P3-17. Hypertension and low BMI are associated to mortality in well treated HIV-positive subjects above 50 years: Preliminary findings at 4, 5 years follow up
Göran Bratt , Catharina Missalidis , Fabian Söderdahl  and Anders Blaxhult 
Affiliates:  Venhälsan, Department of Infectious Disease, Södersjukhuset, Stockholm.  Statisticon, Uppsala.
Despite a marked reduction in mortality and morbidity among HIV-positive persons on effective antiretroviral treatment (ART) co-morbidity is frequent and may predict the long-term prognosis. The aim of this study was to examine factors that might influence morbidity and mortality in an aging well treated HIV infected population.
Between 1.1 2012 and 1.1 2013 all HIV-positive patients born 1964 or earlier were included in a study of co-morbidity. The cohort included 570 patients: 12% females; 88% males. Median time with known HIV was 17 years, 20% had had an AIDS diagnosis and 54% a CD4 nadir <200. 98% were on ART. The median CD4 was 600 and 94% had RNA<50 cop/ml. The data at an average of 4,4 years follow up has now been analyzed. Mortality was assessed using Cox regression and the development of cancer, severe cardiovascular disease and diabetes was analyzed using uni and multivariate logistic regression. All analyses were controlled for age, sex and a history of AIDS.
5% (n=29) died during follow up. The standardized mortality ratio was 1.77 indicating that there were approximately 77 % more deaths than expected with regards to the general population. Associated factors in univariate analysis were: Framingham score (p<0,005). hypertension (p<0,005) and a low BMI (p <0,05). In multivariate analysis, only hypertension (p<0,05) and a low BMI (p <0,05) were significant.
Non-AIDS, non-skin cancer had been diagnosed in 9% in 2012/2013. A new non-AIDS, non-skin cancer developed in 4% (n=24) during follow up. Associated factors in univariate analysis were: HIV-months (p<0,05), months without ART (p<0,05), increased HOMA or diabetes (p<0,05) and increased triglycerides (p<0,05). In multivariate analysis, only time with HIV (p<0,05) was significant.
Hypertension was present in 43% in 2012/2013. Severe cardiovascular disease (myocardial infarction, unstable angina, repeated TIA, stroke or severe peripheral arterial disease) had occurred in 16% after HIV-diagnosis up to 2012/2013 and further 3% (n=17) were diagnosed with one or more of these events during this follow up. Associated factors in univariate analysis were time with CD4<200 (p<0,05), increased triglycerides (p<0,005), increased total cholesterol ((p<0,05) and increased hsCRP (p<0,01). In multivariate analysis, only increased triglycerides (p<0,01) and increased hsCRP (p<0,05) were significant.
11% had diabetes in 2012/2013 and a further 4% (n=21) developed diabetes during this follow up Associated factors in univariate analysis were: increased BMI (p<0,001), increased triglycerides (p<0,05), increased HOMA (p<0,005) and central obesity (p<0,005). In multivariate analysis, only increased triglycerides (p<0,05) and increased BMI ( 50 years shows a continued increased mortality rate. Hypertension and low BMI were particularly associated with mortality. In this study with limited numbers of events, only the duration of the HIV infection was related to cancer development and low grade inflammation and increased triglycerides related to severe cardiovascular disease. Surprisingly, smoking was not a significant factor in these analyses. The development of diabetes was associated with increased BMI and triglycerides. Hypertension, over and underweight and hyperlipidemia should receive particular attention in patients with well controlled HIV-infections.