M. Safwan Badr

Abstract Objectives We hypothesized that benzodiazepine/z-drug (BZRA) dose, insomnia symptoms, and obstructive sleep apnea (OSA) risk are associated with cognitive performance above and beyond other modifiable dementia risk factors. Methods In this cross-sectional study, we used baseline data from recruited patients without dementia from one VA site and one academic site to analyze regression models based on previously published risk factors for cognitive impairment, adapted to include sleep and benzodiazepine usage interactions. Changes in R 2 and model fit (F-tests) were examined for four nested regression models predicting Mini-Mental State Examination (MMSE), Trail Making Test (TMT)-A, TMT-B, and Digit Symbol Substitution Test (DSST). Base Models included common dementia risk factors. Model-2 = Base + Insomnia Severity Index; Model-3 = Model-2 + dose; Model-4 = Model-3 + OSA risk; and Model-5 = Model-4 + OSA-BZRA. Results R 2 for the base model predicting MMSE was 0.139 (n = 313), 0.241 for TMT-A (n = 298), 0.229 for TMT-B (n = 287), and 0.189 for DSST (n = 285). Model-5 R 2 was 0.142 for MMSE, 0.246 for TMT-A, 0.238 for TMT-B, and 0.190 for DSST, with all F-test p-values >.145. Conclusions Sleep disorder variables did not significantly explain additional variation in cognition performance. Studies with a wider range of cognitive performance are needed.