Introduction

Resting heart rate (RHR) is a simple and accessible cardiovascular parameter. More than a decade ago, epidemiologic studies and clinical trial data suggest that RHR could be an important prognostic factor and potential therapeutic target for cardiovascular and all-cause mortality in people with or without diagnosed cardiovascular disease [1,2,3]. However, because of the complex pathophysiology of RHR, defining an optimal RHR for a given individual for preventing and controlling cardiovascular disease is difficult.

Hypertension is the most important and well-known risk factor for cardiovascular disease and the leading cause of death worldwide [4]. A better understanding of the association between RHR and blood pressure (BP) could provide evidence for the primary screening of hypertension in general population. However, the association of RHR and hypertension remains unclear. Some studies suggested an increased risk of hypertension with high RHR [5, 6], whereas others found no significant association [7,8,9]. Furthermore, sex differences were observed in the RHR-hypertension association [10].

Although RHR is pointed as a risk factor of hypertension in Chinese population [11, 12], no age- and sex-specific assessment of RHR-hypertension association has been conducted to date. Because RHR could be influenced by some unmodifiable (age and sex) and modifiable factors (smoking and obesity) [13], we wondered (1) whether the RHR and its dynamic change-hypertension association would differ by sex and (2) whether age, smoking, and obesity would influence the association. Therefore, the present study aimed to investigate the association between RHR and risk of hypertension in a rural Chinese cohort in terms of age and sex status and provide a medical-based preventive strategy for individuals.

Methods

Study population

The Rural Chinese Cohort Study randomly enrolled 20,194 participants older than 18 years living in the rural area of Luoyang city in the middle of China during July to August 2007 and July to August 2008. Details concerning the study design, methods, and inclusion criteria of the study population were published elsewhere [14]. Briefly, baseline examination involved self-reported questionnaires and anthropometric and laboratory measurements with the consent of participants. These participants were then followed up with the same procedures as at baseline during July to August of 2013 and July to October of 2014. Ultimately, 17,265 participants were successfully followed up (response rate 85.5%). The study was approved by the ethics committee of the Shenzhen University Health Sciences Center, and written informed consent was obtained from all participants.

Among the 20,194 participants, 6299 with prevalent hypertension, 407 with cardiovascular diseases (stroke, myocardial infarction, and heart failure), 87 with missing RHR at baseline, 458 who died before follow-up, as well as, 2974 who refused to participate in the follow-up examination were excluded. Thus, 9969 participants were available for the final analysis.

Data collection

Data on demographic characteristics (age, sex, income, and education), lifestyle (smoking, drinking, and physical activity), and personal and family history of cardiovascular disease (stroke, myocardial infarction, and heart failure) were collected by questionnaire interviews. Smoking was defined as currently smoking and/or having smoked at least 100 cigarettes during the lifetime and dichotomized as ever/current versus never smoker [15]. Drinking alcohol was defined as having consumed alcohol 12 or more times in the last year [15]. Physical activity level was assessed by the International Physical Activity Questionnaire and classified as low, moderate, and high [16].

Anthropometric variables were measured with participants wearing light clothing. Weight and height were measured twice to the nearest 0.5 kg and 0.1 cm, respectively. Body mass index (BMI), as an index of general obesity, was calculated as the ratio of weight (kg) to the square of height (m2). BP and RHR were measured by using an electronic brachial sphygmomanometer (HEM-770AFuzzy, Omron, Japan), measuring BP on the upper arm. BP and RHR were measured three times with participants in a seated position after at least 5 min of rest and at intervals of 30 s [17]. Furthermore, smoking; alcohol, coffee, and tea consumption; and excessive exercise were prohibited 30 min before measuring BP, and participants were not allowed to talk during measurement. The mean BP and RHR of each participant were calculated from the recorded measurements and used for the analyses. New cases of hypertension were defined as currently taking antihypertensive medication or systolic BP (SBP) ≥ 140 mmHg or diastolic BP (DBP) ≥90 mmHg at follow-up [18].

Overnight fasting venous blood samples were taken for measuring fasting glucose, total cholesterol (TC), triglycerides (TG), and high-density lipoprotein cholesterol (HDL-C). Details about the storage and measurement methods were reported elsewhere [14, 15]. Diabetes mellitus was defined as fasting glucose ≥7.0 mmol/L or the use of insulin or oral hypoglycemic agents and/or self-reported history of diabetes [19].

Statistical analysis

Baseline characteristics of study participants were analyzed by sex and RHR tertiles. Continuous data were reported with median (interquartile range [IQR]) and categorical data with frequency (%). Linear regression analysis was used to examine the linear trend for baseline continuous variables and logistic regression for categorical variables.

Modified Poisson regression with robust error variances was used to estimate the relative risks (RRs) and 95% confidence intervals (CIs) of incident hypertension associated with baseline RHR and its dynamic change during study period after adjustment for age, income, education, smoking, drinking, physical activity, family history of hypertension, BMI, and fasting glucose, TC, TG, and HDL-C levels. Interaction analyses were conducted to examine the influence of factors (sex, age, BMI, smoking, drinking, and physical activity) on association of RHR and hypertension. Subgroup analyses were further performed after classifying participants by age and sex status. To assess the robustness of the results, three sensitivity analyses were conducted by rerunning the aforementioned model excluding ever/current smokers, alcohol drinkers, or baseline diabetic patients.

All statistical analyses involved using SAS v9.1 (SAS Inst., Cary, NC). Two-sided p < 0.05 was considered statistically significant.

Results

A total of 3930 men and 6039 women were included in the analysis. Median (IQR) age for men and women was 50 (41–59) and 45 (39–55) years and median RHR at baseline was 71 (65–78) and 76 (70–83) beats/min. Table 1 showed the baseline demographics of study participants by tertiles of RHR and by sex. For both men and women, those with higher RHR were younger, more physically inactive, and had higher DBP and fasting glucose (all p < 0.05) (Table 1).

Table 1 Baseline characteristics of study participants by tertiles of resting heart rate (RHR) at baseline for men and women
Full size table

During follow-up (median, 6.01 years), 797 (20.28%) men and 1178 (19.51%) women developed hypertension. Among them, 351 (44.04%) men and 578 (49.07%) women were diagnosed as hypertension according to their current antihypertensive medication at follow-up assessment. The remaining hypertension cases were identified based on the BP values. For men, the cumulative incidence of hypertension with RHR tertile 1 (≤67 beats/min), tertile 2 (68–76 beats/min), and tertile 3 (≥77 beats/min) were 20.10%, 19.64%, and 21.29%, respectively (Table 2). When adjusted for covariates in model 3 (age, income, education, smoking, alcohol drinking, physical activity, family history of hypertension, BMI, fasting glucose, TC, TG, and HDL-C levels), as compared with RHR tertile 1, with tertiles 2 and 3, the adjusted risk of hypertension was 0.98 (95%CI: 0.83–1.15) and 1.02 (95%CI: 0.86–1.20), respectively. For women, the cumulative incidence of hypertension with RHR tertile 1 (≤72 beats/min), tertile 2 (73–80 beats/min), and tertile 3 (≥81 beats/min) was 19.13%, 19.83%, and 19.61%, respectively. For women, as compared with RHR tertile 1, with tertiles 2 and 3, the adjusted risk of hypertension was 1.09 (95%CI: 0.95–1.24) and 1.19 (95%CI: 1.04–1.36) (model 3), respectively.

Table 2 Risk of hypertension by tertiles of resting heart rate (RHR) at baseline
Full size table

Results from interaction analyses indicated that sex, age and BMI may modify the effect of RHR on incident hypertension (all p for interaction term < 0.05, data not shown). Figure 1 showed the trend in RR of hypertension per 10-beat/min increase in RHR among men and women by sex and age status. For women, risk of hypertension per 10-beat/min increase in RHR was associated with young age (18–39 years), middle age (40–59 years), and older age (≥60 years) (adjusted RR per 10-beat/min RHR increase: 1.25 (95%CI: 1.09–1.43), 1.06 (0.99–1.14), and 1.11 (1.01–1.23), respectively. Men showed no association between risk of hypertension and RHR.

Fig. 1
figure 1

Association of per 10 beats/min increase in resting heart rate and risk of hypertension by age and sex status. Data are relative risks (RRs) and 95% confidence intervals (CIs). RRs and 95% CIs were calculated based on modified Poisson regression with adjustment for age, income, education level, smoking, alcohol drinking, physical activity, family history of hypertension, body mass index, and fasting glucose, triglycerides, total cholesterol, and high-density lipoprotein cholesterol levels at baseline

Full size image

We further considered RHR ≤80-beat/min at both baseline and follow-up as stable RHR during study period. Compared with stable RHR, risk of hypertension was significantly higher (RR: 1.22, 95%CI: 1.04–1.42) in women with high RHR (>80-beat/min) throughout the study (Table 3). Women who had normal RHR at baseline but high RHR at follow-up showed non-significantly higher risk of developing hypertension (RR: 1.11, 95%CI: 0.92–1.34). Similar results were found in women with high RHR at baseline but normal RHR at follow-up (RR: 1.10, 95%CI: 0.95–1.27). However, we still found no significant association between dynamic change in RHR and incident hypertension in men.

Table 3 Risk of hypertension by resting heart rate (RHR) at baseline and follow-up
Full size table

Finally, with three sensitivity analyses performed to eliminate the potential confounding effect of smoking, alcohol drinking, and diabetes, the results were not changed. Risk of hypertension was still associated with increased RHR for women (Table 4).

Table 4 Sensitivity analysis of relative risks per 10-beat/min increase in resting heart rate for incident hypertension
Full size table

Discussion

To our knowledge, our study is the first to prospectively analyze the age- and sex-specific relation in the association of RHR and risk of hypertension in Chinese people. Risk of hypertension was increased by 10% with every 10-beat/min increase in RHR for women but not men. Persistently high RHR was positively associated with risk of hypertension in women. Increased RHR was positively associated with varying degrees of hypertension risk in young, middle-aged, and older women. These results provide evidence that strategies for hypertension prevention based on RHR should fully consider age and sex status of individuals.

Increased RHR has previously identified as a risk factor for hypertension in epidemiological studies. However, some unmodifiable factors (age, sex, and race) can affect the impact of RHR on hypertension, which results in inconsistent results in different populations [13]. Three US population-based cohort studies demonstrated a positive [5], male-positive [10], and negative [20] association between RHR and hypertension. Three Japanese cohort studies came to almost completely opposite conclusions (positive [6] or null [8, 9] association). Another two Korean cross-sectional studies reported a positive [21] and male-positive [22] association between RHR and hypertension. Notably, these studies were limited by a short follow-up [6, 8], small sample size [8, 9, 20], and lack of analyses of confounding biases [9, 20], which may also contribute to different results in addition to the ambiguous effect of race and sex.

Only two studies have examined the impact of RHR on hypertension in Chinese adults. Results from The China National Hypertension Survey showed baseline heart rate is an independent risk factor for incident hypertension in 5280 male and 5245 female Chinese adults over a mean of 8.2 years of follow-up [11]. The Kailuan Cohort Study demonstrated a 10-beat/min increase in RHR associated with an 8% increase in hypertension risk among 31 507 non-hypertensive Chinese adults after a mean follow-up of 3.5 years [12]. Although the Kailuan study had a large sample size, it was a coal mining company-based investigation, which over-proportionally consisted of men and restricted further sex-specific analysis. In our community population-based study, the results showed RHR could be considered as a predictor of incident hypertension in women. Moreover, women with persistently high RHR had 22% greater risk of developing hypertension than those had stable normal RHR. Our findings indicated that hypertension screening program should focus more on women with higher RHR to prevent hypertension efficiently in rural China.

RHR progressively decreases with age [23]. However, previous studies seldom considered the impact of age on RHR-hypertension association. Our study first demonstrated an interaction of age in the effect of RHR on risk of hypertension in Chinese people and found elevated RHR more pronounced in young participants, which was consistent with a US study [5]. As compared with young people, older people may have a decreased physiologic and chronotropic response to sympathetic stimulation, with more likely development of hypertension regardless of RHR [5].

Our study also examined the interactive effect between RHR and modifiable determinants of hypertension (smoking, alcohol drinking, obesity, and physical activity). Only obesity status measured by BMI may have a potential role in RHR-hypertension association, which has also found in previous studies [21, 24]. Obesity is a major risk factor for hypertension and also an outcome of higher RHR, which may mask the real effect of RHR on hypertension development. Similar to obesity, smoking is a confounder related to both RHR and hypertension [13]. Considering the high prevalence of smoking (over 70%) among rural men, we performed a sensitivity analysis excluding ever/current smokers and found RHR still not associated with hypertension in rural Chinese men. In addition, although we excluded people with cardiovascular disease at baseline to reduce the selection bias, we further excluded people with diabetes at baseline because insulin resistance is a plausible biologic mechanism for a potential causal relation between increased RHR and hypertension. Finally, our study supported that elevated RHR is significantly associated with increased risk of hypertension in rural Chinese women.

From a pathophysiologic viewpoint, important factors including increased sympathetic activity, vascular injury, activation of inflammatory, and arterial stiffness are generally involved in the causal relationship of increased RHR and hypertension [2]. For example, sympathetic over-activation increases the vasoconstriction of resistance blood vessels and leads to insulin resistance via adrenergic stimulation, thereby elevating BP and promoting a cascade of cardiovascular risk factors [25]. Chronically increased RHR exerts sustained pulsatile stress to the arterial wall, which leads to increased arterial stiffening and arterial hypertension [26].

The major advantages of our study are its prospective design, which avoids reverse causation bias; the large sample size; the collection of data on sociodemographic characteristics and anthropometric and laboratory measurements, which allowed for controlling potential confounding; and the subgroup and sensitivity analyses, which allowed for a better understanding of the effect of potential confounding variables on the RHR-hypertension association. However, some limitations should be noted. First, this study is a single-center study mainly based on the Chinese Han population, which may have implied selection bias. Further multiethnic and multicenter studies are needed to confirm our finding. Second, RHR measurements can be affected by participants’ posture, measurement time, and resting state. To better ensure the accuracy of RHR, our participants remained in a seated position after at least a 5-min rest for RHR measurement by pulse three times during 30-s intervals, which was recommended in a consensus meeting [27]. Third, this is an observational study, so definite causal inference is limited. Fourth, several potential confounders such as mental stress, nutrient consumption, and menopausal status could not be considered properly because of incomplete data at baseline examination.

In conclusion, our findings provide epidemiologic support that RHR may be an important risk factor in the development of hypertension in Chinese women. The dose-response association between RHR and hypertension in women could be affected by age and obesity status. As a noninvasive marker, RHR modified by lifestyle changes may have potential role in preventing hypertension in the women population. Nevertheless, further studies are needed to identify the causal mechanisms of RHR and hypertension and its influencing factors.

Summary

What is known about this topic

  • Resting heart rate has been pointed as a risk factor of hypertension; however, the findings were controversial.

  • No study has assessed the age- and sex-relation in the RHR-hypertension association in Chinese population.

What this study adds

  • We found a significant positive trend of increased risk of hypertension with high RHR in rural Chinese women but not men. Persistently high RHR is associated with increased hypertension risk in women.

  • The dose-response association between RHR and hypertension could be affected by sex and age status.

  • As a non-invasive predictor, RHR modified by lifestyle changes may help prevent hypertension in the women population.