The baseline characteristics of the 4405 participants were shown in Table 1. Within a mean follow-up of 12.2 years, 627 (14.2%) of the individuals progressed to at least one condition, of whom 304(48.5%) had diabetes, 162 (25.8%) had MI, and 242(38.6%) had stroke; 72(11.5%) had 2 or 3 of these conditions. At baseline, age, BMI and prevalence rate of taking antihypertensive drugs increased with SBP category.
Table 2 displays the association of SBP category with MI, stroke and diabetes. For the incidence of MI, compared with participants with SBP 80–119 mmHg, the multivariable adjusted HRs[95% CIs] were 1.20[0.79–1.81], 1.93[1.18–3.16], and 2.35[1.26–4.39] for MI in participants with SBP 120–139, 140–159, 160–249 mmHg, respectively; a significant positive trend in categorized SBP for incident MI(P for trend 0.001) was observed; each additional 10 mmHg of SBP was associated with a 16% higher probability of developing MI(HRs[95% CIs]: 1.16[1.07–1.27]). For the incidence of stroke, compared with participants with SBP 80–119 mmHg, the multivariable adjusted HRs[95% CIs] were 1.64[1.12–2.41], 3.12[2.03–4.81], and 5.65[3.47–9.20] for stroke in participants with SBP 120–139, 140–159, and 160–249 mmHg, respectively; a significant positive trend in categorized SBP for incident stroke(P for trend < 0.001) was observed; each additional 10 mmHg of SBP was associated with a 31% higher probability of developing stroke(HRs[95% CIs]: 1.31[1.23–1.40]). For the incidence of diabetes, compared with participants with SBP 80–119 mmHg, the multivariable adjusted HRs[95% CIs] were 1.48[1.10-2.00], 1.60[1.10–2.34], and 1.54[0.93–2.56] for diabetes in participants with SBP 120–139, 140–159, and 160–249 mmHg, respectively; a significant positive trend in categorized SBP for incident diabetes(P for trend 0.030) was observed; each additional 10 mmHg of SBP was associated with a 10% higher probability of developing diabetes(HRs[95% CIs]: 1.10[1.03–1.17]). Multivariable adjusted restricted cubic spline analyses showed a linear association of SBP with MI (P for linearity 0.001), stroke (P for linearity < 0.001), diabetes(P for linearity 0.005) and composite outcome(P for linearity < 0.001), respectively (Fig. 1).
In subgroup analysis, either SBP or DBP was more strongly associated with MI and stroke in young age group than elder age group(Fig. 2 and Fig. 3; Figure S2 and Figure s3 in appendix). For the incidence of diabetes, either SBP or DBP was more associated with diabetes in female participants, but not associated with diabetes in male participants(Fig. 4 and Figure s4 in appendix). The associations for DBP were broadly consistent with those for SBP: there were linear associations of usual DBP with stroke, diabetes and composite outcome, and J shape relation of DBP with MI throughout the BP range examined(50–120 mm Hg usual DBP)(Figure s1 in appendix). The strengths of these associations for 5 mmHg higher usual DBP were equivalent to about 10 mm Hg higher usual SBP(Figure s2, s3 and s4 in appendix).
Table 1
Baseline clinical characteristics of participants
Characteristics | Baseline SBP level(mmHg) |
80–119(n = 1835) | 120–139(n = 1800) | 140–159(n = 553) | 160–249(n = 217) |
Age(year) | 51.14 ± 8.33 | 54.08 ± 9.43 | 59.48 ± 9.92 | 60.46 ± 9.89 |
Gender | | | | |
female | 1075(58.58%) | 852(47.33%) | 275(49.73%) | 120(55.3%) |
male | 760(41.42%) | 948(52.67%) | 278(50.27%) | 97(44.7%) |
BMI(kg/m^2) | 22.35 ± 2.89 | 23.43 ± 3.19 | 24.20 ± 3.38 | 25.10 ± 4.14 |
DBP(mmHg) | 71.75 ± 8.11 | 81.53 ± 7.44 | 89.67 ± 9.64 | 98.98 ± 12.21 |
Antihypertensive drugs | | | | |
no | 1793(97.71%) | 1687(93.72%) | 458(82.82%) | 140(64.52%) |
yes | 15(0.82%) | 72(4%) | 87(15.73%) | 73(33.64%) |
Current smoker | | | | |
no | 1253(68.28%) | 1130(62.78%) | 373(67.45%) | 146(67.28%) |
yes | 552(30.08%) | 632(35.11%) | 171(30.92%) | 67(30.88%) |
Current drinker | | | | |
no | 1169(63.71%) | 1054(58.56%) | 351(63.47%) | 141(64.98%) |
yes | 614(33.46%) | 695(38.61%) | 185(33.45%) | 69(31.8%) |
Site | | | | |
urban | 558(30.41%) | 617(34.28%) | 200(36.17%) | 68(31.34%) |
rural | 1277(69.59%) | 1183(65.72%) | 353(63.83%) | 149(68.66%) |
Follw-up(year) | 12.57 ± 3.54 | 11.64 ± 3.95 | 10.53 ± 4.31 | 9.56 ± 4.64 |
Composite outcome | | | | |
no | 1688(91.99%) | 1542(85.67%) | 412(74.5%) | 136(62.67%) |
yes | 147(8.01%) | 258(14.33%) | 141(25.5%) | 81(37.33%) |
Number of individuals across categories may not sum to the given number because of missing data. |
Table 2
Associations of SBP category with MI, stroke and diabetes
| MI | stroke | diabetes |
HR[95%CI] | P | HR[95%CI] | P | HR[95%CI] | P |
SBP(mmHg) | | | | | | |
80–119 | 1.00 | | 1.00 | | 1.00 | |
120–139 | 1.20[0.79–1.81] | 0.388 | 1.64[1.12–2.41] | 0.011 | 1.48[1.10-2.00] | 0.010 |
140–159 | 1.93[1.18–3.16] | 0.009 | 3.12[2.03–4.81] | < .001 | 1.60[1.10–2.34] | 0.014 |
160–249 | 2.35[1.26–4.39] | 0.007 | 5.65[3.47–9.20] | < .001 | 1.54[0.93–2.56] | 0.094 |
P for trend | | 0.001 | | < .001 | | 0.030 |
10mmHg increase | 1.16[1.07–1.27] | 0.001 | 1.31[1.23–1.40] | < .001 | 1.10[1.03–1.17] | 0.006 |
All models were adjusted for age, gender, BMI, antihypertensive drugs, smoking status, drinking status, and site. |