Morbidity and aging in HIV-infected persons: the Swiss HIV cohort study.

BACKGROUND: Patterns of morbidity and mortality among human immunodeficiency virus (HIV)-infected individuals taking antiretroviral therapy are changing as a result of immune reconstitution and improved survival. We studied the influence of aging on the epidemiology of non-AIDS diseases in the Swiss HIV Cohort Study. METHODS: The Swiss HIV Cohort Study is a prospective observational cohort established in 1988 with continuous enrollment. We determined the incidence of clinical events (per 1000 person-years) from January 2008 (when a new questionnaire on non-AIDS-related morbidity was introduced) through December 2010. Differences across age groups were analyzed using Cox regression, adjusted for CD4 cell count, viral load, sex, injection drug use, smoking, and years of HIV infection. RESULTS: Overall, 8444 (96%) of 8848 participants contributed data from 40,720 semiannual visits; 2233 individuals (26.4%) were aged 50-64 years, and 450 (5.3%) were aged ≥65 years. The median duration of HIV infection was 15.4 years (95% confidence interval [CI], 9.59-22.0 years); 23.2% had prior clinical AIDS. We observed 994 incident non-AIDS events in the reference period: 201 cases of bacterial pneumonia, 55 myocardial infarctions, 39 strokes, 70 cases of diabetes mellitus, 123 trauma-associated fractures, 37 fractures without adequate trauma, and 115 non-AIDS malignancies. Multivariable hazard ratios for stroke (17.7; CI, 7.06-44.5), myocardial infarction (5.89; 95% CI, 2.17-16.0), diabetes mellitus (3.75; 95% CI, 1.80-7.85), bone fractures without adequate trauma (10.5; 95% CI, 3.58-30.5), osteoporosis (9.13; 95% CI, 4.10-20.3), and non-AIDS-defining malignancies (6.88; 95% CI, 3.89-12.2) were elevated for persons aged ≥65 years. CONCLUSIONS: Comorbidity and multimorbidity because of non-AIDS diseases, particularly diabetes mellitus, cardiovascular disease, non-AIDS-defining malignancies, and osteoporosis, become more important in care of HIV-infected persons and increase with older age.

Missing value imputation in longitudinal measures of alcohol consumption

Attrition in longitudinal studies can lead to biased results. The study is motivated by the unexpected observation that alcohol consumption decreased despite increased availability, which may be due to sample attrition of heavy drinkers. Several imputation methods have been proposed, but rarely compared in longitudinal studies of alcohol consumption. The imputation of consumption level measurements is computationally particularly challenging due to alcohol consumption being a semi-continuous variable (dichotomous drinking status and continuous volume among drinkers), and the non-normality of data in the continuous part. Data come from a longitudinal study in Denmark with four waves (2003-2006) and 1771 individuals at baseline. Five techniques for missing data are compared: Last value carried forward (LVCF) was used as a single, and Hotdeck, Heckman modelling, multivariate imputation by chained equations (MICE), and a Bayesian approach as multiple imputation methods. Predictive mean matching was used to account for non-normality, where instead of imputing regression estimates, "real" observed values from similar cases are imputed. Methods were also compared by means of a simulated dataset. The simulation showed that the Bayesian approach yielded the most unbiased estimates for imputation. The finding of no increase in consumption levels despite a higher availability remained unaltered. Copyright (C) 2011 John Wiley & Sons, Ltd.