C - reactive protein and interleukin - 6 levels among human immunodeficiency virus -infected patients with dysglycemia in Tanzania

Improved access to antiretroviral therapy (ART) and advances in HIV care in sub-Saharan Africa (SSA) has significantly increased life expectancy for people living with HIV (PLHIV) [1]. However, with aging and more time on ART, PLHIV are now more susceptible to non-communicable diseases (NCD) such as dysglycemia, that is, the presence of impaired fasting glucose (IFG), impaired glucose tolerance (IGT), or type 2 diabetes mellitus (T2DM) [2, 3]. Published reports from SSA reveal an escalating burden of dysglycemia among PLHIV [2, 4, 5], up to four times more compared to HIV negative controls [5]. This may be due to a mix of HIV related risk factors including chronic systemic inflammation [6, 7], use of ART such as stavudine [8], zidovudine [8], and protease inhibitors [4], and traditional risk factors such as physical inactivity, harmful alcohol use, smoking, overweight, and obesity [9‐11].

It has been hypothesized that dysglycemia in PLHIV may be a result of an ongoing chronic inflammatory response [12, 13]. Several studies from high income countries (HIC) among PLHIV have demonstrated correlation between high levels of pro-inflammatory cytokines such as IL-6, CRP and tumour necrosis factor (TNF α), and dysglycemia, independent of body mass index (BMI) and age [12, 13]. In SSA, inflammation is reported to increase the risk of mortality [7, 14] and cardiovascular disease (CVD) among PLHIV [15], however, data on inflammation and dysglycemia is limited. Furthermore, extrapolating findings from HIC to PLHIV in SSA may not be reliable because of the existence of heterogeneity between inflammatory markers and type 2 diabetes based on race/ethnicity [16], a lower prevalence of traditional risk factors such as obesity in SSA compared to HIC which may influence systemic inflammation [17], and a higher background inflammatory state in the general SSA population [17]. SSA regional data on the association between chronic inflammation and dysglycemia among PLHIV is therefore needed.

The inflammatory markers of concern are many [18] but due to the biological and laboratory reproducibility of IL-6 and CRP, and their ability to predict the occurrence of non AIDS events [19], this study assessed the levels of CRP and IL-6 among PLHIV with dysglycemia in Tanzania.

The prevalence of dysglycemia among our sampled PLHIV population on ART was 32%. We found that high CRP or being on a stavudine based regimen were significantly associated with having dysglycemia in multivariate analysis. A trend towards increased WHR (89%) and being overweight/obese BMI (44%), was observed among those with dysglycemia, but these did not reach statistical significance.

Although the prevalence of overt DM was low at 0.8%, the prevalence of pre-diabetes mellitus (pre-DM) comprising of IFG and IGT was high at 31.7%. In a recent review, the prevalence of DM and pre-DM was estimated to range from 1 to 26% and 19–47% among PLHIV in SSA [17]. The highest prevalence of DM and pre-DM was reported in Cameroon at 47 and 27% respectively among ART naive PLHIV, but no association with inflammatory markers was done [2]. In Tanzania, Maganga et al. found 18 and 14.7% DM and pre-DM prevalence amongst PLHIV respectively, however this burden could not be explained by known risk factors such as BMI, age, or ART duration among those receiving therapy, suggesting a potential role of inflammation [5]. Estimates from the HIV negative population in rural communities in northern Tanzania reported a comparatively lower prevalence of DM, and pre-DM, 2.8 and 2.5% respectively [27] which highlights the increased risk of dysglycemia among PLHIV. Furthermore, PrayGod et al. reported 1.5% and over 20% prevalence for DM and pre-DM respectively among underweight HIV-infected participants [28], and Mohammed et al. found a prevalence of 6.4 and 19.6% for DM and pre-DM respectively among Ethiopians [29]. While largely consistent, our study participants had longer times on ART (8 years vs 2 and 5 years) [28, 29]. The consistently high prevalence of pre-DM presents an opportunity for research on interventions to interrupt disease progression, as well as the urgent integration of routine screening for dysglycemia among PLHIV in SSA.

In our study, high CRP was associated with a 2 fold increased odds of having dysglycemia among PLHIV on 1st line ART with undetectable viral load. In SSA, studies on inflammation and dysglycemia among PLHIV are limited to a follow up cross sectional study [28] which did not find association between dysglycemia and either baseline or follow-up CRP after 2–3 years on ART. Inability to mount a CRP response could be explained by low CD4 count of 127 ± 99 μl/ml in their study as compared to 520 ± 236 μl/ml in the current study, as severe immunodeficiency (< 50 μl/ml) is reported to significantly limit inflammatory response, in addition to an increased risk of death [30]. This is observed in the above study [28] where only 57% out of 478 were alive at trial conclusion [28]. Several studies from HIC have reported higher levels of CRP and IL-6 to be associated with the incidence of dysglycemia [31, 32]. However how molecular mediators of inflammation lead to dysglycemia is not fully understood [33]. According to Lufti et al., immune inflammatory response may result to a change in metabolism by initiating insulin resistance in order to reduce energy consumption for body activities other than host defence [34]. Persistent inflammation among PLHIV due to inability to clear the virus or dysbiosis induce prolonged insulin resistance causing chronic restructure of metabolic pathways that may lead to diabetes [34, 35]. This is evidenced by a follow up study showing graded increases of high sensitive CRP (hs-CRP) and IL-6 to be associated with the increased risk of developing diabetes; those in the highest quartile of baseline plasma hsCRP had 5 times greater risk than those in the lowest quartile (HR = 5.13;95%CI 2.6–10.1), for those in the highest quartile of IL-6 rates were 3 times greater than those in the lowest quartile (HR 3.45;95% CI 1.91–6.23) [12]. Furthermore Brown et al. reported association between hs-CRP and incidence of diabetes at 48 weeks of ART initiation independent of known diabetes risk factors, however no significant associations with IL-6 were reported in this study [13]. These findings support chronic low-grade inflammation hypothesis in the development of DM, suggesting that CRP might be a useful marker in prediction of glucose abnormalities among PLHIV [28, 29], however, causation cannot be established through our current research design.

Of note, in this study, IL-6 had no association with dysglycemia. Weaker associations between dysglycemia and IL-6 compared to dysglycemia and CRP have also been observed in other studies [12, 36]. This may be because IL-6 has a much lower half life in circulation compared to CRP [37]. On the other hand, there are studies that reported IL-6 to be strongly associated to non AIDS events compared to CRP during HIV infection [38], therefore variability observed in our study may be due to chance. We did not find studies in SSA that looked at IL-6 as a potential risk factor for dysglycemia among PLHIV, and to the best of our knowledge, this is the first study that looked at both CRP and IL-6 and dysglycemia among SSA PLHIV.

Stavudine use was associated with a 1.99 fold risk of having dysglycemia. Twenty nine percent of our study participants were on this regimen despite WHO recommendations to discontinue the use of thymidine analogue NRTI [39]. Stavudine use has been consistently associated with insulin resistance and a 19% increased risk of developing DM [8, 40]. The proposed mechanism involves inhibition of enzyme DNA polymerase-γ responsible for mitochondrial function [8]. This results to mitochondrial dysfunction that leads to insulin resistance [41]. Studies conducted among PLHIV have been complicated by the confounding effect of HIV infection where a reduction of mitochondrial DNA is observed in absence of ART [42]. However data from HIV uninfected populations have demonstrated a 52% reduction of muscle mitochondrial DNA after 1 month of stavudine exposure [43]. Therefore consumption of stavudine for PLHIV may exacerbate mitochondrial dysfunction, hastening the occurrence of diabetes. The risk is ever increasing with cumulative exposure. This was not observed in our study with a median ART duration of 8 years. Abrahams et al. [44] also did not find association with dysglycemia in PLHIV with a median duration of use of 6.8 years on stavudine. On the contrary, a 9 year follow up study in Senegal that involved participants who received stavudine among other regimens reported a 10% increased risk of dysglycemia after 4 years of treatment [45]. The majority of studies that have reported a lack of association between ART consumption and dysglycemia have had either a small sample size or a small number of different ART combinations [5].The overwhelming benefits of ART outweigh the occurrence of non AIDS events [46], therefore efforts to counteract the adverse events may be more useful to reduce the burden of co-morbidities.

Traditional risk factors for dysglycemia in PLHIV include: older age [4, 47], male gender [9, 48], family history of diabetes [49], higher BMI [9, 50], high WHR [51], smoking [52], alcohol consumption [52], hypertension [29] and physical inactivity [28]. We did not find a statistically significant association for any of the above with dysglycemia in this study. Our study participants were majority female (75%), reported low prevalence of smoking (3%), alcohol consumption (15%), and family history of DM (9%), which may also have limited our ability to find associations between these risk factors and dysglycemia. There was a trend towards increased WHR (89%) and being overweight/obese BMI (42%) among those with dysglycemia, but this did not reach statistical significance. Participants with dysglycemia were slightly younger with the majority falling in the 30–49 age category, however the association was also not statistically significant.

While our findings are among the few studies on inflammation and dysglycemia in Africa, they remain correlative. Research gaps that can be explored in the future include: i) conducting a longitudinal study that will further assess the relationship between inflammation and incident dysglycemia, and ii)intervention studies among the pre-DM PLHIV population to prevent disease progression to DM. Different interrupters can be explored such as nutrition, anti-inflammatory agents and lifestyle modifications.

One of the main limitations of this study was that 75% of study participants were female. It was challenging to get men to participate, possibly because of the 2 h wait period for the OGTT test. Poor male participation in health services is reported in several studies in Africa and this has been associated with a higher mortality among HIV positive males compared to females in SSA [53]. A small sample size may have also limited the ability to detect other significant associations. Finally, our study did not use high sensitive CRP (hs-CRP) to measure CRP, which may have underestimated the burden of inflammation.

This study showed high CRP was significantly associated with dysglycemia among PLHIV with undetectable viral load. Taking stavudine was associated with increased odds of dysglycemia.