Characteristics of Study Population
We excluded 34 participants with missing baseline data, 1782 having baseline systolic blood pressure within BP lowering target and 35 having less than 2 blood pressure measurements during 0-4 month. Finally, 2882 participants were included into the present study (Supplemental Figure S1). The average baseline age of overall study population was 63.0±6.8 years old, 1400 (48.6%) participants were female, 691 (24.0%) were black race and 974 (33.8%) participants had cardiovascular disease at baseline.
During the exposure time (0-4 month), the mean systolic blood pressure achieved was 139.6±13.3 mm Hg in standard group and 131.2±11.9 mm Hg in the intensive group (P<0.001), and the systolic blood pressure variability was 12.3±6.9 mm Hg for the standard arm and 13.6±5.5 mm Hg for the intensive arm (P<0.001). Meanwhile, the systolic blood pressure TTR was 45.9%±34.4% in the standard group and 45.6%±29.7% in the intensive group (P=0.76).
Compared with the lowest TTR group, the participants in the highest TTR group were younger, received higher education (P=0.021), and had lower level of baseline glycated hemoglobin, low-density lipoprotein cholesterol and blood pressure and higher level of estimated glomerular filtration rate (P≤0.040). The proportion of black race were higher in participants with TTR between 75% to 100% than those of 0 to 25% (P<0.001, Table 1). The frequency of baseline cardiovascular disease, smokers and intensive glycemic therapy did not show difference across TTR groups (Table 1). The relationship between TTR groups and mean systolic blood pressure was shown in Figure 1, which presented greater TTR was correlated with higher proportion of blood pressure within target range (P<0.001).
Associations of Systolic Blood Pressure Time in Target Range and Outcomes
The primary outcome occurred in 280 participants with median follow-up of 4.96 years (incidence rate: 20.65 per 1000 person-years, 95% CI: 18.30-23.21). Systolic blood pressure TTR was significantly associated with 20.0% decreased risk of first primary events per 1-SD increase in unadjusted model (HR 0.80, 95% CI: 0.71-0.90, P<0.001). In both two adjusted models, systolic blood pressure TTR was also significantly associated with primary outcomes (HR 0.81, 95% CI: 0.72-0.91, P<0.001; HR 0.83, 95% CI: 0.74-0.94, P=0.0026, Table 2).
The associations of systolic blood pressure TTR and secondary outcomes were also assessed in the study. In fully adjusted models, systolic blood pressure TTR was significantly associated with all-cause mortality (HR 0.83, 95% CI: 0.72-0.97, P=0.018), cardiovascular death (HR 0.70, 95% CI: 0.54-0.89, P=0.0045) and nonfatal myocardial infarction (HR 0.85, 95% CI: 0.73-0.99, P=0.034) (Table 2).
The relationships among systolic blood pressure TTR and outcomes did not significantly differ between the standard and intensive therapy (P for interaction ≥0.086, Supplemental Table S1-4). In fully adjusted models, systolic blood pressure TTR was significantly associated with primary outcome, and all-cause and cardiovascular mortality (P≤0.042) in the standard arm, no matter whether systolic blood pressure average and variability were included in the regression models (Supplemental Table S1-4).
To estimate the independency of the associations between systolic blood pressure TTR and outcomes, we further adjusted systolic blood pressure average and variability during exposure time based on full models. Results showed that systolic blood pressure TTR was independently associated with primary outcome (P≤0.012), all-cause mortality (P≤0.018) and cardiovascular death (P≤0.022) when systolic blood pressure average and variability were included in the regression models (Table 3).
Fully adjusted models were also conducted for systolic blood pressure TTR groups stratified by the threshold of 25%, 50% and 75%. The participants of systolic blood pressure TTR of 75% to 100% had significantly decreased risk of primary outcome (HR 0.57, 95% CI: 0.41-0.80, P=0.0012), all-cause mortality (HR 0.62, 95% CI: 0.41-0.95, P=0.027), cardiovascular death (HR 0.36, 95% CI: 0.17-0.76, P=0.0071) and nonfatal myocardial infarction (HR 0.54, 95% CI: 0.35-0.82, P=0.0042). A linear trend was found for primary outcome, all-cause mortality, cardiovascular death and nonfatal myocardial infarction across TTR groups (All P for trend ≤0.025). When further adjusted with systolic blood pressure average and variability, the highest systolic blood pressure TTR group was also significantly associated with decreased risk of various outcomes (P≤0.027 for all), as similar as the results in full models (Table 4).
In order to better understand the relationship of systolic blood pressure TTR and outcomes, restricted cubic spline was further conducted (Figure 2). The associations between TTR and outcomes showed linear relationships and no nonlinearity was detected in the analysis (P≥0.23 for all).
Subgroup Analysis of Systolic Blood Pressure Time in Target Range and Outcomes
For the further assessment of the prognostic value of systolic blood pressure TTR, similar analyses were conducted in participants with different characteristics (Supplemental Table S5). The interactions between subgroups and systolic blood pressure TTR were also detected. Though the correlation of systolic blood pressure TTR and outcomes did not significantly differ across age and gender groups (P for interaction ≥0.086), there was a significant interaction of systolic blood pressure TTR on primary outcome and total stroke between different BMI groups (BMI ≤30 kg/m2 vs BMI >30 kg/m2; primary outcome: HR, 0.69 vs 0.94, P for interaction=0.022; total stroke: HR, 0.55 vs 1.01; P for interaction=0.022), indicating better protective effect of systolic blood pressure TTR in thinner participants.
The distribution of systolic blood pressure TTR in different blood pressure control groups was shown in Supplemental Figure S2. Among participants whose mean systolic blood pressure levels beyond target range, 1-SD increase of systolic blood pressure TTR was correlated with 20% and 41% decreased risk of primary outcome and cardiovascular death (HR, 0.80, 95% CI: 0.66-0.99, P=0.035; HR, 0.59, 95% CI: 0.37-0.93, P=0.024, respectively) in full model, and sustained significance when systolic blood pressure average and variability (P ≤0.018) were included into the models (Supplemental Table S6).
In sensitivity analysis, systolic blood pressure TTR was significantly associated with primary outcome (HR, 0.83, 95% CI: 0.74-0.94, P=0.0031), all-cause mortality (HR, 0.83, 95% CI: 0.71-0.96, P=0.015), chronic heart failure (HR, 0.79, 95% CI: 0.65-0.96, P=0.015), nonfatal myocardial infarction (HR, 0.85, 95% CI: 0.73-0.99, P=0.034) and total stroke (HR, 0.74, 95% CI: 0.57-0.95, P=0.020) in full models when exposure time expanded to 8th month (Supplemental Table S7), and it sustained the significant prognostic values of primary outcome (P≤0.017) and all-cause mortality (P≤0.022) after adjusting for systolic blood pressure average and variability (Supplemental Table S8).