Blood pressure data analysis

A total of 2 164 321 individuals in Japan registered with the WellnessLINK system as of the end of September 2015. Among these users, 110 454 sent BP data to the system automatically from BP measurement devices, or manually. Self-reported data that were manually inputted by users were excluded from the data analysis in this study. In this report, we analysed 41.7 million BP measurement data from 64 536 registered users (51 335 men, 13 201 women; mean age 52.9 years) who sent their measured BP data automatically. Among these users, 24% were treated with antihypertensive medication.

BP values measured at home between 04:00 and 09:59 were defined as ‘morning BP’, while those measured at home between 19:00 and 01:59 were defined as ‘evening BP’. Individual mean morning BP values were calculated using the first measured BP value in the morning, and those measured during the following 10 minutes. Individual mean evening BP values were calculated using the last measured BP value in the evening, and the values measured during the previous 10 minutes. The monthly mean home BP values were calculated by averaging individuals’ home BP values for all participants.

Descriptive statistical values such as mean monthly systolic BP (SBP) and diastolic BP (DBP) (and standard errors) were calculated by sex and age group for 2 years, from September 2013 to September 2015. Monthly mean BP changes for 2 years are shown in figures along with mean outdoor temperatures. Inter-month differences in BP were assessed using analysis of variance. The t test for unpaired observations was used to determine the significance of the difference, after Bonferroni correction for multiple comparisons. Multiple regression analysis was preformed to assess whether the association between temperature and BP is independent of age (continuous variable), gender (dichotomous; men: 1, women: 0) and geological location (Northern Japan/Eastern Japan/Central Japan/Kansai region/Western Japan). Four dummy variables for five geological locations were assigned and Eastern Japan was defined as the reference area. All P values were two sided, and P values < 0.05 were considered to be statistically significant. Mean outdoor temperatures were calculated using outdoor temperatures at the time of each measurement in the capital city of the prefecture in which each participant lived, obtained from the database of the Japan Meteorological Agency.

The participants of this study had voluntarily filled out the online consent form, agreed to send their data to the server managed by OMRON Healthcare Co., Ltd. using the WellnessLINK system service, and agreed with the condition that the data obtained by the WellnessLINK system would be analysed for medical research purposes after data anonymisation. The ethics committee of the Omron Healthcare Co., Ltd. approved the study protocol.