The variability of blood pressure (BPV) has been suggested as a clinical indicator for cognitive dysfunction, yet the results from clinical studies are variable. This study investigated the relationship between BPV and the risk of cognitive decline or dementia. Bibliographic databases, including PubMed, Scopus, and Embase, were searched systematically for longitudinal cohort studies with BPV measurements and neuropsychological examinations or dementia diagnosis. A traditional meta‐analysis with subgroup analysis, and a further dose‐response meta‐analysis were conducted. Twenty cohort studies with 7 924 168 persons were included in this review. The results showed that a higher systolic BPV (SBPV), when measured with the coefficient of variation (SBP‐CV) or standard deviation (SBP‐SD), was associated with a higher risk of all‐cause dementia diagnosis but not incidence of cognitive decline on neuropsychological examinations. In subgroup analysis, the effect was more prominent when using BPV of shorter timeframes, during shorter follow‐ups, or among the elderly aged more than 65 years. No dose‐response relationship could be found. Our study suggested possible positive associations between SBPV and the risk of dementia. Further studies are required to validate these findings.

To the Editor, we appreciate the comments from Tully and de Heus and would like to respond to their concerns. As widely recognized, systematic review and meta-analysis, due to its timely summary and analysis through comprehensive search for the best available research, is an useful research strategy to summarize the totality of evidence. 1The a priori registration of protocols for a systematic review and metaanalysis is recommended by the reporting guideline of PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses).

Objective:An accurate blood pressure measurement (BPM) is important for diagnosing and managing hypertension. Several monitoring techniques of BPM, including office BPM (OBPM), ambulatory BPM (ABPM), and home BPM (HBPM) have been widely utilized; however, the corresponding agreement and conversion between these modalities remained elusive. With emerging evidence suggesting the utility of out-of-office blood pressure (BP) for hypertension management, an investigation into the relationship between BPMs is crucial. Therefore, we performed a systematic review and network meta-analysis from published studies and compared the reported BP values between different BPM modalities.Design and method:We searched MEDLINE and Scopus for observational studies and andomised controlled trials reporting more than one type of BPM by Apr 6, 2021. Studies with participants that were over 18 years old and received at least two different BPM within a week were included. The risk of bias of included studies was assessed by the quality of the study and BPM techniques. We calculated the mean differences (MD) between modalities and pooled the measurements using random-effects network meta-analysis. Furthermore, the agreement between BPMs within different BP levels was investigated using network meta-regression.Results:We identified 10970 studies and included 486/482 studies (467654/466167 individuals) for SBP/DBP measurements. With reference to OBPM, the MDs of SBP of casual OBPM, HBPM, 24 hours ABPM, daytime ABPM, and night time ABPM, were 2.8, -5.5, -10.0, -6.4, -19.5mmHg, respectively. For DBP the MDs were 1.3, -3.0, -6.3, -3.19, and -13.76mmHg. Network meta-regression (see figure) showed that the MDs between OBPM and casual OBPM were rather fixed; however, those between OBPM/HBPM or OBPM/ABPM followed a negatively associated linear relationship across BP levels. The diagnostic thresholds of different BPMs corresponding to OBPM 140/90mmHg and 130/80mmHg were generally in agreement with current guidelines.Conclusions:Our study demonstrated the agreement between common BPM modalities by synthesizing the body of evidence from a comprehensive search. Furthermore, the regression model showed that the MDs of measurements between modalities may vary depending on the mean BP. Our systematic review supports the diagnostic thresholds for different BPM recommended by current guidelines.

No abstract