Obesity was the most important risk factor for developing diabetes in a large US HIV clinic cohort between 2008 and 2018, researchers from the University of Alabama at Birmingham report in a pre-print article released ahead of peer review this month.
The study investigators say that preventing excess weight gain is critical for prevention of diabetes in people with HIV.
Having a body mass in the obese or morbidly obese category at the beginning of the study had the greatest impact on the risk of developing diabetes. People with the highest body mass index – a BMI over 40 – were ten times more likely to develop diabetes than people with a weight in the normal range.
In comparison, after controlling for baseline weight, taking an integrase inhibitor raised the risk of developing diabetes by 48%.
A second study, looking at the impact of the antiretrovirals used by people starting treatment between 2007 and 2018, found that people taking an integrase inhibitor had a 31% higher risk of developing diabetes or raised blood sugar in the first six months after starting treatment.
Type 2 diabetes, or diabetes mellitus, develops when the body’s mechanism for controlling blood sugar levels fails and blood glucose levels rise to harmful levels. The long-term consequences of diabetes include an increased risk of cardiovascular disease, kidney failure, blindness and nerve damage in the limbs.
In the population as a whole, diabetes is associated with being overweight. People of South Asian, African and Afro-Caribbean descent are at higher risk of developing diabetes. People with HIV are at higher risk, although it’s unclear if this is because the established risk factors for diabetes are more common in people with HIV or if HIV-related factors raise the risk too. Risks may be higher for women with HIV than for men with HIV.
Studies have shown mixed results regarding the impact of antiretroviral treatment on the risk of developing diabetes, and several have suggested that if the risk of diabetes is raised, it is a consequence of weight gain after starting treatment rather than a direct effect of antiretrovirals on glucose control.
Risk factors in a large US clinic
Physicians at the University of Alabama at Birmingham carried out an analysis of risk factors for new-onset diabetes in people with HIV attending the university’s HIV clinic, one of the largest in the south-eastern United States.
Between 2008 and 2018, 5,339 people with HIV attended at least one clinic visit. People were eligible for inclusion in the analysis if they did not have diabetes at the time of their first visit after 2008, attended the clinic at least twice during the follow-up period and did not have an extended delay between clinic visits; 4,113 were eligible for inclusion in the analysis.
The study population was predominantly Black (62%) and male (77%). Eighty-two percent of the study population was below the age of 50, the median age at study entry was 38 and just under half were either overweight (body mass index 25-29.99) (29%) or obese (BMI of 30 or above) (20%).
Most participants entered care in 2008 or later (79%) and had been living with HIV for less than five years at study entry (64%). Just under half (47%) had a CD4 count below 350 at study entry.
Several comorbidities were common in the study population. Forty-four percent had high blood pressure (hypertension), 23% were taking an anti-hypertensive medication, 31% had dyslipidemia, and 8% were taking a statin or other lipid-lowering medication.
A new diagnosis of diabetes was defined as an Hgb A1C measurement of 6.5% or above and/or two glucose results greater than 200mg/dl (11.1 mmol/l) at least 30 days apart.
During the follow-up period, 252 people developed diabetes at a median age of 51 years. The incidence of diabetes increased over time from 1.04 cases per 1000 person-years of follow-up in 2008 to 1.55 cases per 1000 person-years in 2018. The prevalence of diabetes in the cohort rose from 8% in 2008 to 14% in 2018.
In multivariable analysis diabetes was strongly associated with morbid obesity (BMI > 40) (adjusted hazard ratio 10.5), obesity (BMI 30-39.99) (aHR 3.9) and being underweight (BMI < 18.5) (aHR 2.8).
New-onset diabetes was also associated with cirrhosis or end-stage liver disease (aHR 1.9), dyslipidemia (1.6), non-statin lipid-lowering drug exposure (aHR 1.54) and glucocorticoid treatment (aHR 1.46).
Several HIV-related factors were also associated with new-onset diabetes. Living with HIV for more than ten years (aHR 1.6), a baseline CD4 count between 200 and 350 (aHR 1.55) and integrase inhibitor treatment of any duration (aHR 1.48) each raised the risk of developing diabetes.
To reduce the risk of diabetes and other metabolic complications, “more resources should be devoted to curb excess body fat gain and related conditions in this particularly vulnerable population,” the study investigators conclude.
The investigators note that integrase inhibitor exposure was associated with the development of diabetes independent of obesity, suggesting that in addition to any impact of the drug class on body weight, integrase inhibitors may also lead to increases in visceral fat and truncal obesity that promote insulin resistance and diabetes.
The impact of antiretroviral choice
A more focussed analysis of a larger US population, looking only at the relationship between antiretroviral treatment and new-onset diabetes or hyperglycaemia within six months of starting treatment found an association between integrase inhibitor treatment and diabetes incidence of a similar magnitude to the University of Alabama study.
The study looked at the 6-month period after starting treatment to capture early effects of integrase inhibitors on glucose control and before significant weight gain was likely to occur.
Dr Jane O’Halloran and colleagues at University of Washington School of Medicine, St Louis, looked at the incidence of diabetes after initiating treatment in 42,382 people with HIV between 2007 and 2018. The study population was 74% male and had a mean age of 38 years. As the study collected information about treatment starts from an insurance claims database, no details of ethnicity were available. Ethnicity might affect the risk of developing diabetes and the researchers acknowledge that this lack of data is a limitation of the study.
Fifty-three percent of participants started treatment with an integrase inhibitor, 28% with a non-nucleoside reverse transcriptase inhibitor and 19% with a protease inhibitor. Of those on integrase inhibitors, raltegravir was prescribed more often (39%) than elvitegravir (28%), dolutegravir (26%) or bictegravir (5%).
Approximately two-thirds of participants underwent glucose testing in the six months after initiating treatment and 2.1% of participants were diagnosed with new-onset diabetes or hyperglycaemia during that period (2.3% in the integrase inhibitor group and 1.8% in the non-integrase inhibitor group).
The risk of being diagnosed with diabetes or hyperglycaemia was 31% higher in people who started treatment with an integrase inhibitor compared to other regimens. When the risk of developing diabetes was assessed according to the integrase inhibitor used, the analysis showed a trend towards a higher risk of diabetes in people who took elvitegravir or dolutegravir.
Although use of an integrase inhibitor with tenofovir alafenamide (TAF) has been associated with greater weight gain in other studies, TAF use did not affect the risk of developing diabetes in people taking an integrase inhibitor in this study.
The study authors say that their findings “add to a growing body of evidence that [integrase inhibitors] impact on glycemic control.”
Spieler G et al. Trends in diabetes incidence and associated risk factors in people living with HIV in the current treatment era, 2008-2018. MedRxiv pre-publication, 1 May 2022.
DOI: https://doi.org/10.1101/2022.04.29.22274506
O’ Halloran J et al. Integrase strand transfer inhibitors are associated with incident diabetes mellitus in people with HIV. Clinical Infectious Diseases, published online, 6 May 2022.