Vascular risk factors for depression and apathy | Part II 86 Determinants Vascular risk factors - During a standardized medical examination at the 6-year followup 33 25, we assessed smoking status and measured systolic and diastolic blood pressure (mmHg) twice in a supine position using an electronic Omron sphygmomanometer. Length (cm) and weight (kg) were measured to calculate the body-mass index (BMI) as a measure of obesity. Doppler assessment of ankle and blood pressure allowed calculation of the ankle-brachial index (ABI)34 as a measure of atherosclerosis 35. Vascular diseases - The presence of diabetes mellitus, and past/current history of stroke and cardiac disease were derived from the answers to the self-report questions of the CBS/LASA-questionnaire (NESDO) 36 and on information provided by the primary-care physician (NESDA) at the 6-year follow-up. Covariates The following covariates were controlled for as these might confound the association between vascular risk factors and diseases and apathy. Age, sex, and highest level of education were documented at the baseline interview. The highest level of education was categorized as basic, intermediate and high. The use of antipsychotics, antidepressants and benzodiazepines was asked during the interview and checked for by inspection of medication containers. Handgrip strength (kg) was assessed with a dynamometer as indicator of physical performance 37. Finally, the mood factor (see above) was included as a covariate to adjust for residual (and overlap of apathy with) depression. Statistical analyses Multiple linear regression models were built for the SAS sum score as well as for both apathy factors (dependent variables) separately. All vascular risk factors and vascular diseases were examined as cerebrovascular-related independent variables in separate models, adjusted for covariates described above including the level of residual depressive symptoms (based on either the IDS or the mood factor identified by the PCA). Missing values on any of the covariates were replaced by their means. To adjust for multiple testing, we considered p<.01 as statistically significant. Since the vascular apathy hypothesis is primarily based on findings in older persons, we post-hoc examined the influence of age on associations by introducing interaction factors between age and any of the vascular factors/diseases; and by repeating analyses in participants of 50 years of age and above and of 70 years of age and above. Two sensitivity analyses were carried out. A first set of sensitivity analyses were conducted by not replacing missing data on covariates by their mean.
RkJQdWJsaXNoZXIy MjY0ODMw