Author response: Association of orthostatic hypotension with incident dementia, stroke, and cognitive decline

We thank Dr. Young for his insightful comment. While there is a biologically plausible mechanism by which orthostatic hypotension (OH) may affect cognition, we agree that our results also support the idea that OH may operate primarily through vascular processes and may simply be a biomarker of elevated vascular risk; this is supported by our results that OH appeared detrimental in persons with hypertension and diabetes, as noted in the Discussion.¹

As Dr. Young mentions, OH is strongly associated with other cardiovascular risk factors, including smoking, diabetes, and hypertension—factors that are also highly prevalent in this population of middle-aged adults. In our sample,¹ 3,012 people (76 with OH) were never smokers with no diabetes and no hypertension at baseline (25% of our original sample). We ran our analysis of incident dementia in this population, yielding an adjusted hazard ratio of 1.28 (95% confidence interval 0.67, 2.42). While we likely have limited power in this group, the magnitude of this hazard ratio suggests some association of OH with dementia, independent of cardiovascular risk factors. Sample sizes in this group were too small to examine 20-year cognitive decline; we agree that further studies are needed to clarify the causal nature of these associations.

We also thank Dr. Kawada for the comment. We propose several explanations for the discrepancy in the association between OH and dementia/cognitive decline.

First, while the cognitive decline results were not statistically significant, the magnitude of the association was relatively large and similar in size to that seen in the Atherosclerosis Risk in Communities (ARIC) study from hypertension.² Given the small number of persons with OH at baseline (5%) or the fact that few (21%) with OH attended the last follow-up examination, our power may be limited to find significant differences. Alternatively, the discrepancy between dementia and cognitive decline could be due to floor effects. Persons with OH had lower cognitive scores at baseline, with less room to decline, but may be most likely to cross the threshold for dementia. We agree that long-term prospective studies may clarify these associations.

Second, we agree that our single measurement of OH may represent repeated exposure to cerebral hypoperfusion, which could result in the cascade of events leading to cognitive dysfunction.

Third, we agree with Dr. Kawada's comment on further exploring racial differences, since risk factor prevalence and control differs across race and ethnicities.^3,4 In ARIC, we are unable to examine associations in persons of Asian descent.