Patients on hydroxychloroquine had significantly lower blood pressure variability.
The investigators measured disease activity with the SELENA revision of the Safety of Estrogens in Lupus Erythematosus National Assessment–Systemic Lupus Erythematosus Disease Activity Index instrument score and the Physician Global Assessment. At each visit, they measured blood pressure using a Carescape Dinamap V100 monitor calibrated once every 12 months according to manufacturer guidelines with patients seated and the arm supported at heart level.
The cohort was divided into subsets of patients, with each subset approximately 92% female. Most (60%) were younger than age 40 at entry to the cohort, although a small proportion (6%) was 60 or older. Most were Caucasian (55%) or African American (38%).
Mean systolic and diastolic blood pressure measures were estimated using random intercept models, fit by restricted maximum likelihood. Using this approach, the investigators got estimates of the mean blood pressure, the within-person standard deviation (i.e., the standard deviation of an individual’s blood pressure around their personal mean) and the between-person standard deviation (i.e., the standard deviation of the person-specific mean blood pressures). They compared means with national data collected from 2001–2008 by analyzing data from 22,672 clinic visits of 1,509 cohort members between 2001 and 2008.
The researchers used data from the National Health Statistics Reports (19,921 adults aged 18 and over with blood pressure estimates calculated using the mean of up to three measurements) to compare the estimated means to the mean systolic and diastolic pressures in the general population.7 They used a larger sample, including data from 52,791 cohort visits of 2,128 patients seen from 1987 to 2013, to obtain more precise estimates of the means and variances of blood pressure by age.
To analyze the relationship between clinical and demographic characteristics and blood pressure variability, the researchers used random intercept models to estimate within-patient and between-patient BPV. They used likelihood ratio tests to determine the statistical significance of differences between clinical subgroups with respect to BPV. The analysis was based on 63,890 clinic visits of 2,525 cohort members from 1987 to 2018.
The researchers analyzed the relationship between blood pressure parameters and cardiovascular events based on 1,340 cohort members who had at least eight clinical assessments of blood pressure in the cohort between 1987 and 2013. Ninety-two percent of the cohort members were women, and the average duration of follow-up was 6.2 years. The researchers defined cardiovascular events as stroke, myocardial infarction, incident angina, a coronary procedure (coronary artery bypass graft surgery or percutaneous coronary intervention) or claudication. Considering only the first cardiovascular event for each person, 105 events occurred.
The researchers aggregated person-months and calculated the risks of a cardiovascular event by monthly characteristics. They excluded person-months after a previous cardiovascular event. For each month of follow-up for a patient, the previous eight blood pressure measurements were included in the analysis. The following variables were calculated based on those most recent past eight measures, and the information was later linked to whether the patient experienced a cardiovascular event that month:
