 |
|
|
| |
6% New Onset Diabetes in D.A.D. Study: Risk factors for new onset diabetes mellitus (DM) in HIV patients
|
| |
| |
Reported by Jules Levin
IAS-Rio July 2005
ED NOTE from Jules Levin: this data does not consider the rates of insulin resistance among these study patients. Development of insulin resistance starts before diabetes and can predict the development of diabetes. These glucose abnormalities are associated with developing HAART disease and can be caused by body changes. As well, diabetes is a risk factor for renal dysfunction. As people with HIV age these become more serious considerations. Long-term patient cohorts and studies are needed to assess these problems and better characterize the concerns and the treatment interventions.
Sabin C.1, Friis-Moller N.2, Reiss P.3, Weber R.4, d'Arminio Monforte A.5, Dabis F.6, El-Sadr W.7, de Wit S.8, Mateu S.9, Kirk O.10, Pradier C.11, Morfeldt
L.12, Law M.13, Lundgren J.10
1Royal Free & UC Medical School, London, United Kingdom, 2Copenhagen HIV Programme (CHIP), Hvidovre University Hospital, Copenhagen,
Denmark, 3ATHENA, HIV Monitoring Foundation, Academic Medical Center, Amsterdam, Netherlands, 4SHCS, University Hospital, Zurich, Switzerland,
5ICONA, L Sacco Hospital, University of Milan, Milan, Italy, 6Aquitaine Cohort, Bordeaux University Hospital, INSERM U593, Bordeaux, France,
7CPCRA, Columbia University/Harlem Hospital, New York, United States of America, 8Saint-Pierre Cohort, CHU Nice Hospital de l'Archet, Nice, France,
9BASS, Autonomous University of Barcelona, Barcelona, Spain, 10EuroSIDA, CHIP, Hvidovre University Hospital, Copenhagen, Denmark, 11Nice
Cohort, CHU Nice Hopital de l'Archet, Nice, France, 12HivBivus, Karolinska University Hospital, Stockholm, Sweden, 13AHOD, National Centre in HIV
Epidemiology and Clinical Research, UNSW, Sydney, Australia
The D:A:D study (Data Collection on Adverse Events of Anti-HIV Drugs) is an observational, prospective study. Sabin et al reported on a cohort of 22,749 HIV patients on HAART who did not have diabetes mellitus at D:A:D entry (73,803 person-years).
A total of 435 new diagnoses of diabetes mellitus (incidence: 5.89% [95% CI 5.37-6.45/1000 person-years]) were found. Risk factors for new-onset diabetes mellitus included male sex, older age, greater BMI, black race, and earlier calendar year.
After controlling for these risk factors, PI exposure was associated with a small, but significant increased risk of diabetes mellitus.
Increased risk appeared to be linked to the increases in triglyceride in patients receiving HAART over time.
ABSTRACT
The objectives of this study were to assess the incidence of and risk factors for new onset DM in the D:A:D Study, including an assessment of the relationships between DM and combination antiretroviral therapy (CART) and protease inhibitor (PI) use.
DM (defined as fasting glucose >7.0 mmol/L (126 mg/dL) on >2 consecutive occasions or initiation of anti-diabetic therapy) has been collected
prospectively in D:A:D since its initiation.
Follow-up was considered from D:A:D entry to date of DM diagnosis, death, 1/2/2004 or last clinic visit.
Poisson regression models assessed the relation between DM, CART and PI exposure adjusting for the following factors: age, sex, body-mass index (BMI), race, family history of CVD, smoking, transmission route, cohort, calendar year.
Additional models considered the role of metabolic markers including total
cholesterol [TC], HDL cholesterol [HDL-c], and triglycerides [TG].
Results
435 new diagnoses of DM occurred among 22,749 individuals without DM at D:A:D entry
(73,803 person-years, incidence: 5.89 [95% CI 5.37-6.45]/1000 person-years).
Neither CART exposure (RR per year: 1.01 [0.97-1.06]), nor PI exposure (RR per year: 1.03 [0.98-1.07]) were univariably associated with DM risk.
TC (1.08 [1.01-1.14] per mmol/L), HDL-c (0.32 [0.21-0.49] per mmol/L) and TG (5.20 [3.89-6.94] per log2 higher) were all univariably associated with DM risk.
After adjustment for other factors, PI exposure became an independent predictor of DM risk (RR per year 1.06 [1.01-1.12], p=0.02).
Other independent predictors of DM were: older age, male sex, increased BMI, black race, smoking status, earlier calendar year.
Introducing TC or HDL-c into the model did not modify the relationship between PI and DM, whereas adjusting for TG removed the PI effect.
Conclusions: Our results confirm that PI exposure is associated with the risk of DM, and may be associated with hypertriglyceridaemia. Our results should be interpreted with caution; despite rigorous monitoring, we cannot exclude the possibility of ascertainment bias.
|
| |
|
|
|
|
|
