Abstract
Background:
The recent American College of Cardiology/American Heart Association guidelines for high blood pressure lowered the hypertension criteria from systolic/diastolic blood pressure (SBP/DBP) of 140/90 mmHg or greater to 130/80 mmHg or greater, while the potential impact of the change on Chinese adults remains unclear.
Design:
A pooled prospective cohort analysis.
Methods:
Included were 154,407 Chinese adults from three prospective cohorts, which measured blood pressure at baseline and follow-up visits, and tracked death events by linkages to medical insurance system or vital statistics registries. Cox regression models were used to estimate the hazard ratios (HRs) and 95% confidence intervals (CIs).
Results:
During a total follow-up of 1,718,089 person-years, 14,692 deaths were documented including 5086 cardiovascular deaths (1277 ischaemic heart disease and 2509 cerebrovascular disease deaths). Compared to normal blood pressure (SBP/DBP < 120/80 mmHg), newly defined stage I hypertension (SBP/DBP 130–139/80–89 mmHg) was associated with increased cardiovascular mortality (HR 1.40, 95% CI 1.16–1.69; HR 1.36, 95% CI 1.12–1.65 for ischaemic heart disease mortality; HR 1.53, 95% CI 1.18–2.00 for cerebrovascular mortality), but not with all-cause mortality (HR 1.04, 95% CI 0.89–1.21). Stage 2 hypertension (SBP/DBP ≥ 140/90 mmHg) showed significant associations with cardiovascular disease and all-cause mortality, while elevated blood pressure (SBP 120–129 mmHg and DBP < 80 mmHg) showed null associations. The associations were stronger in adults younger than 65 years and adults without pre-existing cardiovascular disease compared with their counterparts (P for heterogeneity < 0.05).
Conclusions:
The newly defined stage 1 hypertension is associated with an increased risk of cardiovascular disease mortality in the Chinese population, particularly among younger adults and those without a history of cardiovascular disease.
Keywords: Blood pressure, hypertension, mortality, cardiovascular disease, cohort study, Chinese
Introduction
Cardiovascular disease (CVD) remains as the top leading causes of death and years of life lost, and ischaemic heart disease (IHD) and cerebrovascular disease (CBVD) are the main contributors to CVD death.1 High blood pressure (BP) is an important risk factor for CVD morbidity and mortality,2,3 and a meta-analysis of 61 prospective cohort studies has shown that usual BP was strongly and log-linearly associated with coronary heart disease (CHD) and stroke mortality with no apparent threshold down to a BP of 115/75 mmHg.3
In 2017, the American College of Cardiology (ACC)/American Heart Association (AHA) published a new guideline for the prevention, detection, evaluation, and management of high blood pressure in adults.4 In the new ACC/AHA guideline, BP is categorised into four groups, i.e. normal BP (systolic blood pressure (SBP) < 120 mmHg and diastolic blood pressure (DBP) < 80 mmHg), elevated BP (SBP 120–129 mmHg and DBP < 80 mmHg), stage 1 hypertension (SBP 130–139 mmHg or DBP 80–89 mmHg) and stage 2 hypertension (SBP/DBP ≥ 140/90 mmHg), with lower thresholds for hypertension than that defined by the Eighth Joint National Committee (JNC-8)5 or Chinese guidelines6 (SBP/DBP ≥ 140/90 mmHg for hypertension). A recent study reported that the prevalence rate increased from 49.7% based on the JNC-8 guidelines to 63.0% based on the 2017 ACC/AHA guidelines in US adults and from 38.0% to 55.0% in Chinese adults aged between 45 and 75 years.7 Because of its large population, lowering the criteria for hypertension definition in China would exert a greater burden for the management and treatment of hypertension in the primary healthcare system.
However, little is known about whether the criteria change would have a beneficial impact on the prevention of all-cause and CVD mortality. A recent meta-analysis with 17 trials and 55,163 patients reported that on-treatment SBP of less than 130 mmHg had an optimal balance between efficacy and safety, corroborating the recommendation by the new guideline.8 However, a study in Singapore Chinese revealed that the newly defined stage 1 hypertension was not associated with increased CVD mortality, except in those younger than 65 years and without CVD.9 Few studies have specifically explored the impact of new BP categories on all-cause and CVD mortality in mainland China. A previous analysis from the Shanghai Women’s Health Study (SWHS, 2002–2007, 68,438 participants with 1574 deaths) revealed that participants with SBP of 130–139 mmHg and/or DBP of 85–89 mmHg (similar to stage 1 hypertension in the new guideline) did not have increased risks of all-cause or CHD deaths compared with those with optimal BP (SBP < 120 mmHg and DBP < 80 mmHg), but increased risk of stroke deaths.10
Therefore, we aimed to investigate the risks of all-cause and CVD (including IHD and CBVD) mortality related to the new BP criteria in Chinese adults using data from three large prospective cohort studies. We further examined whether the associations were different in various subgroups (age, sex, baseline comorbidities, etc.), and the associations of components of BP (SBP, DBP, pulse pressure (PP) and mean arterial pressure (MAP)) with all-cause and CVD mortality.
Methods
Study population
Data from three prospective cohort studies were used: SWHS, Shanghai Men’s Health Study (SMHS) and the Dongfeng-Tongji (DFTJ) cohort. The SWHS recruited 74,942 women aged 40–70 years in 1996–2000 and measured BP in 2000–2002. The SMHS recruited 61,480 men aged 40–74 years in 2002–2006. The DFTJ enrolled 27,009 retired employees with an average age of 63.6 years from the Dongfeng Motor Corporation (DMC) in 2008–2010. The detailed information about these studies has been published elsewhere,11–13 and is shown in the Supplementary methods file. For the current analysis, participants with missing information of baseline BP measurements (n = 6545 for the SWHS, n = 1106 for the SMHS and n = 645 for the DFTJ cohort), missing or invalid data on vital status or censor date (n = 38 for the SWHS/SMHS and n = 709 for the DFTJ cohort) were excluded. Finally, a total of 128,733 participants from the SWHS/SMHS and 25,674 from the DFTJ cohort remained.
The studies were approved by the institutional review boards of Vanderbilt University and the Shanghai Cancer Institute (for the SWHS/SMHS), and the medical ethics committee of the School of Public Health, Tongji Medical College, Huazhong University of Science and Technology, and Dongfeng Central Hospital, DMC (for the DFTJ cohort). All participants provided informed consent forms.
BP measurements and categorisation
For the SWHS/SMHS, resting BP was measured twice with a mercury sphygmomanometer by trained staff at the participants’ home, and the average of two readings was used for analyses. For the DFTJ cohort, resting BP was measured once with a mercury sphygmomanometer by trained staff at the health examination centre of the Dongfeng Central Hospital. In all cohorts, BP was all taken in the right arm of seated participants with cuffs of appropriate size.
PP was defined as SBP minus DBP, and MAP as DBP plus one third of PP. BP was categorised based on the 2017 ACC/AHA guidelines, as mentioned above.
Follow-up and ascertainment of outcome
For the SWHS/SMHS, participants were followed with a combination of annual linkage to Shanghai Vital Statistics Registry and home visits every 2–4 years. In the DFTJ cohort, vital status was tracked with the DMC medical insurance system and electronic medical records in the affiliated hospitals. The DMC had a full coverage of medical care and insurance, which minimised the loss to follow-up rate.
The underlying causes of death were coded according to the International Classification of Disease (ICD)-9 for the SWHS/SMHS and the ICD-10 for the DFTJ cohort. The outcomes of interest included mortality due to all causes, CVD, IHD and CBVD. The ICD codes for the outcomes are listed in the Supplementary methods file.
Assessment of covariates
The information on covariates (including sociodemographic characteristics, lifestyle factors and comorbidities) was ascertained using standardised questionnaires at baseline and the details are described in the Supplementary methods file.
Statistical analysis
Person-years were counted from the baseline survey (2000–2002 for SWHS, 2002–2006 for SMHS and 2008–2010 for DFTJ cohort) until the date of loss to follow-up, death or 31 December 2014 (SWHS/SMHS) and 31 December 2016 (DFTJ cohort), whichever came first. Continuous and categorical variables were presented as mean (standard deviation; SD) and count (percentage), respectively. Baseline characteristics across BP categories were compared using one-way analysis of variance (ANOVA) for continuous variables and the chi-square test for categorical variables. Cox proportional hazards regression models were used to estimate the adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) of the BP—mortality associations. No significant violations of the proportional hazards assumptions were found. Four multivariate models were constructed: model 1 adjusted for sociodemographic factors (including age, sex, ethnicity (DFTJ cohort only), education level, personal income (SWHS/SMHS only) and marital status); model 2 additionally adjusted for smoking, alcohol drinking, sleep duration (DFTJ cohort only), regular physical activity and BMI; model 3 further adjusted for postmenopausal status (women only), baseline history of diabetes, hyperlipidemia (DFTJ cohort only), CVD and cancer; model 4 finally adjusted for the use of antihypertensive drugs. To test the robustness of the results, we repeated the analyses by excluding participants with baseline cancer or those died within 2 years of follow-up.
Stratified analyses were conducted to investigate whether the associations were modified by some specific baseline characteristics: sex (men, women), age (<65, ≥65 years, in accordance with different treatment strategies in the ACC/AHA guidelines), BMI (<24, ≥24 kg/m2), baseline CVD (no, yes), baseline history of diabetes (no, yes), physician-diagnosed hypertension (no, yes) and the use of antihypertensive medications (no, yes). Effect modifications were assessed by including interaction terms of these variables with BP categories in the final model. Analyses were conducted separately in each cohort, and results were summarised by meta-analyses with a random-effects model. The population attributable risk proportion (PAR%) was calculated to estimate the proportion of CVD death attributable to hypertension with the following equation: PAR% = Pe (HR−1)/[Pe (HR−1)+1]×100%, where Pe was the prevalence of hypertension in our study and HR was the pooled risk estimate for this category. All statistical analyses were performed using SAS 9.3 (SAS Institute, Cary, NC, USA). Statistical significance was defined as two-sided P < 0.05.
Results
Based on the 2017 ACC/AHA guidelines, the prevalence of hypertension was 66.3% (stage 1 hypertension 45.6%, stage 2 hypertension 20.7%) in the whole population with a similar prevalence across cohorts.
The baseline characteristics of participants are presented in Table 1. Those with higher BP were older and more likely to be men, less educated, current drinkers, postmenopausal (women only) and have higher BMI and a higher prevalence of comorbidities (except cancer), but were less likely to be married. In the SWHS/SMHS, participants with higher BP were more likely to be current smokers, have lower personal income but higher physical activity levels. While in the DFTJ cohort, those with higher BP were less likely to be current smokers and slept for fewer hours.
Table 1.
Baseline characteristics according to blood pressure categories in the SWHS/SMHS cohorts and the DFTJ cohort.
| SWHS/SMHS |
P value | DFTJ cohort |
P value | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Blood pressure categories |
Blood pressure categories |
|||||||||
| Normal | Elevated | Stage 1 hypertension | Stage 2 hypertension | Normal | Elevated | Stage 1 hypertension | Stage 2 hypertension | |||
| SBP (mmHg) | <120 | 120–129 | 130–139 | ≥140 | <120 | 120–129 | 130–139 | ≥140 | ||
| AND | AND | OR | OR | AND | AND | OR | OR | |||
| DBP (mmHg) | <80 | <80 | 80–89 | ≥90 | <80 | <80 | 80–89 | ≥90 | ||
| n (%) | 35,026 (27.2) | 8738 (6.8) | 59,096 (45.9) | 25,873 (20.1) | 4886 (19.0) | 3410 (13.3) | 11,321 (44.1) | 6057 (23.6) | ||
| Age, years | 51.5 ± 8.1 | 56.9 ± 9.7 | 56.0 ± 9.4 | 57.9 ± 9.3 | <0.001 | 61.4 ± 8.0 | 64.2 ± 7.8 | 63.8 ± 7.6 | 64.8 ± 7.5 | <0.001 |
| Ethnicity, Han, n (%) | – | – | – | – | 4831 (98.9) | 3351 (98.3) | 11,151 (98.5) | 5973 (98.6) | 0.12 | |
| Women, n (%) | 22,971 (65.6) | 5257 (60.2) | 28,932 (49.0) | 11,202 (43.3) | <0.001 | 2981 (61.0) | 1872 (54.9) | 6229 (55.0) | 3088 (51.0) | <0.001 |
| Education level, n (%) | <0.001 | |||||||||
| Primary school | 2990 (8.5) | 1684 (19.3) | 9248 (15.7) | 5055 (19.5) | <0.001 | 1264 (25.9) | 999 (29.3) | 3456 (30.5) | 2075 (34.3) | |
| Middle school | 13,544 (38.7) | 2930 (33.5) | 20,510 (34.7) | 9028 (34.9) | 1698 (34.8) | 1148 (33.7) | 4123 (36.4) | 2240 (37.0) | ||
| High school | 12,440 (35.5) | 2575 (29.5) | 18,387 (31.1) | 7330 (28.3) | 1356 (27.8) | 842 (24.7) | 2647 (23.4) | 1209 (20.0) | ||
| College/university or more | 6052 (17.3) | 1549 (17.7) | 10,951 (18.5) | 4460 (17.2) | 568 (11.6) | 421 (12.4) | 1095 (9.7) | 533 (8.8) | ||
| Income, n (%) | <0.001 | |||||||||
| Low | 5101 (14.6) | 1308 (15.0) | 8344 (14.1) | 3957 (15.3) | – | – | – | – | ||
| Middle | 26,215 (74.8) | 6636 (75.9) | 45,265 (76.6) | 19,838 (76.7) | – | – | – | – | ||
| High | 3710 (10.6) | 794 (9.1) | 5487 (9.3) | 2078 (8.0) | – | – | – | – | ||
| Marital status, n (%) | <0.001 | <0.001 | ||||||||
| Married | 31,233 (89.2) | 7625 (87.3) | 51,318 (86.8) | 22,120 (85.5) | 4395 (90.0) | 3014 (88.4) | 10211 (90.2) | 5370 (88.7) | ||
| Widowed | 1033 (3.0) | 471 (5.4) | 2549 (4.3) | 1326 (5.1) | 322 (6.6) | 310 (9.1) | 862 (7.6) | 570 (9.4) | ||
| Separated/divorced | 850 (2.4) | 164 (1.9) | 998 (1.7) | 309 (1.2) | 136 (2.8) | 61 (1.8) | 181 (1.6) | 95 (1.6) | ||
| Never married | 408 (1.2) | 68 (0.8) | 573 (1.0) | 223 (0.9) | 18 (0.4) | 12 (0.4) | 33 (0.3) | 13 (0.2) | ||
| Missed values | 1502 (4.3) | 410 (4.7) | 3658 (6.2) | 1895 (7.3) | 15 (0.3) | 13 (0.4) | 34 (0.3) | 9 (0.2) | ||
| Smoking, n (%) | <0.001 | <0.001 | ||||||||
| Never smoker | 25,546 (72.9) | 6184 (70.8) | 37,588 (63.6) | 15,373 (59.4) | 3526 (72.2) | 2340 (68.6) | 8018 (70.8) | 4188 (69.1) | ||
| Former smoker | 978 (2.8) | 458 (5.2) | 3571 (6.0) | 1924 (7.4) | 423 (8.7) | 427 (12.5) | 1403 (12.4) | 789 (13.0) | ||
| Current smoker | 8502 (24.3) | 2096 (24.0) | 17,937 (30.4) | 8576 (33.2) | 937 (19.2) | 643 (18.9) | 1900 (16.8) | 1080 (17.8) | ||
| SWHS/SMHS | P value | DFTJ cohort | P value | |||||||
| Blood pressure categories | Blood pressure categories | |||||||||
| Normal | Elevated | Stage 1 hypertension | Stage 2 hypertension | Normal | Elevated | Stage 1 hypertension | Stage 2 hypertension | |||
| Alcohol drinking, n (%) | <0.001 | <0.001 | ||||||||
| None | 31,013 (88.5) | 7560 (86.5) | 48,253 (81.7) | 19,998 (77.3) | 3662 (75.0) | 2532 (74.3) | 8319 (73.5) | 4294 (70.9) | ||
| Former regular drinker | 470 (1.3) | 158 (1.8) | 1455 (2.5) | 799 (3.1) | 261 (5.3) | 201 (5.9) | 637 (5.6) | 401 (6.6) | ||
| Current regular drinker | 3543 (10.1) | 1020 (11.7) | 9388 (15.9) | 5076 (19.6) | 963 (19.7) | 677 (19.9) | 2365 (20.9) | 1362 (22.5) | ||
| Regular physical activity, n (%) | 10,967 (31.3) | 3561 (40.8) | 23,372 (39.6) | 10,676 (41.3) | <0.001 | 4327 (88.6) | 3049 (89.4) | 10090 (89.1) | 5342 (88.2) | 0.17 |
| Sleep duration, n (%) | <0.001 | |||||||||
| ≤5 hours/day | – | – | – | – | 262 (5.4) | 183 (5.4) | 802 (7.1) | 461 (7.6) | ||
| 6–8 hours/day | – | – | – | – | 2553 (52.3) | 1836 (53.8) | 5998 (53.0) | 3131 (51.7) | ||
| ≥9 hours/day | – | – | – | – | 2071 (42.4) | 1391 (40.8) | 4521 (39.9) | 2465 (40.7) | ||
| Post-menopausal (women only, yes, %) | 9981 (43.5) | 3455 (65.7) | 18,953 (65.5) | 8348 (74.5) | <0.001 | 2577 (86.5) | 1726 (92.2) | 5677 (91.1) | 2870 (92.9) | <0.001 |
| Body mass index, kg/m2 | 22.5 ± 2.9 | 23.6 ± 3.1 | 24.2 ± 3.2 | 25.2 ± 3.3 | <0.001 | 23.4 ± 3.2 | 24.1 ± 3.3 | 24.8 ± 3.3 | 25.3 ± 3.5 | <0.001 |
| SBP, mmHg | 105.0 ± 7.7 | 121.1 ± 2.2 | 127.2 ± 11.8 | 148.4 ± 15.5 | <0.001 | 106.0 ± 6.7 | 121.0 ± 2.0 | 132.0 ± 12.7 | 150.5 ± 14.5 | <0.001 |
| DBP, mmHg | 68.5 ± 5.2 | 72.1 ± 3.9 | 82.1 ± 4.7 | 93.6 ± 8.5 | <0.001 | 67.0 ± 5.3 | 69.5 ± 4.7 | 79.6 ± 6.1 | 88.1 ± 12.3 | <0.001 |
| Self-reported history of hypertension, n (%) | 2377 (6.8) | 1684 (19.3) | 18,561 (31.4) | 15,596 (60.3) | <0.001 | 772 (15.8) | 1002 (29.4) | 4734 (41.8) | 3574 (59.0) | <0.001 |
| Current use of antihypertensive medications, n (%) | 1237 (3.5) | 1018 (11.7) | 12,115 (20.5) | 10,597 (41.0) | <0.001 | 611 (12.5) | 829 (24.3) | 3837 (33.9) | 2820 (46.6) | <0.001 |
| History of diabetes, n (%) | 1089 (3.1) | 696 (8.0) | 4160 (7.0) | 2229 (8.6) | <0.001 | 640 (13.1) | 636 (18.7) | 2172 (19.2) | 1351 (22.3) | <0.001 |
| History of hyperlipidemia, n (%) | – | – | – | – | 1999 (40.9) | 1616 (47.4) | 5864 (51.8) | 3381 (55.8) | <0.001 | |
| History of cardiovascular disease, n (%) | 1404 (4.0) | 840 (9.6) | 5506 (9.3) | 3464 (13.4) | <0.001 | 748 (15.3) | 718 (21.1) | 2366 (20.9) | 1435 (23.7) | <0.001 |
| History of cancer, n (%) | 1688 (4.8) | 429 (4.9) | 3006 (5.1) | 1284 (5.0) | 0.33 | 357 (7.3) | 210 (6.2) | 676 (6.0) | 309 (5.1) | <0.001 |
SBP: systolic blood pressure; DBP: diastolic blood pressure.
Data are shown as n (%) or mean ± standard deviation, unless otherwise specified.
P values were derived from one-way analysis of variance (ANOVA) for continuous variables and the chi-square test for categorical variables.
‘–’ means data not available.
In the SWHS/SMHS, 12,209 deaths (4104 from CVD, 794 from IHD and 2121 from CBVD) were recorded, with a median follow-up of 12.5 years. In the DFTJ cohort, 2483 deaths (982 CVD deaths, 483 IHD deaths and 388 CBVD deaths) were documented during a median follow-up of 8.3 years. The pooled HR (95% CI) associated with stage 1 hypertension was 1.40 (1.16–1.69) for CVD mortality, 1.36 (1.12–1.65) for IHD mortality and 1.53 (1.18–2.00) for CBVD mortality, respectively, compared with the normal BP group (Table 2). However, stage 1 hypertension was not significantly associated with all-cause mortality (HR 1.04, 95% CI 0.89–1.21). Stage 2 hypertension was significantly associated with mortality, ranging from 1.26 (1.02–1.55) for all-cause mortality to 2.49 (1.82–3.40) for CBVD mortality. The cohort-specific associations are shown in Supplementary Tables 1 and 2. When deleting participants with baseline cancer or those who died within 2 years of follow-up, the results did not change materially (Supplementary Table 3).
Table 2.
Blood pressure categories and mortality risk: pooled results from the SWHS/SMHS cohorts and the DFTJ cohort.
| Mortality | Blood pressure categories |
|||
|---|---|---|---|---|
| Normal | Elevated | Stage 1 hypertension | Stage 2 hypertension | |
| All-cause mortality | ||||
| Model 1 | 1 | 0.95 (0.82–1.10) | 1.04 (0.85–1.27) | 1.27 (0.95–1.69) |
| Model 2 | 1 | 0.96 (0.83–1.11) | 1.08 (0.91–1.28) | 1.32 (1.04–1.69) |
| Model 3 | 1 | 0.93 (0.80–1.08) | 1.06 (0.91–1.23) | 1.30 (1.06–1.61) |
| Model 4 | 1 | 0.92 (0.79–1.07) | 1.04 (0.89–1.21) | 1.26 (1.02–1.55) |
| CVD mortality | ||||
| Model 1 | 1 | 1.17 (1.00–1.38) | 1.53 (1.19–1.98) | 2.24 (1.48–3.39) |
| Model 2 | 1 | 1.17 (1.02–1.35) | 1.55 (1.25–1.91) | 2.25 (1.60–3.16) |
| Model 3 | 1 | 1.11 (0.97–1.28) | 1.50 (1.26–1.78) | 2.15 (1.61–2.88) |
| Model 4 | 1 | 1.09 (0.95–1.25) | 1.40 (1.16–1.69) | 1.93 (1.45–2.58) |
| IHD mortality | ||||
| Model 1 | 1 | 1.04 (0.80–1.35) | 1.54 (1.28–1.87) | 1.89 (1.39–2.56) |
| Model 2 | 1 | 1.01 (0.77–1.32) | 1.49 (1.23–1.81) | 1.81 (1.48–2.22) |
| Model 3 | 1 | 0.96 (0.73–1.25) | 1.42 (1.17–1.72) | 1.71 (1.40–2.09) |
| Model 4 | 1 | 0.94 (0.72–1.23) | 1.36 (1.12–1.65) | 1.59 (1.30–1.95) |
| CBVD mortality | ||||
| Model 1 | 1 | 1.27 (1.03–1.57) | 1.72 (1.27–2.33) | 2.98 (2.00–4.46) |
| Model 2 | 1 | 1.26 (1.02–1.55) | 1.73 (1.34–2.24) | 3.00 (2.16–4.17) |
| Model 3 | 1 | 1.20 (0.97–1.48) | 1.68 (1.34–2.10) | 2.86 (2.15–3.81) |
| Model 4 | 1 | 1.16 (0.94–1.44) | 1.53 (1.18–2.00) | 2.49 (1.82–3.40) |
Data are shown as pooled hazard ratio (95% confidence interval) obtained by random-effects meta-analysis.
Model 1: adjusted for age, sex, race (DFTJ cohort only), education level, marital status, and income (SWHS/SMHS only).
Model 2: model 1 plus smoking status, drinking status, physical activity, sleep duration (DFTJ cohort only) and body mass index.
Model 3: model 2 plus post-menopausal (women only), history of diabetes, hyperlipidemia (DFTJ cohort only), cardiovascular disease and cancer.
Model 4: model 3 plus use of antihypertensive medications.
Table 3 and Supplementary Table 4 present the pooled estimates stratified by baseline characteristics. For all-cause mortality, the risk estimates conferred by stage 1 hypertension ranged from 0.96 to 1.12 in subgroups. Notably, the associations with CVD/IHD/CBVD mortality were stronger for participants younger than 65 years, those without baseline CVD or those not taking antihypertensive medications than their counterparts (P for heterogeneity <0.05). For participants younger than 65 years, even the elevated BP category was related to increased risks of CVD mortality (1.55, 1.04–2.29) and CBVD mortality (2.03, 1.31–3.15). Cohort-specific stratified analyses are shown in Supplementary Tables 5 and 6, and the results were generally consistent across cohorts.
Table 3.
Stratified analysis of blood pressure categories and mortality risk: pooled results from the SWHS/SMHS cohorts and the DFTJ cohort.
| Mortality | Blood pressure categories |
P for heterogeneity |
|||
|---|---|---|---|---|---|
| Normal | Elevated | Stage 1 hypertension | Stage 2 hypertension | ||
| All-cause mortality | |||||
| Men | 1 | 0.96 (0.75–1.23) | 1.03 (0.84–1.26) | 1.26 (0.91–1.75) | 0.86 |
| Women | 1 | 0.91 (0.83–1.00) | 1.10 (1.03–1.17) | 1.31 (1.22–1.42) | |
| Age < 65 years | 1 | 0.98 (0.87–1.12) | 1.09 (0.90–1.31) | 1.55 (1.42–1.69) | 0.03 |
| Age ≥ 65 years | 1 | 0.90 (0.81–1.00) | 1.01 (0.91–1.12) | 1.15 (0.93–1.42) | |
| Without history of CVD | 1 | 0.95 (0.87–1.03) | 1.02 (0.84–1.25) | 1.27 (1.07–1.51) | 0.73 |
| With history of CVD | 1 | 0.92 (0.68–1.24) | 1.09 (0.98–1.22) | 1.26 (0.92–1.74) | |
| No use of antihypertensive medications | 1 | 0.93 (0.80–1.07) | 1.04 (0.89–1.22) | 1.33 (1.14–1.55) | 0.88 |
| Use of antihypertensive medications | 1 | 0.89 (0.75–1.06) | 0.99 (0.81–1.21) | 1.14 (0.83–1.56) | |
| CVD mortality | |||||
| Men | 1 | 1.21 (0.95–1.53) | 1.38 (1.19–1.59) | 1.97 (1.37–2.84) | 0.92 |
| Women | 1 | 0.96 (0.78–1.18) | 1.40 (1.03–1.89) | 1.99 (1.72–2.30) | |
| Age < 65 years | 1 | 1.55 (1.04–2.29) | 1.85 (1.54–2.22) | 3.44 (2.84–4.15) | <0.001 |
| Age ≥ 65 years | 1 | 0.90 (0.74–1.10) | 1.22 (1.04–1.44) | 1.51 (1.14–2.01) | |
| Without history of CVD | 1 | 1.15 (0.96–1.38) | 1.52 (1.34–1.72) | 2.35 (2.06–2.67) | 0.03 |
| With history of CVD | 1 | 0.92 (0.67–1.27) | 1.24 (1.00–1.54) | 1.52 (1.06–2.17) | |
| No use of antihypertensive medications | 1 | 1.04 (0.88–1.24) | 1.48 (1.29–1.69) | 2.37 (2.08–2.69) | 0.05 |
| Use of antihypertensive medications | 1 | 0.99 (0.77–1.27) | 1.13 (0.93–1.38) | 1.41 (1.01–1.97) | |
| IHD mortality | |||||
| Men | 1 | 0.96 (0.68–1.36) | 1.26 (0.98–1.63) | 1.42 (1.09–1.87) | 0.21 |
| Women | 1 | 0.90 (0.59–1.38) | 1.46 (1.08–1.97) | 1.79 (1.31–2.45) | |
| Age < 65 years | 1 | 1.70 (0.47–6.17) | 1.61 (1.03–2.54) | 2.71 (1.44–5.10) | 0.01 |
| Age ≥ 65 years | 1 | 0.75 (0.55–1.03) | 1.23 (0.98–1.55) | 1.28 (1.00–1.63) | |
| Without history of CVD | 1 | 1.12 (0.78–1.61) | 1.56 (1.21–2.02) | 1.95 (1.48–2.56) | 0.03 |
| With history of CVD | 1 | 0.74 (0.50–1.11) | 1.07 (0.80–1.43) | 1.16 (0.85–1.58) | |
| No use of antihypertensive medications | 1 | 1.13 (0.81–1.58) | 1.59 (1.25–2.02) | 2.01 (1.55–2.62) | 0.004 |
| Use of antihypertensive medications | 1 | 0.61 (0.39–0.95) | 0.86 (0.62–1.19) | 0.96 (0.69–1.33) | |
| CBVD mortality | |||||
| Men | 1 | 1.23 (0.89–1.70) | 1.50 (1.19–1.89) | 2.88 (2.27–3.64) | 0.29 |
| Women | 1 | 0.99 (0.58–1.69) | 1.53 (0.99–2.36) | 2.18 (1.39–3.43) | |
| Age < 65 years | 1 | 2.03 (1.31–3.15) | 2.80 (2.06–3.80) | 6.26 (4.58–8.56) | <0.001 |
| Age ≥ 65 years | 1 | 0.88 (0.69–1.12) | 1.22 (1.03–1.45) | 1.75 (1.46–2.10) | |
| Without history of CVD | 1 | 1.26 (0.96–1.65) | 1.76 (1.46–2.13) | 3.16 (2.60–3.85) | 0.005 |
| With history of CVD | 1 | 0.94 (0.65–1.36) | 1.28 (1.00–1.65) | 1.78 (1.26–2.52) | |
| No use of antihypertensive medications | 1 | 1.02 (0.78–1.33) | 1.29 (0.71–2.35) | 2.98 (2.33–3.80) | 0.004 |
| Use of antihypertensive medications | 1 | 1.16 (0.80–1.70) | 1.33 (0.99–1.80) | 1.86 (1.38–2.52) | |
Data are shown as pooled hazard ratio (95% confidence interval) of model 4 obtained by random-effects meta-analysis.
With the prevalence of 45.6% for stage 1 hypertension in our study, the PAR% was 15.4% for CVD mortality, 14.1% for IHD mortality and 19.5% for CBVD mortality. The corresponding PAR% related to stage 2 hypertension was 16.1%, 10.9% and 23.6%, respectively.
Higher SBP, DBP, PP and MAP were all positively associated with mortality risk (see Supplementary Tables 7–9 for the pooled and cohort-specific results), and the associations were substantially greater in younger participants than their older counterparts (see Supplementary Tables 10–12).
Discussion
With data from three population-based cohorts in mainland China, we found that stage 1 hypertension defined by the new ACC/AHA guidelines was associated with an elevated risk of CVD, IHD and CBVD mortality but not with all-cause mortality. Of all the CVD deaths, 15.4% could be prevented if stage 1 hypertension was eliminated. In the stratified analyses, we found that the BP—mortality associations were stronger in younger adults and participants without a history of CVD. Blood pressure components were generally linearly associated with increased mortality risks, and the associations weakened with increasing age.
The term ‘prehypertension (120–139/80–89 mmHg)’ was first introduced in the JNC-7,14 which could be further grouped into low-range prehypertension (120–129/80–84 mmHg, similar to elevated BP in the new guideline) and high-range prehypertension (130–139/85–89 mmHg, similar to stage 1 hypertension in the new guideline). In our study, we found that stage 1 hypertension was associated with an elevated risk of CVD mortality but not with all-cause mortality. The findings are generally consistent with results from a meta-analysis of 20 cohort studies, which reported that high-range prehypertension was associated with an elevated risk of CVD mortality (1.28, 1.16–1.41) but not all-cause mortality (1.03, 0.95–1.12).15 In this meta-analysis, low-range prehypertension was neither associated with all-cause mortality nor with CVD mortality, which is in agreement with our results.
Several studies have been conducted to explore the association between prehypertension and mortality among Asian populations, with conflicting conclusions.10,16,17 A Singapore study with 5830 adults (62.2% were Chinese) found no significant associations between prehypertension and all-cause and CVD mortality, even for high-range prehypertension.16 However, the statistical power of that study may be limited (419 all-cause deaths including 151 CVD deaths). Another prospective cohort study with 169,871 adults aged 40 years and above from 17 provinces in China found that prehypertension was related to an increased risk of CVD mortality (1.22, 1.15–1.30), CHD mortality (1.47, 1.23–1.75) and stroke mortality (1.67, 1.50–1.86).17 However, that study did not examine low-range and high-range prehypertension further. A previous analysis in the SWHS reported that high-range prehypertension was significantly associated with stroke mortality but not with CHD or all-cause mortality.10 However, that analysis only included 91 CHD deaths and 247 stroke deaths due to a short follow-up (median follow-up of 5.1 years). In our current analysis, with extended follow-up years of the SWHS cohort and combination with two other cohorts, we observed that stage 1 hypertension was associated with approximately 40–50% increased risks of CVD/IHD/CBVD mortality.
We also found that the mortality risk estimates were greater for participants younger than 65 years compared with their counterparts. Analyses of individual BP components also showed that the BP—mortality associations were attenuated with increasing age. Similar findings were reported from the Singapore Chinese Health Study of 30,636 participants9 and a Japanese study of 67,309 participants.18 The China Kadoorie Biobank study (CKB) with about half a million participants also demonstrated that the risk of incident vascular diseases in relation to SBP weakened with increasing age.19 It has been shown that BP may provide less information in predicting CVD risk in old people with multiple comorbidities, and the direct measurement of arterial function (e.g. pulse wave velocity) may be more informative in the elderly.20 The findings suggest that stricter BP control may be more beneficial for younger adults. As indicated in our study, 27.4% of CVD deaths could be prevented if stage 1 hypertension is eliminated among adults less than 65 years old, while the estimate was only 9.8% for those aged 65 years and older.
In our analysis, stage 1 hypertension was related to a higher risk of CVD mortality for those with and without baseline CVD, but was more evident for those without pre-existing CVD, which was in agreement with the Singapore Chinese Health study.9 The weaker association among those with previous CVD may be because atherosclerotic peripheral arterial disease may coexist with CVD, and conventional arm BP could not be accurately measured if the atherosclerosis occurred in subclavian and brachial arteries.21 Survival bias and confounding by treatment strategies were also possible among those with prior CVD. Our findings highlighted the importance of BP monitoring and control in general populations without CVD history. To be noted, the ACC/AHA guideline does not recommend drug prescription for the general population with stage 1 hypertension, except for those with CVD or with an estimated 10-year CVD risk of 10% or greater, which only accounted for 2.0–5.5% in the Chinese population.22 Unfortunately, we did not have data to calculate the 10-year CVD risk and further studies are still needed to evaluate whether stage 1 hypertension was associated with a higher mortality risk among those with a 10-year CVD risk less than 10%. Our current analyses found significant associations between stage 1 hypertension and CVD mortality among those not taking antihypertensive medications, but not among those taking antihypertensive medications, suggesting the potential benefits of treating stage 1 hypertension, but more studies are still needed to confirm this finding. With the implementation of the 2017 ACC/AHA guideline, it is undoubted that the prevalence of hypertension would increase substantially,7,23 as well as the cost incurred by treating the subgroups of the newly diagnosed stage 1 hypertension and intense treatment for those who already take medications to meet the new guideline.22,23 Therefore, a careful evaluation is urgently needed of the cost-effectiveness of the 2017 ACC/AHA guideline before any implementation.22–24
This is one of the first prospective studies to assess the application of the new ACC/AHA guideline in mainland China. Our study has several strengths. First, it had a large sample size from three different cohorts with long-term follow-ups, which ensured the statistical power and made subgroup analyses feasible. Second, detailed information on comorbidities and baseline characteristics allowed us to adjust for a broad range of potential confounders and to conduct multiple stratification analyses among individuals with a different baseline CVD risk. Third, deaths were ascertained by linkage with well-established insurance system or vital registries to minimise loss to follow-up, and the misclassification of potential causes of mortality was minimal.
There are some limitations to be acknowledged. First, we did not have information and thus could not account for the classes of antihypertensive drugs, which may have differential effects.25 Second, residual confounding could not be fully eliminated in an observational study such as ours. Third, there are heterogeneities of BP measurement settings (clinic vs. in-home) between different cohorts, although BP measurements were all performed by trained staff. Fourth, we only had BP measurements at a single visit and could not capture BP changes over time. Furthermore, analysis on incident CVD was not conducted because the validation work on the self-reported non-fatal CVD events is still in progress.
Conclusions
In three prospective cohorts of over 150,000 Chinese adults, we found that the stage 1 hypertension newly defined by the 2017 ACC/AHA guideline was associated with increased risks of CVD, IHD and CBVD mortality. For individuals younger than 65 years and individuals without pre-existing CVD, comprehensive management (home blood pressure monitoring, lifestyle modifications, etc.) of BP to less than 130/80 mmHg may be optimal for the prevention of CVD deaths.
Supplementary Material
Acknowledgements
The authors thank all the participants and research staff who took part in the Shanghai Women’s and Men’s Health Studies and the Dongfeng-Tongji cohort for their contributions. This work has not been presented before, in whole or in part, in any language.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: the study was supported by the National Key Research and Development Program of China (2017YFC0907504). The DFTJ cohort was supported by the National Key Research and Development Program of China (2016YFC0900800, 2016YFC0900801, 2017YFC0907500 and 2017YFC0907501) and the Natural National Scientific Foundation of China (91643202, 81230069 and 81390542). The SWHS/SMHS cohorts were supported by grants from US National Institutes of Health (R37 CA070867 and UM1 CA182910; R01 CA082729 and UM1 CA173640). DY is supported by Vanderbilt University Medical Center Faculty Research Scholars Program.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
References
- 1.Global Burden of Disease 2016; Causes of Death Collaborators. Global, regional, and national age-sex specific mortality for 264 causes of death, 1980–2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet 2017; 390: 1151–1210. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Zhang Y, Jiang X, Bo J, et al. Risk of stroke and coronary heart disease among various levels of blood pressure in diabetic and nondiabetic Chinese patients. J Hypertens 2018; 36: 93–100. [DOI] [PubMed] [Google Scholar]
- 3.Lewington S, Clarke R, Qizilbash N, et al. Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies. Lancet 2002; 360: 1903–1913. [DOI] [PubMed] [Google Scholar]
- 4.Whelton PK, Carey RM, Aronow WS, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: a report of the American College of Cardiology/American Heart Association task force on clinical practice guidelines. Hypertension 2018; 71: e13–e115. [DOI] [PubMed] [Google Scholar]
- 5.James PA, Oparil S, Carter BL, et al. 2014 Evidence-based guideline for the management of high blood pressure in adults: report from the panel members appointed to the Eighth Joint National Committee (JNC 8). JAMA 2014; 311: 507–520. [DOI] [PubMed] [Google Scholar]
- 6.Writing Group of 2010 Chinese Guidelines for the Management of Hypertension. 2010 Chinese guidelines for the management of hypertension. Chin J Hypertens 2011; 19: 701–743. [Google Scholar]
- 7.Khera R, Lu Y, Lu J, et al. Impact of 2017 ACC/AHA guidelines on prevalence of hypertension and eligibility for antihypertensive treatment in United States and China: nationally representative cross sectional study. BMJ 2018; 362: k2357. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Bangalore S, Toklu B, Gianos E, et al. Optimal systolic blood pressure target after SPRINT: insights from a network meta-analysis of randomized trials. Am J Med 2017; 130: 707–719.e8. [DOI] [PubMed] [Google Scholar]
- 9.Talaei M, Hosseini N, Koh AS, et al. Association of “elevated blood pressure” and “stage 1 hypertension” with cardiovascular mortality among an Asian population. J Am Heart Assoc 2018; 7: e008911. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Dorjgochoo T, Shu XO, Zhang X, et al. Relation of blood pressure components and categories and all-cause, stroke and coronary heart disease mortality in urban Chinese women: a population-based prospective study. J Hypertens 2009; 27: 468–475. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Zheng W, Chow WH, Yang G, et al. The Shanghai Women’s Health Study: rationale, study design, and baseline characteristics. Am J Epidemiol 2005; 162: 1123–1131. [DOI] [PubMed] [Google Scholar]
- 12.Shu XO, Li H, Yang G, et al. Cohort profile: the Shanghai Men’s Health Study. Int J Epidemiol 2015; 44: 810–818. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Wang F, Zhu J, Yao P, et al. Cohort profile: the Dongfeng-Tongji cohort study of retired workers. Int J Epidemiol 2013; 42: 731–740. [DOI] [PubMed] [Google Scholar]
- 14.Chobanian AV, Bakris GL, Black HR, et al. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. JAMA 2003; 289: 2560–2572. [DOI] [PubMed] [Google Scholar]
- 15.Huang Y, Su L, Cai X, et al. Association of all-cause and cardiovascular mortality with prehypertension: a meta-analysis. Am Heart J 2014; 167: 160–168.e1. [DOI] [PubMed] [Google Scholar]
- 16.Lee J, Heng D, Ma S, et al. Influence of pre-hypertension on all-cause and cardiovascular mortality: the Singapore Cardiovascular Cohort Study. Int J Cardiol 2009; 135: 331–337. [DOI] [PubMed] [Google Scholar]
- 17.Gu D, Chen J, Wu X, et al. Prehypertension and risk of cardiovascular disease in Chinese adults. J Hypertens 2009; 27: 721–729. [DOI] [PubMed] [Google Scholar]
- 18.Fujiyoshi A, Ohkubo T, Miura K, et al. Blood pressure categories and long-term risk of cardiovascular disease according to age group in Japanese men and women. Hypertens Res 2012; 35: 947–953. [DOI] [PubMed] [Google Scholar]
- 19.Lacey B, Lewington S, Clarke R, et al. Age-specific association between blood pressure and vascular and non-vascular chronic diseases in 0.5 million adults in China: a prospective cohort study. Lancet Glob Health 2018; 6: e641–e649. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Benetos A, Petrovic M and Strandberg T. Hypertension management in older and frail older patients. Circ Res 2019; 124: 1045–1060. [DOI] [PubMed] [Google Scholar]
- 21.Clark CE, Taylor RS, Shore AC, et al. The difference in blood pressure readings between arms and survival: primary care cohort study. BMJ 2012; 344: e1327. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Wang JG and Liu L. Global impact of 2017 American College of Cardiology/American Heart Association hypertension guidelines: a perspective from China. Circulation 2018; 137: 546–548. [DOI] [PubMed] [Google Scholar]
- 23.Vaucher J, Marques-Vidal P, Waeber G, et al. Population impact of the 2017 ACC/AHA guidelines compared with the 2013 ESH/ESC guidelines for hypertension management. Eur J Prev Cardiol 2018; 25: 1111–1113. [DOI] [PubMed] [Google Scholar]
- 24.Van Kleef ME and Spiering W. Hypertension: overly important but under-controlled. Eur J Prev Cardiol 2017; 24: 36–43. [DOI] [PubMed] [Google Scholar]
- 25.Ettehad D, Emdin CA, Kiran A, et al. Blood pressure lowering for prevention of cardiovascular disease and death: a systematic review and meta-analysis. Lancet 2016; 387: 957–967. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.