To investigate the reliability of self-measured home blood pressure (BP) obtained with wrist devices.Design and method:
Office and home upper-arm and wrist BP was obtained in 597 representative subjects from general population using two validated devices. Parti-cipants received an individual 20-min training in the office and were asked to repeat the upper-arm and wrist measurements at home for 7 consecutive days using the same procedures. In particular, they were recommended to keep the wrist at heart level during wrist BP measurement. The difference between upper-arm BP and wrist BP measured by the observer in the office was calculated for each individual and compared with the 7-day averaged self-measured upper-arm - wrist BP difference. Anthropometric measures and a complete cognitive assessment were also obtained.Results:
In the office, under observer supervision, after adjustment for confounders, wrist BP was lower than upper-arm BP (systolic −2.4%, p < 0.0001; diastolic −0.6%, NS), while at home BP was higher at wrist than at upper-arm (systolic +5.9%, p < 0.0001; diastolic +5.6%, p < 0.0001). Both systolic and diastolic upper-arm - wrist BP differences obtained in office significantly differed from the upper-arm - wrist differences obtained by the participant at home (systolic 3.0 ± 13.7 vs. −6.7 ± 0.1 mmHg, p < 0.0001; diastolic 0.6 ± 9.4 vs. −4.0 ± 6.6 mmHg, p < 0.0001). Office (r = 0.37, p < 0.0001) but not home (r = −0.05, p = 0.9) wrist systolic BP correlated with office BP, the two regression coefficients differing significantly (p < 0.0001). Similar results were obtained for diastolic BP. In multivariable regression analyses, using the difference between the upper-arm - wrist BP discrepancies obtained in the office and at home as the dependent variable, forearm length (p < 0.06 for systolic, p < 0.04 for diastolic) and praxic abilities (p < 0.008, p < 0.0001, respectively) were independent determinants of the office-home difference.Conclusions:
These data indicate that in spite of adequate training wrist BP measurement at home is unreliable. This is likely due to a cognitive (particularly praxic) deficit leading to incorrect forearm position, a problem that, for hydraulic reasons, is magnified in people with longer forearm. Wrist BP measurement should be discouraged chiefly in subjects presumed to have impaired cognitive function.