Preplanned Studies: Impacts of PM_2.5 on Ambulatory Blood Pressure Monitoring Indicators Attenuated by Blood Pressure Control Status and Treatment — Two Cities and Two Municipalities, China, 2017−2019

High blood pressure (BP) is the leading risk factor for cardiovascular disease (CVD) in China. Previous studies from either lowly or highly polluted regions have revealed that increased fine particulate matter (particles with aerodynamic diameter ≤2.5 µm; PM_2.5) exposure was associated with elevated BP levels and higher risk of hypertension (1). In recent years, using personal portable devices to record individual PM_2.5 exposure and ambulatory BP monitoring (ABPM) to measure BP levels periodically, panel studies also identified the relationship of short-term PM_2.5 exposure with hourly, 24-hour, daytime, and nighttime average BP levels (2-3). However, few studies assessed the adverse impacts of PM_2.5 with other ABPM indicators, such as BP variability, circadian rhythm of BP, and BP load, and the modifying effects of BP control status and antihypertensive therapy were also unclear. To clarify these issues, these associations were assessed using a multicity and municipality panel study conducted from 2017 to 2019 in China. BP levels and BP load elevation were associated with short-term PM_2.5 exposure. Further stratified analyses demonstrated that the adverse impacts of PM_2.5 on BP variability, circadian rhythm of BP, and BP load were attenuated among patients with well-controlled BP or taking angiotensin receptor blockers (ARBs), which provided hints about the prevention of PM_2.5-related adverse outcomes.

This study used data from a three-phase (winter, summer, spring/autumn) panel study conducted in Beijing, Shanghai, Wuhan, and Xi’an from 2017 to 2019. This panel study recruited participants at intermediate-to-high risk of CVD who were defined as having prehypertension or hypertension combined with at least 1 of the 3 conditions (i.e., central obesity, diabetes mellitus, and dyslipidemia) from the community (2). Questionnaires were administrated to collect demographic characteristics, lifestyle and risk factors, medication uses, etc. Each participant carried a portable monitor to measure real-time individual PM_2.5 from the first day to the third day during each phase, and a 24-hour ABPM was scheduled on the third day. This study was approved by the Institutional Review Board of Fuwai Hospital in Beijing, and written informed consent was obtained from all participants before data collection.

This study assessed the association of short-term PM_2.5 exposure with several ABPM indicators, including 24-hour, daytime, and nighttime average systolic BP (SBP) and diastolic BP (DBP), SBP and DBP variability (i.e., the standard deviation of BP), SBP and DBP load [i.e. the percentage of the counts for BP readings beyond certain threshold (130/80 mmHg, 135/85 mmHg, and 120/70 mmHg for 24-hour, daytime and nighttime, respectively)], and the percentage of nocturnal BP decline [i.e., (daytime BP-nighttime BP)/daytime BP×100%]. In addition, three methods were used to calculate morning BP surge, including morning BP (i.e., the average BP during the first 2 hours after waking), pre-waking BP surge (i.e., the mean BP during the 2 hours after waking minus the mean BP during the 2 hours before waking), and sleep-trough BP surge (the mean BP during the 2 hours after waking minus the mean BP of the lowest BP and the BP before and after the lowest one during sleeping). The exposure windows of PM_2.5 included the concurrent day with ABPM (lag0d), lag 1 day (lag1d), lag 2 days (lag2d), and moving average of previous 2 days (MA2d). Estimated changes and 95% confidence intervals (CIs) with per interquartile range (IQR) increment of PM_2.5 exposure (41.96 μg/m³) were calculated using a mixed linear model adjusted for age, gender, current alcohol drinking status, ideal level of physical activity, body mass index (BMI), antihypertensive medication uses, diabetes mellitus, dyslipidemia, and natural splines with 3 degrees of freedom for the daily average of personal environmental temperature and relative humidity at the same exposure windows as PM_2.5. City-specific and subject-specific random intercepts were used to account for the variations in different cities and within-subjects correlations. Stratified analysis of BP control status (hypertensive patients with SBP/DBP of less than 140/90 mmHg were defined as individuals with controlled BP, otherwise as uncontrolled BP), and for those with uncontrolled BP, a further stratified analysis of ARB uses was conducted. Z tests were conducted to explore the group difference. All analyses were performed using the SAS software (version 9.4; SAS Institute, Cary, NC, USA). A 2-side P of <0.05 was demonstrated statistical significance.

After excluding participants attending only one visit, those missing ABPM data or not fulfilling the valid criteria of ABPM data and prehypertensive patients, a total of 277 hypertensive patients with 802 visits were included for analysis. At baseline, the average age was 59.1±8.5 years and 113 (40.8%) participants were men. The mean office SBP and DBP were 134.5 mmHg and 80.4 mmHg, respectively. Overall, 91.7% patients were taking antihypertensive medications, with 33.2% being taking ARBs (Table 1). Across the three phases, a total of 557 (69.5%) visits reached well-controlled BP status. Among those without controlled BP, 65 (26.5%) visits had been taking ARBs. Generally, short-term PM_2.5 exposure was associated with BP levels and BP load; for example, the 24-hour, daytime, and nighttime average SBP increased by 0.68 mmHg (95% CI: 0.07–1.29), 0.82 mmHg (95% CI: 0.19–1.45) and 0.93 mmHg (95% CI: 0.17–1.70) per IQR increment of the lag2d PM_2.5 concentration (Figure 1A). The 24-hour SBP load and DBP load increased by 1.28% (95% CI: 0.01%–2.56%) and 1.38% (95% CI: 0.22%–2.54%), respectively, with per IQR increment of the lag1d PM_2.5 concentration (Figure 1C). However, short-term PM_2.5 exposure may not be associated with BP variability, morning BP surge, or the percentage of nocturnal BP decline (Figure 1B, 1D, 1E). When stratified by BP control status, the increment of 24-hour or daytime BP levels associated with PM_2.5 exposure tended to be higher among those with uncontrolled BP, but the difference was not statistically significant (Figure 2A). Those with uncontrolled BP were more prone to have higher morning BP surges than those with controlled BP. For example, per IQR increment of the lag2d PM_2.5 concentration was associated with 2.19 mmHg (95% CI: 0.47–3.91), 1.72 mmHg (95% CI: 0.20–3.24) and 1.47 mmHg (95% CI: -0.14–3.07) elevation for morning SBP, pre-waking SBP surge, and sleep-trough SBP surge, respectively, among patients without controlled BP, while no significant changes were noted in patients with controlled BP (all P<0.05) (Figure 2D). Among patients with uncontrolled BP, the association of PM_2.5 exposure with 24-hour, daytime SBP variability was modified by treatment with ARB. For example, 24-hour SBP variability increased by 0.49 mmHg (95% CI: -0.22–1.20) per IQR increment of lag0d PM_2.5 exposure among patients taking ARB, which was significantly higher than those not taking ARB (Figure 2G). In addition, patients taking ARB were more likely to have better circadian rhythms for DBP (Figure 2J).

Figure 1. 

Estimated changes and 95% CIs of BP levels (A), BP variability (B), BP load (C), morning BP surge (D) and the percentage of nocturnal BP decline (E) with per IQR (41.96 μg/m³) of PM_2.5 exposure among participants in two cities and two municipalities, China, 2017−2019

Note: Statistically significant changes were shown in red. Abbreviations: CI=confidence interval; BP=blood pressure; IQR=interquartile range; PM_2.5=particles with aerodynamic diameter ≤2.5 µm; SBP=systolic blood pressure; DBP=diastolic blood pressure; lag0d=current day with BP measurement; lag1d=lag of 1 day; lag2d=lag of 2 days; MA2d=the 2-day moving average of PM_2.5.

Figure 2. 

Estimated changes and 95% CIs of BP levels (A, F), BP variability (B, G), BP load (C, H), morning BP surge (D, I) and the percentage of nocturnal BP decline (E, J) with per IQR (41.96 μg/m³) of the PM_2.5 exposure, stratified by BP control status in overall population and by ARB use in patients without controlled BP in two cities and two municipalities, China, 2017−2019.

* P<0.05 for differences between two groups. Abbreviations: CI=confidence interval; BP=blood pressure; IQR=interquartile range; PM_2.5=particles with aerodynamic diameter ≤2.5 µm; SBP=systolic blood pressure; DBP=diastolic blood pressure; ARB=angiotensin receptor blockers; lag0d=current day with BP measurement; lag1d=lag of 1 day; lag2d=lag of 2 days; MA2d=the 2-day moving average of PM_2.5.