs continuing smoking during follow-up, INSTI vs no-INSTI regimens, nadir CD4 cell count <200 vs >200 cells/µL, time since HIV diagnosis >20 years vs <20 years and CD4/CD8 ratio >1 vs <1.Citation11 The significance of the tests was set to 0.05. The statistical program R, v. 3.6.0 in GNU Linux environment was used to analyze the data.
This study was approved by the University of Modena and Reggio Emilia ethics committee according to the Helsinki declaration.
Results
We included 281 PWH (74% males) with a mean follow-up was 4.2 years (±1.8 SD). At baseline, age 50.3 (±7.9 SD) years, median time since HIV diagnosis 22.3 years (IQR 12.1–23.2), CD4 cell count 630 cells/µL (IQR 482–823); 91.5% had undetectable HIV-1 RNA. 118 PLWH switched to INSTI, 163 remained on current ART.
shows the baseline clinical characteristics by INSTI exposure. INSTI-s had lower nadir CD4 cell count, lower median CD4 cell count, longer time since HIV diagnosis and higher burden of multimorbidity. The prevalence of obesity was 7.4% among INSTI-n and 5.1% among INSTI-s (p = 0.60).
Table 1. The anthropometric and clinical characteristics of PLWH at baseline who remained INSTI naïve and PLWH who switched to an INSTI-based regimen.
Over 4 years of follow up, weight increased 1.6 (±5.7) kg in INSTI-n and 3.1 (±8.2) kg in INSTI-s (p = 0.20), with no difference in the proportion of new cases of obesity (1 vs. 2, respectively) (data presented within the text only).
Seventy-five PLWH experienced a WG of ≥5%: 47% were INSTI-s and 53% INSTI-n (p = 0.42). Other demographic, anthropometric, HIV-related and immuno-metabolic differences between those who had <5% or ≥5% WG are shown in . Those with a significant weight gain were more likely to be women, obese, have a higher CD4 and CD4/CD8, and be less physically active at follow-up.
Table 2. The anthropometric and clinical characteristics of PLWH who had <5% of weight gain and PLWH who had ≥5% of weight gain at baseline and follow-up.
Next, we determined the PAFs for modifiable and HIV- related factors contributing to WG ≥5%. PAF for weight gain was the greatest for high BMI (45%, 95% CI: 27–59, p < 0.001), followed by high CD4/CD8 ratio (41%, 21–57, p < 0.001) and lower physical activity (32%, 95% CI 5–52, p = 0.03). PAF was not significant for daily caloric intake (-1%, −9-13, p = 0.45), smoking cessation during follow-up (5%, 0–12, p = 0.10), INSTI switch (11%, −19-36; p = 0.34) ().
Figure 1. Population attributable fraction for weight gain in PLWH.
Discussion
This study assessed the relative contribution of modifiable lifestyle factors and INSTI regimens to predict weight gain over 4 years of follow-up. Using a clinically relevant gain of ≥5% weight from baseline, we found that one of every four patients (26.4%) experienced significant WG, but the prevalence of obesity remained 7.5%, below the 10% of general Italian population.Citation12 In contrast to prior studies which demonstrated a significant effect of INSTI use on weight gain, we found that higher BMI, higher CD4/CD8, and lower physical activity had a significant impact on weight gain while INSTI switch did not.
Of the factors that did significantly contribute to ≥5% weight gain, many are consistent with prior studies. While lower BMI is a significant risk factor for weight gain at ART initiation, a higher BMI at the time of switch to INSTI was a risk for INSTI-associated weight gain in the HAILO Cohort of middle-aged and older individuals on suppressive ART.Citation13 A higher CD4/CD8 ratio, was independently associated with WG, consistent with prior studies of weight gain in ART-naïve persons. While CD4 count at the time of ART switch is not a modifiable risk, this immunological biomarker may identify those with additional risk for WG.Citation14 Lastly, lower physical activity was the most modifiable risk factor that could affect WG. A MET score ≥600 decreased the risk of weight gain by 31% the risk of WG, which is equivalent to the current recommendations of 150 weekly minutes of moderate or 75 weekly minutes of vigorous exercise. To the best of our knowledge, this is the first study to show the contribution of physical activity to ART-associated weight gain in PLWH.
Using a similar approach paper, Althoff et al. estimated the PAFs of preventable or modifiable risk factors for non-AIDS-defining cancers, myocardial infarction, end-stage liver disease, and end-stage renal disease.Citation6 A high proportion of these conditions could be prevented with interventions on traditional, modifiable risk factors.
Notably, several modifiable factors were not associated with weight gain by PAF, including caloric intake and smoking cessation. Very few patients in our cohort had high caloric intake, and Mediterranean diet is still common in the Italian population. Similarly, while active smokers may have low weight and smoking cessation may be associated with WG, very few patients quit smoking during the follow-up period.
This study has several limitations, many of which are intrinsic in observational cohorts. Change in other ART regimens were not included in this analysis. Switch in ART did not always coincide with collection of variables and risk factors. Switch in ART was determined by the treating provider, and may have associated biases that cannot be completely adjusted for within this analysis. While this cohort is representative of the HIV epidemic in Italy, our data may not be generalizable to other populations with more demographic diversity, that include more people of Afro-American/African or Hispanic descent, or women, who might be at higher risk of WG.
The strengths of the study include the available data on caloric intake and physical activity by mean of validated questionnaires, the duration of follow-up (4 years), and clinical cohort more representative of real-world switch that occur outside of the setting of randomized clinical trials. Moreover, weight was assessed using DEXA scans, which is a more reliable measure of body weight. Further research fat and lean mass should provide additional information on weight changes in both ART-naïve and experienced PLWH.
In summary, our findings emphasize the importance of physical activity in maintaining a healthy weight with HIV, regardless of switch in ART regimen. Guidance on physical activity in the management of obesity in PLWH,Citation1 and guidance on exercise recommendations for older PLWH are available and should be utilized.Citation15 Counselling on the importance of physical activity with ART initiation or switch should routinely be incorporated with every ART prescription. Meeting or exceeding current World Health Organization recommended targets for moderate to intense physical activity (MET > 600) may be a critical step in preventing an obesity epidemic among PLWH.
Disclosure statement
GG and CM received research grant and speaker honorarium from Gilead, ViiV, MERCK and Jansen. GG and CM attended advisory boards of Gilead, ViiV and MERCK. JL has served as a consultant to Merck, GSK, Gilead Sciences and Theratechnologies, and receives grant funding from Gilead Sciences. KME has received grant funding from Gilead, and has served as a consultant to ViiV and Theratechnologies. Other authors reported no conflict of interest.
Data availability statement
The data that support the findings of this study are available on request from the corresponding author (GG).
References
- Lake JE, Stanley TL, Apovian CM, et al. Practical review of recognition and management of obesity and lipohypertrophy in human immunodeficiency virus infection. Clin Infect Dis. 2017;64(10):1422–1429.
- Erlandson KM, Taejaroenkul S, Smeaton L, et al. A randomized comparison of anthropomorphic changes with preferred and alternative efavirenz-based antiretroviral regimens in diverse multinational settings. Open Forum Infect Dis. 2015;2(3):ofv095.
- Venter WDF, Moorhouse M, Sokhela S, et al. Dolutegravir plus two different prodrugs of tenofovir to treat HIV. N Engl J Med. 2019;381(9):803–815. 10.1056/NEJMoa1902824
- Erlandson KM, MaWhinney S, Wilson M, et al. Physical function improvements with moderate or high-intensity exercise among older adults with or without HIV infection. AIDS. 2018;32(16):2317–2326.
- 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):1379–1389.
- Althoff KN, Gebo KA, Moore RD, North American AIDS Cohort Collaboration on Research and Design, et al. Contributions of traditional and HIV-related risk factors on non-AIDS-defining cancer, myocardial infarction, and end-stage liver and renal diseases in adults with HIV in the USA and Canada: a collaboration of cohort studies. Lancet HIV. 2019;6(2):e93–e104.
- Shepherd J, Ng B, Sommer M, Heymsfield SB, Francisco S. HHS public access body composition by DXA. Bone 2017;104:101–105.
- Craig CL, Marshall AL, Sjostrom M, et al. International physical activity questionnaire: 12-country reliability and validity. Med Sci Sports Exerc. 2003;35(8):1381–1395.
- U.S. Department of Health and Human Services and U.S. Department of Agriculture. 2015 – 2020 Dietary Guidelines for Americans. 8th ed. https://health.gov/our-work/food-nutrition/previous-dietary-guidelines/2015.
- 10th EACS Guidelines. (Assessed 24th December 2019). https://www.eacsociety.org/media/guidelines-10.0_final_2_2.pdf
- Mussini C, Lorenzini P, Cozzi-Lepri A, et al. CD4/CD8 ratio normalisation and non-AIDS-related events in individuals with HIV who achieve viral load suppression with antiretroviral therapy: an observational cohort study. Lancet HIV. Netherlands. 2015;2(3):e98–e106.
- State of Health in the EU - Italy country health profile 2017 https://www.euro.who.int/__data/assets/pdf_file/0008/355985/Health-Profile-Italy-Eng.pdf. Accessed May 9, 2021.
- Lake JE, Wu K, Bares SH, et al. Risk factors for weight gain following switch to integrase inhibitor–based antiretroviral therapy. Clin Infect Dis. 2020;71(9):e471–e477.
- Guaraldi G, Zona S, Silva AR, et al. The dynamic association between frailty, CD4 and CD4/CD8 ratio in people aging with HIV. PLoS One. 2019;14(2):e0212283–11.
- Montoya JL, Jankowski CM, O'Brien KK, et al. Evidence-informed practical recommendations for increasing physical activity among persons living with HIV. AIDS. 2019;33(6):931–939.