Continuous variables are reported as mean ± standard deviation or median [25th–75th interquartile range], as appropriate. Categorical variables are reported in frequency and percentages. Student t test or Mann–Whitney U test was used to compare normally and nonnormally distributed variables between participants with HIV and controls and chi-squared test was used to compare categorical variables.
The association of EF density with cardiovascular risk factors, as well as with HIV related parameters, was assessed using univariable and multivariable linear regression models, with and without inclusion of EF volume into the models. Multivariable model included all HIV-related parameters and cardiovascular risk factors risk factors significantly associated with EF density.
The association of EF density with coronary plaque (CAC, total, calcified, noncalcified and mixed plaque volume) was evaluated using zero-inflated Poisson regression, as previously described.[10 (link)] The use of zero-inflated model is made necessary by the excess of 0 in the distribution of total coronary plaque volumes. Coronary plaque volume variables were natural-log transformed prior to statistical analysis. Multivariable models were adjusted for CVD risk factors.
In addition and as an exploratory analysis, the association of EF with inflammatory biomarkers was evaluated using univariable and multivariable linear regression models. Multivariable models included inflammatory biomarkers and CVD risk factors if they showed a univariable association with EF. As this analysis was hypothesis-generating, we did multiple tests and did not correct for multiple comparisons.
Effect modification by HIV of each association was assessed by inclusion of an interaction term to the fully adjusted models, and significance of the interaction term was assessed by a likelihood ratio test.
For patients with incomplete data, the mean or median value was used to impute the missing data. Values for the following number of participants were missing and imputed: Body mass index (BMI),[5 (link)] smoking exposure,[14 (link)] high density lipoprotein-cholesterol,[4 (link)] low density lipoprotein-cholesterol,[9 (link)] antiretroviral therapy (ART) exposure duration,[7 (link)] non-nucleoside reverse transcriptase inhibitors exposure duration.[7 (link)] A P value <.05 was considered statistically significant. Statistical analyses were performed using R (version 3.3.2; R Foundation for Statistical Computing, Vienna, Austria).
The association of EF density with cardiovascular risk factors, as well as with HIV related parameters, was assessed using univariable and multivariable linear regression models, with and without inclusion of EF volume into the models. Multivariable model included all HIV-related parameters and cardiovascular risk factors risk factors significantly associated with EF density.
The association of EF density with coronary plaque (CAC, total, calcified, noncalcified and mixed plaque volume) was evaluated using zero-inflated Poisson regression, as previously described.[10 (link)] The use of zero-inflated model is made necessary by the excess of 0 in the distribution of total coronary plaque volumes. Coronary plaque volume variables were natural-log transformed prior to statistical analysis. Multivariable models were adjusted for CVD risk factors.
In addition and as an exploratory analysis, the association of EF with inflammatory biomarkers was evaluated using univariable and multivariable linear regression models. Multivariable models included inflammatory biomarkers and CVD risk factors if they showed a univariable association with EF. As this analysis was hypothesis-generating, we did multiple tests and did not correct for multiple comparisons.
Effect modification by HIV of each association was assessed by inclusion of an interaction term to the fully adjusted models, and significance of the interaction term was assessed by a likelihood ratio test.
For patients with incomplete data, the mean or median value was used to impute the missing data. Values for the following number of participants were missing and imputed: Body mass index (BMI),[5 (link)] smoking exposure,[14 (link)] high density lipoprotein-cholesterol,[4 (link)] low density lipoprotein-cholesterol,[9 (link)] antiretroviral therapy (ART) exposure duration,[7 (link)] non-nucleoside reverse transcriptase inhibitors exposure duration.[7 (link)] A P value <.05 was considered statistically significant. Statistical analyses were performed using R (version 3.3.2; R Foundation for Statistical Computing, Vienna, Austria).
Free full text: Click here
