| |
HIV and Obesity Comorbidity Increase Interleukin
6 but Not Soluble CD14 or D-Dimer
|
| |
| |
Download the PDF here
JAIDS Aug 15 2017 Taylor, Barbara S. MD, MS*; So-Armah, Kaku PhD; Tate, Janet P. ScD; Marconi, Vincent C. MD; Koethe, John R. MD Bedimo, Roger J. MD; Butt, Adeel A. MD, MS#; Gibert, Cynthia L. MD**; Goetz, Matthew B. MD; Rodriguez-Barradas, Maria C. MD; Womack, Julie A. CNM, APRN, PhD; Gerschenson, Mariana PhD; Lo Re, Vincent III MD, MSCE; Rimland, David MD; Yin, Michael T. MD, MS##; Leaf, David MD***; Tracy, Russell P. PhD; Justice, Amy C. MD, PhD; Freiberg, Matthew S. MD
Objectives: Obesity prevalence among people living with HIV (HIV+) is rising. HIV and obesity are proinflammatory states, but their combined effect on inflammation (measured by interleukin 6, IL-6), altered coagulation (D-dimer), and monocyte activation (soluble CD14, sCD14) is unknown. We hypothesized inflammation increases when obesity and HIV infection co-occur.
Methods: The Veterans Aging Cohort Study survey cohort is a prospective, observational study of predominantly male HIV+ veterans and veterans uninfected with HIV; a subset provided blood samples. Inclusion criteria for this analysis were body mass index ≥ 18.5 kg/m2 and biomarker measurement. Dependent variables were IL-6, sCD14, and D-dimer quartiles. Obesity/HIV status was the primary predictor. Unadjusted and adjusted logistic regression models were constructed.
Results: Data were analyzed for 1477 HIV+ and 823 uninfected participants. Unadjusted median IL-6 levels were significantly higher and sCD14 levels significantly lower in obese/HIV+ compared with nonobese/uninfected (P <0.01 for both). In adjusted analyses, the odds ratio for increased IL-6 in obese/HIV+ patients was 1.76 (95% confidence interval: 1.18 to 2.47) compared with nonobese/uninfected, and obesity/HIV+ remained associated with lower odds of elevated sCD14. We did not detect a synergistic association of co-occurring HIV and obesity on IL-6 or sCD14 elevation. D-dimer levels did not differ significantly between body mass index/HIV status groups.
Conclusions: HIV-obesity comorbidity is associated with elevated IL-6, decreases in sCD14, and no significant difference in D-dimer. These findings are clinically significant, as previous studies associated these biomarkers with mortality. Future studies should assess whether other biomarkers show similar trends and potential mechanisms for unanticipated sCD14 and D-dimer findings.
INTRODUCTION
Obesity is a leading health threat in the United States1-3 and is more common in minorities and those with lower socioeconomic status.4-9 These same individuals are also most severely impacted by the HIV epidemic.10,11 Because potent antiretroviral therapy (ART) is now widely available, weight loss previously associated with untreated HIV infection has been supplanted by weight gain. Consequently, obesity is increasingly common in people living with HIV infection.12-17
Both obesity and HIV infection independently increase risk for cardiovascular disease (CVD).18-22 A hypothesized mechanism for the increased cardiovascular risk, common to both, is immune system alteration.23-29 Although obesity has well-established associations with traditional risk factors for CVD including hypertension, hyperlipidemia, and diabetes, both HIV and obesity are associated with immune system alterations that may independently impact risk.23-28 Although both HIV and obesity are associated with altered immunity, their combined effect on inflammation, altered coagulation, and monocyte activation is unclear.
We hypothesized that the combination of obesity and HIV infection (obesity/HIV) is associated with increased inflammation when compared with either condition alone or the absence of both conditions. In a subset of participants from the Veterans Aging Cohort Study (VACS) survey cohort, we examined the relationship between obesity/HIV and biomarkers of inflammation (interleukin 6), altered coagulation (D-dimer), and monocyte activation (soluble CD14, sCD14). These 3 biomarkers were chosen because they represent different inflammatory pathways and are altered in the context of HIV and obesity.23,30-33
DISCUSSION
In this cross-sectional examination of inflammatory biomarkers in HIV+ and uninfected participants, we found that obesity/HIV comorbidity is associated with elevated IL-6 and decreased sCD14, particularly in the setting of uncontrolled viral replication. Obese/HIV+ participants, when compared with nonobese/uninfected participants, had statistically significant elevations in IL-6, a biomarker of inflammation, and decreases in sCD14, a biomarker of monocyte activation. We did not detect a significant association between obesity/HIV status and D-dimer, a biomarker of altered coagulation.
To our knowledge, this is the largest study comparing IL-6, sCD14, and D-dimer by BMI and HIV status in a cohort including HIV+ and uninfected participants. A recent investigation compared 35 HIV+ with 30 matched, uninfected, obese participants, and found that IL-6 did not significantly differ according to HIV status. They also found sCD14 was increased in obese/HIV+ when compared with obese/uninfected participants, the opposite of the association seen in our study. The characteristics of this smaller cohort differed from those in our cohort in terms of age and gender distributions, and D-dimer levels were not measured.30
Multiple prior studies investigated the association between these biomarkers and obesity in HIV+ participants without comparison with an uninfected population. IL-6 did not have a consistent association with adipose tissue mass in HIV+ participants in 1 investigation,49 whereas in other reports, IL-6 was associated with increases in BMI, particularly for HIV+ men.31,50 sCD14 was found to increase with weight gain in virologically suppressed, predominately normal weight HIV+ participants between 0 and 48 weeks after antiretroviral initiation across 9 countries.29 In the Strategies for Management of Antiretroviral Therapy (SMART) Study, investigators did not detect an association between obesity and D-dimer in HIV+ participants,32 similar to our findings. Likewise, D-dimer was not associated with obesity in a cohort of uninfected participants.33 These studies, which did not compare HIV+ with uninfected participants in the same cohort, have limited ability to comment on the differential impact of obesity on these biomarkers in those living with HIV versus without HIV infection.
Interleukin 6
We found that obesity and HIV infection contributed to elevations in IL-6, an inflammatory cytokine strongly associated with morbidity and mortality in HIV+ cohorts.51-55 Median IL-6 levels were significantly higher in obese/HIV+, when compared with nonobese/uninfected, but there was no statistical difference between IL-6 levels in obese/HIV+ and those with either condition alone (nonobese/HIV+ or obese/uninfected, Fig. 1). Thus, we did not find evidence for a synergistic effect of HIV/obesity comorbidity on IL-6 elevations.
Obesity without HIV and HIV viremia without obesity were each associated with IL-6 elevation, and obese/uninfected individuals were more likely to have elevated IL-6 levels than nonobese/uninfected in all 3 logistic regression models. Interestingly, nonobese/HIV+ individuals with HIV-1 RNA <500 copies/mL had similar IL-6 levels to nonobese/uninfected (Table 3), implying that controlled HIV disease alone, without obesity, is not associated with significant elevations in IL-6 in this cohort. Although obese/HIV+ participants with HIV-1 RNA ≥500 copies/mL had a higher prevalence of elevated IL-6 than nonobese/uninfected, this association did not reach statistical significance. The anticipated elevation of IL-6 derived from uncontrolled viral replication may have been present, but the small sample size of only 83 obese/HIV+ participants with HIV-1 RNA ≥500 copies/mL may have limited our ability to detect an association.
The source of IL-6 in obesity/HIV comorbidity may be an important consideration in defining the implications of and future hypotheses generated from our findings. IL-6 can be derived from activated macrophages in tissues, such as blood vessels, or it can be produced by adipocytes, where levels increase in association with adipocyte diameter.56-58 In obesity, IL-6 is predominately adipose tissue derived, which may differ from chronic HIV infection, where it is likely induced by immune cell activation. Whether (1) adipocyte-derived IL-6 has more local effects on tissues while IL-6 production from immune cell activation has more systemic effects and whether (2) the distribution of local versus systemic effects has impact on morbidity or mortality are important questions for future research.
Soluble CD14
We observed lower sCD14 levels in obese/HIV+ participants compared with nonobese/HIV+ participants, which is in line with results from some previous studies.45,47,58 Other studies have focused on the association between HIV status and sCD14 elevation in normal weight participants and found increased sCD14 in HIV+ participants.59-61 A possible explanation for these discordant findings is that hepatic steatosis in obesity leads to lower sCD14 levels because of decreases in hepatocyte-derived sCD14.62 When HIV and obesity co-occur, sCD14 levels may be influenced by multiple factors including monocyte activation within adipose tissue, other co-occurring inflammatory conditions (eg, diabetes, hepatitis C), and the presence of hepatic steatosis.63,64 The etiology of the unanticipated findings that sCD14 decreases in the setting of obesity/HIV comorbidity deserve further exploration, and other biomarkers of monocyte activation, such as soluble CD163 or macrophage inflammatory protein-1α, should be examined to determine whether they follow a similar trend.45 Other illustrative lines of research would repeat this study accounting for differences in hepatic steatosis between groups.
D-Dimer
In fully adjusted models, obesity/HIV comorbidity was not associated with D-dimer except when HIV status was stratified by viral suppression (ie, HIV-1 RNA less or greater than 500 copies/mL). Unsuppressed HIV viremia was associated with increased D-dimer, although this was only statistically significant among nonobese people. The positive association between obesity and D-dimer among HIV-uninfected people did not reach statistical significance. Overall, we did not detect a synergistic effect of obesity/HIV comorbidity on D-dimer elevations.
There are limitations to this analysis. This is an observational, cross-sectional study, preventing us from examining trends over time in inflammatory parameters. It would be useful to examine trends in relation to the duration of virologic suppression or in the setting of weight change or aging. BMI is used as a measure of adiposity rather than a measurement of waist circumference or visceral adipose tissue, known to be more strongly correlated with cardiovascular outcomes. The population is overwhelmingly male and older, limiting generalizability to women and younger HIV+ people. The individual biomarkers measured may not capture the full complexity of the immune processes they represent, although all 3 are well studied in both HIV+ and uninfected cohorts. Data on diet and physical activity, which can confound associations of obesity with inflammatory biomarkers, were not included in this analysis. As with any observational study, there is the possibility of unmeasured confounding from drug use, multiple comorbidities, or other factors.
In summary, obesity/HIV comorbidity is associated with elevated IL-6 and decreased sCD14. However, we did not detect a synergistic effect of obesity/HIV comorbidity on IL-6, sCD14, or D-dimer. This may suggest overlap in the mechanisms by which obesity and HIV contribute to alterations in these biomarkers and the immune processes they represent. The clinical implications of these findings deserve further exploration, as the impact of obesity on morbidity and mortality in HIV+ individuals remains unclear. Recent studies from the North American AIDS Cohort Collaboration on Research and Design suggest that weight gain and obesity are associated with improved CD4+ cell recovery and decreased incidence of noncommunicable diseases.65,66However, data from the Data Collection on Adverse Events of Anti-HIV Drugs cohort study show that short-term weight gain after ART initiation, particularly in those with a normal baseline BMI, was associated with increased risk of incident CVD and diabetes.67
Despite uncertainty surrounding the impact of rising BMI in this population, the difference in median IL-6 levels between obese/HIV+ and nonobese/uninfected participants our analysis is similar to or greater than differences associated with increased mortality risk in other prospective studies.53,54 As obesity prevalence among people living with HIV rises,17 further investigation into the long-term implications of this weight gain is needed. Future prospective studies should assess whether intentional weight loss or similar interventions affect inflammatory processes similarly in HIV-infected and uninfected people.68-70
|
|
| |
| |
|
|
|
