Review Article | Published:

Body mass index, abdominal fatness, and hypertension incidence: a dose-response meta-analysis of prospective studies

Journal of Human Hypertensionvolume 32pages321333 (2018) | Download Citation


Despite the established relationship of obesity to hypertension, the question as to whether there is a linear association between these two morbidities is unanswered. To quantitatively evaluate the relationship between obesity and hypertension, we carried out a dose–response meta-analysis of studies that looked at the relationship of different adiposity measures to hypertension. We searched PubMed, Embase, and Web of Science databases for articles published before 27 June 2017. A random-effects model was used to pool relative risks and 95% confidence intervals. Restricted cubic spline analysis was used to model the relationship. A total of 59 studies were included. Fifty-seven cohort studies with 125,071 incident cases among 830,685 participants were included in the analysis of body mass index and hypertension with the summary relative risk for per 5-unit increment in body mass index of 1.50 (95% confidence interval: 1.40–1.59). We found that the risk of hypertension in the body mass index analysis was greater in populations where the baseline body mass index was <25 kg/m2. The summary relative risk for a 10-cm increase in waist circumference was 1.25 (95% confidence interval: 1.19–1.32) and per 0.1-unit increase in waist-to-hip ratio was 1.27 (95% confidence interval: 1.18–1.37). This meta-analysis suggests that in normal range of obesity indexes, as lean as possible may be the best suggestion to prevent hypertension incidence.


Globally the prevalence of obesity has risen rapidly over the past decades [1]. Several studies have shown that overweight and obesity are associated with increased rates of cardiovascular disease [2], heart failure, and mortality [3], type 2 diabetes [4], as well as hypertension [5].

Hypertension is a global health problem, putting a growing burden on public health agencies through skyrocketing healthcare costs and high mortality rates [6, 7]. Though many studies have shown an increased risk of hypertension with obesity as measured by body mass index (BMI), individual studies may have had limited statistical power to examine the association of waist circumference (WC) and waist-to-hip ratio (WHR) with the risk of hypertension. Moreover, questions remain about the shape of the association curve. In addition, the association between obesity and hypertension remains controversial in many studies. Some studies have suggested an increased risk of hypertension with higher BMI, WC, and WHR [5, 8], whereas three studies found that the association was not significant [9,10,11].

Previous meta-analyses have also explored the association of obesity with hypertension; however, the studies included in one meta-analysis [12] were cross-sectional studies, while the other meta-analysis reported only the cutoff value of BMI and WC with the risk for hypertension [13], and neither explored the quantitative relationship of obesity to hypertension.

BMI, WC, and WHR are closely related to body fat and are the most common indices used to assess the prevalence of obesity. While BMI is used to assess body size and general obesity, WC and WHR are used to assess abdominal obesity [14]. We used BMI, WC, and WHR as indicators of adiposity.

To provide a more powerful and complete assessment of the available evidence in relation to hypertension, we conducted a dose–response meta-analysis of prospective studies to investigate the association between BMI, WC, WHR, and the risk of hypertension.


Literature search strategy

The meta-analysis was conducted and reported in accordance with the Meta-analysis of Observational Studies in Epidemiology (MOOSE) guidelines [15]. We searched the electronic databases PubMed, Embase and Web of Science for studies that were published to 27 June 2017, using wide search terms (Supplementary Table S1-S2). All published studies in English and Chinese were considered. We also manually searched the reference lists to identify additional relevant studies when they were eligible [16, 17]. Study quality was assessed with the Newcastle-Ottawa scale [18].

Study selection

Studies were included in the analyses if they met the following criteria: (1) the exposure of interest was BMI, WC, or WHR; (2) the outcome was hypertension; (3) reported relative risks (RRs), odds ratios (ORs), or hazard ratios (HRs) with 95% confidence intervals (CIs) or sufficient data (such as logistic regression coefficients and standard errors) were provided to calculate them; and (4) the total number of cases and person-years or participants were reported in the publication. Studies were excluded if they met any of the following criteria: (1) the studies were not cohort studies; (2) less than three BMI categories of exposure; (3) reviews or meta-analysis; (4) participants were hypertension patients at baseline. A list of the excluded studies and exclusion reasons is given in the Supplement Table S3.

Data extraction

Two authors (Y.Y.S. and W.Z.) extracted data on the first author, gender, age, country, baseline mean BMI, follow-up years, sample size, number of cases, BMI, WC, and WHR quantity, method of hypertension assessment, assessment for categories of BMI, WC, and WHR, study quality, and variables adjusted for in the analysis. When multiple publications were published from the same study, in general we used the publication with newest and the most detailed data for the relationship analyses [8, 11, 19,20,21,22,23,24,25,26]. If multiple models with different adjusted variables were available in original study, we included the effect value with most covariates adjustment. Any disagreement was discussed until agreement was reached.

Statistical methods

For the studies reporting HRs or ORs for hypertension, we assumed that the HRs and ORs were approximately RRs, and the term RR was used as a generic term for OR and HR [27]. We used a random-effects model [28] which considers both within- and between-study variations to calculate the summary RRs and 95% CIs. If studies reported results separately for race and age, we combined the subgroup estimates using a fixed-effects model before inclusion in the meta-analysis. If the number of cases in each category was missing, these data were inferred on the basis of the number of total cases and the reported effect size. If the exposed person-years or participant numbers were not reported in each category, groups were assumed to be of equal sizes [29]. In addition, to explore the different effects of obesity to hypertension in relative lean and not-lean population, we divided the included study populations into two groups according to whether the reported mean BMI ≥ 25 kg/m2.

The method of generalized least-squares (GLS) regression was used to estimate study-specific dose–response associations [30]. The GLS regression model estimates the linear dose–response relationship coefficient taking into account the covariance for each exposure category within each study, because they are estimated relative to a common referent BMI, WC, and WHR exposure category [31]. This method requires the total number of participants/person-years/cases and RR estimates with their variance estimates for at least three quantitative exposure categories. The median BMI, WC, and WHR level in each category was assigned to the corresponding risk estimate for each study; we extracted means for our analysis if medians were not reported, and for studies that reported the exposures in ranges we used the midpoint of the upper and the lower cutoff point. If the lowest or highest category was open-ended, we used the width of the adjacent category to calculate an upper or lower bound [32]. When the reference category used in the analysis was not the lowest category, we used the method of Hamling et al. [33] to convert risk estimates. The potential non-linear relationship between BMI, WC, or WHR and risk of hypertension was investigated by modeling BMI, WC, or WHR exposure levels by using restricted cubic splines with three knots at percentiles 25, 50, and 75% of the distribution [34]. The P-value for nonlinearity was calculated by testing the null hypothesis that the coefficient of the second spline is equal to zero. All statistical analyses used Stata v12.1 (Stata Corp, College Station, TX, USA). All P values were two-sided, and the level of significance was set at < 0.05.

Heterogeneity was tested by Cochran Q and I2 statistics [35]. I2 values of ~25%, 50%, and 75% were considered to reflect low, moderate, and high heterogeneity, separately; P < 0.1 was considered statistically significant for the Q statistic. To explore the potential sources of heterogeneity, we conducted meta-regression (only for BMI analysis) and subgroup analyses, the variables such as sex, assessment of independent factors, duration of follow-up, geographic location, age, cases, publication year, study quality, baseline BMI, the adjusted confounders (age, gender, smoking, alcohol, physical activity, family history of hypertension, education), and potential intermediates (diabetes, hypercholesterolemia, dyslipidemia, systolic and diastolic blood pressure) in the analysis were included in the subgroup analysis and meta-regression model. We performed a sensitivity analysis by excluding one study at a time to assess the potential sources of heterogeneity and the stability of results. Potential publication bias was evaluated by the Egger’s test and Begg’s test [36, 37], and publication bias was indicated at P < 0.05.


Characteristics of studies

We identified 59 independents studies (45 publications) [5, 8, 9, 11, 17, 20,21,22, 24,25,26, 38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71] that included 57 (44 publications), 14 (10 publications) and 10 (7 publications) cohort studies in the analysis of BMI, WC and WHR and risk of hypertension incidence, respectively (Supplementary Table S4 and Fig. 1). Characteristics of the included studies are provided in Supplementary Table S4. Most of the studies were from Asia and America and were measured weight and height. Analyses of the quality of the studies yielded an average Newcastle-Ottawa Scale score of 8.0 (Supplementary Table S5).

Fig. 1
Fig. 1

Flow-chart of study selection

Body mass index

Fifty-seven cohort studies (44 publications) [5, 8, 11, 17, 20,21,22, 24,25,26, 38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71] were identified in the dose–response analysis of BMI and hypertension incidence and included 125071 incident cases among 830685 participants. Thirty-one studies were from Asia, 14 studies were from America, 8 were from Europe, 2 were from Australia and the remaining 2 were from Mauritius (Supplemental Table S4). The summary RR for per 5 units increment in BMI was 1.50 (95% CI: 1.40–1.59; I2 = 97.9%; Pheterogeneity < 0.0001; and it was similar when stratified by sex and geographic location (Fig. 2 and the Table 1). The risk of hypertension incidence was increased by 25% when the group baseline BMI < 25 kg/m2 (RR: 1.53; 95% CI: 1.38–1.69) than when group baseline BMI ≥ 25 kg/m2 (RR: 1.28; 95% CI: 1.18–1.39). In sensitivity analyses removing one study at a time, none of the individual studies changed the pooled risk substantially for the association of BMI with hypertension. There was no evidence of publication bias by Egger’s test (P = 0.298) or Begg’s test (P = 0.305) or by inspection of the funnel plot (Supplementary Figure S1A). Results from the cubic spline model suggested that there was no evidence of a non-linear association between BMI and hypertension (Pnonlinearity = 0.256; Fig. 3a). Therefore, we found a linear positively correlation at the risk of hypertension with BMI. The magnitude of the negative effect for hypertension incidence was substantially greater with the increasing of BMI. For example, for participants with BMI at 18.5 kg/m2, 25 kg/m2, and 30 kg/m2, the risk of hypertension was increased by 27% (RR: 1.27; 95% CI: 1.20–1.35), 107% (RR: 2.07; 95% CI: 1.34–2.46) and 213% (RR: 3.13; 95% CI: 2.49–3.93), respectively. Moreover, the results from cubic spline model showed that the shape of the linear association curve was steeper in the analysis restricted to studies of baseline BMI < 25 kg/m2 than BMI ≥ 25 kg/m2 (Fig. 4).

Fig. 2
Fig. 2

Linear relationship analysis of the risk of hypertension incidence with a 5-unit increase in BMI by sex. RR relative risk, CI confidence interval

Table 1 Subgroup analyses of BMI, waist circumference, and waist-to-hip ratio and hypertension
Fig. 3
Fig. 3

A–C Association between obesity indexes and hypertension incidence modeled by using restricted cubic splines (A results from BMI and hypertension; B results from WC and hypertension; C results from WHR and hypertension)

Fig. 4
Fig. 4

A, B Association between BMI and hypertension modeled by using restricted cubic splines (A results pooled for studies of BMI < 25 kg/m2; B results pooled for studies of BMI ≥ 25 kg/m2)

Waist circumference

Fourteen cohort studies (10 publications) [5, 8, 11, 24, 49, 52, 58, 61, 64, 70] were included in the analysis of WC and risk of hypertension and included 9898 cases among 94953 participants. Six cohort studies were from America, 4 from Asia, 2 from Europe and 2 from Mauritius. The summary RR for a 10-cm increase in WC was 1.25 (95% CI: 1.19–1.32; I2 = 56.7%; Pheterogeneity = 0.005), however, the risk of hypertension incidence was similar when stratified by sex, geographic location, and baseline BMI (Supplementary Figure S2 and the Table 1). The results from sensitivity analyses showed that removal of studies did not materially affect the overall risk estimates. There was an indication of publication bias with Egger’s test, P = 0.038, but not with Begg’s test, P = 0.743 (Supplementary Figure S1B). We found no evidence of a non-linear association between WC and hypertension risk (Pnonlinearity = 0.114; Fig. 3b). The RRs (95% CIs) of hypertension risk were 1.80 (1.30–2.47), 2.35 (1.55–3.57), and 3.02 (1.86–4.92) for 80, 90, and 100 cm, respectively. Only one study [64] provided sufficient data to conduct cubic spline model in the group of baseline BMI ≥ 25 kg/m2, thus it was not possible to reported the results by BMI.

Waist-to-hip ratio

Ten cohort studies (7 publications) [5, 8, 9, 49, 50, 58, 70] were included in the analysis of WHR and risk of hypertension and included 6998 cases in 50308 participants. Five studies were from America, 2 from Asia, 1 from Europe and 2 from Mauritius. The summary RR per 0.1-unit increment in WHR was 1.27 (95% CI: 1.18–1.37; I2 = 65.5%; Pheterogeneity = 0.0002, and it was similar when stratified by sex and baseline BMI, however, it seemed no significant association between WHR and the risk of hypertension in Asia. (Supplementary Figure S3 and Table 1). In sensitivity analyses removing one study at a time, none of the individual studies changed the pooled risk substantially for the association of WHR with hypertension. There was no evidence of publication bias with the Egger’s test (P = 0.06) or the Begg’s test (P = 1.00; Supplementary Figure S1C). There was no evidence of a non-linear association between WHR and hypertension incidence (Pnonlinearity = 0.056; Supplementary Figure 3C). The RRs (95% CIs) of hypertension risk were 1.25 (1.04–1.50), 1.59 (1.17–2.17), and 1.88 (1.31–2.72) for 0.82, 0.92, and 1.02, respectively.

Heterogeneity of the meta-analysis

To explore the sources of heterogeneity among these studies, we performed meta-regression (only for BMI analysis) and subgroup analyses by sex, assessment of independent factors, duration of follow-up, geographic location, age, cases, publication year, study quality, baseline BMI, the adjusted confounders, and potential intermediates in the analysis (Table 1). In general, the results were stable in the most analyses of BMI and hypertension, except the subgroups of assessment of independent factors (P = 0.018) and cases (P = 0.012) were found might be mainly attributable to variability in meta-regression. In the analyses of WC, WHR, and hypertension, we found that the heterogeneity seemed to be lower in man and assessment of independent factors by measured (Table 1). Besides, no significant changes of heterogeneity occurred in other subgroup analyses.


This is the first meta-analysis of cohort studies to quantify the relationship between multiple adiposity measures and the incidence of hypertension. We found a positive association between the risk of hypertension and BMI, WC and WHR, with a 50%, 25%, and 27% increase in the relative risk in BMI (per 5 units), WC (per 10 cm), and WHR (per 0.1 unit). Meanwhile, the positive associations between multiple adiposity measures and incident hypertension were observed across all geographic locations, suggesting that adiposity is a risk factor for hypertension across populations. Results from the cubic spline model showed that there were linear dose–response relationships with risk of hypertension incidence and BMI, WC, and WHR. Moreover, we found that the risk of hypertension associated with increased BMI was greater in the relative lean populations where the baseline BMI < 25 kg/m2 than in those with baseline BMI ≥ 25 kg/m2.

The results are consistent with a previous meta-analysis that included 1 cohort study and 54 cross-sectional studies, with the conclusion that the risk of hypertension associated with adiposity is greater in lean than in not-lean populations when the cut point of baseline BMI is 25 kg/m2[12]; however, the differences were not obvious in risk of hypertension and WC and WHR. In contrast to the previous meta-analysis, the present analysis included a much larger number of cases and participants (59 studies with more than 100,000 incident cases from diverse populations) and thus provides a much more accurate and reliable estimate of the association, in addition to a more comprehensive assessment of different adiposity measures in relation to the risk of hypertension. In conclusion, we find that BMI, WC, and WHR all have predictive value for the development of hypertension. However, owing to the higher I2 for BMI, tighter CI intervals and measuring easily for waist circumference, these results trend to suggest that the index of waist circumference is more commendable to be implemented by clinicians to predict the risk for hypertension.

The biological mechanisms linking adiposity to increased hypertension incidence is remain unclear. Several metabolic and neurohormonal pathways are likely to have complex interactions underlying the development of hypertension among the overweight and obese, including alterations in insulin resistance, the renin–angiotensin–aldosterone system and sympathetic tone [72]. Furthermore, early nutrition and growth, environment factors (including smoking, alcohol, physical activity, etc.), and potential intermediates (diabetes, SBP, DBP, hypercholesterolemia and dyslipidemia) also have a major impact on the development of hypertension [21, 43, 44, 48, 54, 73].

In our meta-analysis, the results showed a significant dose–response relation between different obesity indexes and the risk of hypertension, even within the normal range of them. We have also explored the non-linear relation in different subgroups such as sex and geographic location, which got similar results (results were not showed). However, previous studies have reported that normal weight lower the risk of hypertension than underweight and overweight/obesity [20, 65]. The results seemed contradictory. The greater rate of mortality in the lean population than not-lean population might account for the phenomenon. In addition, overt or latent hyperthyroidism associated with a low body weight might also increase risk of hypertension. Therefore, these results should be interpreted with caution. The better description of our results may be that in normal range of obesity indexes, as thinner as possible be the best suggestion to prevent hypertension incidence.

Moreover, the risk of hypertension associated with adiposity is greater and the shape of the linear dose–response curve was steeper in the group of baseline BMI < 25 kg/m2 than BMI ≥ 25 kg/m2. The result suggests that obesity have a stronger effect in relative lean population than not-lean population. The underlying mechanisms may be explained by the status of early insufficient fetal nutrition. In relative lean population, the adaptations caused by early insufficient nutrition that provide advantage in the short-term, however, chronic disease can be predisposed to individuals later in life, particularly if they are then exposed to excessive energy intake [74]. In addition, in the subgroup analysis we found stronger associations among the subgroup without adjusted the factors (family history of hypertension, SBP and diabetes) than with adjusted. Although the direction of association remained unchanged, these variables also can be considered as intermediate risk factors since partly increased the risk of hypertension.

To discover potential sources of heterogeneity, we performed meta-regression and various subgroup analyses, and the results generally supported our overall findings. In the analysis of BMI, the association did not vary by subgroups except with assessment of independent factors and number of cases. We found that the risk of hypertension in studies that used self-reported anthropometric measures were obviously higher than measured, and test for meta-regression was significant between these subgroups (P = 0.018). The difference may cause by recall bias, reported and measurement errors in the assessment of height and weight. These errors may exaggerate the effect of BMI to hypertension risk. Moreover, the different number of cases of included studies were another source of heterogeneity. As the sample size increased, the effect value decreased gradually. Sampling bias may play a core role in the difference. When the sample size is large enough, the effect value is close to the fact and CI intervals is tighter. In the studies of WC and WHR, we also found the source of potential heterogeneity, the main reasons may derive from diversity from model assumptions, clinical characteristics, and design features, and limited numbers of studies in WC and WHR analyses. Nevertheless, a similar association remained after adjustment for confounders. Hence, the effect of BMI, WC, and WHR on hypertension may be independent of the confounders.

In addition, our meta-analysis also contains several potential limitations that need to be mentioned. Firstly, the main limitation is the lower number of cohort studies available that report on the risk of hypertension with WC and WHR, which limit the ability to conduct subgroup analyses and cubic spline modeling (including stratification by baseline BMI). Second, the definitions of hypertension varied between studies, including different BP cut offs, diagnosed only by self-reports or using medication history, however, we did subgroup analysis found the risk of hypertension caused by obesity were similar. (data no shown). Thirdly, unhealthy diets are typically associated with obesity but very few studies adjusted for diet and energy intake; thus, these subgroup analyses are difficult to interpret.

Our meta-analysis has several strengths. Firstly, we conducted a complete literature search and minimized potential publication bias. Secondly, all included studies were cohort studies with the high study quality, which reduces the possibility for selection bias and avoids recall bias. Thirdly, we confirmed the robustness of our study findings by performing several sensitivity and subgroup analyses, and we observed no significant change in the magnitude or the direction of the effect for the association of obesity levels with hypertension risk. In addition, we conducted detailed relationship analyses, which clarified the shape of these relationships.


Our meta-analyses showed that BMI, WC, and WHR were all good indicators for an increased risk of hypertension, and the results from the cubic spline model suggest that risk of hypertension increases in a linear fashion with obesity. Future meta-analyses involving additional cohort studies are needed to explore the effect of early nutrition to obesity and hypertension incidence and explore the optimal dose of obesity indexes for hypertension prevention.


In conclusion, despite the largely unexplained heterogeneity of relative risk, our findings consistently confirmed the linear relationships between BMI, WC, and WHR and the risk of hypertension.

Summary Table

What is known about the topic?

  • To our knowledge, this is the first dose–response meta-analysis based on prospective studies to quantify the association between multiple adiposity measures and hypertension.

  • The studies with different adiposity measures and incremental characteristics were converted into consistent units (5 kg/m2 in BMI, 10 cm in WC, and 0.1 unit in WHR), which is appropriate to achieve data harmonization.

What this study adds?

  • In the dose–response analysis the risk of hypertension, we found a positive association between the risk of hypertension and BMI, WC, and WHR, with a 50%, 25%, and 27% increase in the relative risk in BMI (per 5 kg/m2), WC (per 10 cm), and WHR (per 0.1 unit).

  • Restricted cubic spline results showed that there were all linear association between multiple adiposity measures and risk of hypertension.


  1. 1.

    Ng M, Fleming T, Robinson M, Thomson B, Graetz N, Margono C, et al. Global, regional, and national prevalence of overweight and obesity in children and adults during 1980-2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet. 2014;384:766–81.

  2. 2.

    Whitlock G, Lewington S, Sherliker P, Clarke R, Emberson J, Halsey J, et al. Body-mass index and cause-specific mortality in 900 000 adults: collaborative analyses of 57 prospective studies. Lancet. 2009;373:1083–96.

  3. 3.

    Aune D, Sen A, Norat T, Janszky I, Romundstad P, Tonstad S, et al. Body mass index, abdominal fatness, and heart failure incidence and mortality: a systematic review and dose–response meta-analysis of prospective studies. Circulation. 2016;133:639–49.

  4. 4.

    Vazquez G, Duval S, Jacobs DR Jr, Silventoinen K. Comparison of body mass index, waist circumference, and waist/hip ratio in predicting incident diabetes: a meta-analysis. Epidemiol Rev. 2007;29:115–28.

  5. 5.

    Grootveld LR, Van Valkengoed IG, Peters RJ, Ujcic-Voortman JK, Brewster LM, Stronks K, et al. The role of body weight, fat distribution and weight change in ethnic differences in the 9-year incidence of hypertension. J Hypertens. 2014;32:990–6.

  6. 6.

    GBD 2015 Disease and Injury Incidence and Prevalence Collaborators. Global, regional, and national incidence, prevalence, and years lived with disability for 310 diseases and injuries, 1990-2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet. 2016;388:1545–602.

  7. 7.

    GBD 2015 Mortality and Causes of Death Collaborators. Global, regional, and national life expectancy, all-cause mortality, and cause-specific mortality for 249 causes of death, 1980-2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet. 2016;388:1459–544.

  8. 8.

    Fuchs FD, Gus M, Moreira LB, Moraes RS, Wiehe M, Pereira GM, et al. Anthropometric indices and the incidence of hypertension: a comparative analysis. Obes Res. 2005;13:1515–7.

  9. 9.

    Moliner-Urdiales D, Artero EG, Sui X, Espana-Romero V, Lee D, Blair SN. Body adiposity index and incident hypertension: the Aerobics Center Longitudinal Study. Nutr Metab Cardiovasc Dis. 2014;24:969–75.

  10. 10.

    Chei CL, Iso H, Yamagishi K, Tanigawa T, Cui R, Imano H, et al. Body fat distribution and the risk of hypertension and diabetes among Japanese men and women. Hypertens Res. 2008;31:851–7.

  11. 11.

    Williams PT, Hoffman K, La I. Weight-related increases in hypertension, hypercholesterolemia, and diabetes risk in normal weight male and female runners. Arterioscler Thromb Vasc Biol. 2007;27:1811–9.

  12. 12.

    Arabshahi S, Busingye D, Subasinghe AK, Evans RG, Riddell MA, Thrift AG. Adiposity has a greater impact on hypertension in lean than not-lean populations: a systematic review and meta-analysis. Eur J Epidemiol. 2014;29:311–24.

  13. 13.

    Seo DC, Choe S, Torabi MR. Is waist circumference ≥102/88cm better than body mass index ≥30 to predict hypertension and diabetes development regardless of gender, age group, and race/ethnicity? Meta-analysis. Prev Med. 2017;97:100–8.

  14. 14.

    World Health Organization. Obesity: preventing and managing the global epidemic of obesity, report of a WHO consultation. Geneva: World Health Organisation; 2000 (Report No: 894).

  15. 15.

    Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis of Observational Studies in Epidemiology (MOOSE) Group. JAMA. 2000;283:2008–12.

  16. 16.

    Ohnishi H, Saitoh S, Akasaka H, Mitsumata K, Chiba M, Furugen M, et al. Incidence of hypertension in individuals with abdominal obesity in a rural Japanese population: the Tanno and Sobetsu study. Hypertens Res. 2008;31:1385–90.

  17. 17.

    Heianza Y, Kodama S, Arase Y, Hsieh SD, Yoshizawa S, Tsuji H, et al. Role of body mass index history in predicting risk of the development of hypertension in Japanese individuals: Toranomon Hospital Health Management Center Study 18 (TOPICS 18). Hypertension. 2014;64:247–52.

  18. 18.

    Wells GSB, O’Connell D, Peterson J, Welch V, Losos M, Tugwell P. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. http://www.ohrica/programs/clinical_epidemiology/oxford.asp. 2014.

  19. 19.

    Niu J, Seo DC. Central obesity and hypertension in Chinese adults: a 12-year longitudinal examination. Prev Med. 2014;62:113–8.

  20. 20.

    Feng BY, Chen JC, Li Y, Huang JF, Li JX, Zhao LC, et al. Relationship between overweight/obesity and hypertension among adults in China: a prospective study. Zhonghua Liu Xing Bing Xue Za Zhi. 2016;37:606–11.

  21. 21.

    Zheng L, Zhang Z, Sun Z, Li J, Zhang X, Xu C, et al. The association between body mass index and incident hypertension in rural women in China. Eur J Clin Nutr. 2010;64:769–75.

  22. 22.

    Banda JA, Clouston K, Sui X, Hooker SP, Lee CD, Blair SN. Protective health factors and incident hypertension in men. Am J Hypertens. 2010;23:599–605.

  23. 23.

    Zhao D, Wang W, Liu J, Cheng J, Liu J, Qin LP. Association between body mass index and ten-year-accumulative-risk of hypertension. Zhonghua Liu Xing Bing Xue Za Zhi. 2009;30:435–8.

  24. 24.

    Shuger SL, Sui X, Church TS, Meriwether RA, Blair SN. Body mass index as a predictor of hypertension incidence among initially healthy normotensive women. Am J Hypertens. 2008;21:613–9.

  25. 25.

    Parikh NI, Pencina MJ, Wang TJ, Benjamin EJ, Lanier KJ, Levy D, et al. A risk score for predicting near-term incidence of hypertension: the Framingham Heart Study. Ann Intern Med. 2008;148:102–10.

  26. 26.

    Luo W, Guo Z, Hao C, Yao X, Zhou Z, Wu M, et al. Interaction of current alcohol consumption and abdominal obesity on hypertension risk. Physiol Behav. 2013;122:182–6.

  27. 27.

    Orsini N, Li R, Wolk A, Khudyakov P, Spiegelman D. Meta-analysis for linear and non-linear dose–response relations: examples, an evaluation of approximations, and software. Am J Epidemiol. 2012;175:66–73.

  28. 28.

    DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177–88.

  29. 29.

    Bekkering GE, Harris RJ, Thomas S, Mayer AM, Beynon R, Ness AR, et al. How much of the data published in observational studies of the association between diet and prostate or bladder cancer is usable for meta-analysis? Am J Epidemiol. 2008;167:1017–26.

  30. 30.

    Orsini N, Bellocco R, Greenland S. Generalized least squares for trend estimation of summarized dose–response data. Stata J. 2006;6:40–57.

  31. 31.

    Bagnardi V. Flexible meta-regression functions for modeling aggregate dose–response data, with an application to alcohol and mortality. Am J Epidemiol. 2004;159:1077–86.

  32. 32.

    Hartemink N, Boshuizen HC, Nagelkerke NJ, Jacobs MA, van Houwelingen HC. Combining risk estimates from observational studies with different exposure cutpoints: a meta-analysis on body mass index and diabetes type 2. Am J Epidemiol. 2006;163:1042–52.

  33. 33.

    Hamling J, Lee P, Weitkunat R, Ambuhl M. Facilitating meta-analyses by deriving relative effect and precision estimates for alternative comparisons from a set of estimates presented by exposure level or disease category. Stat Med. 2008;27:954–70.

  34. 34.

    Greenland S. Dose–response and trend analysis in epidemiology: alternatives to categorical analysis. Epidemiology. 1995;6:356–65.

  35. 35.

    Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. Br Med J. 2003;327:557–60.

  36. 36.

    Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. Bmj. 1997;315:629–34.

  37. 37.

    Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994;50:1088–101.

  38. 38.

    Park CS, Ha KH, Kim HC, Park S, Ihm SH, Lee H-Y. The association between parameters of socioeconomic status and hypertension in Korea: The Korean Genome and Epidemiology Study. J Korean Med Sci. 2016;31:1922.

  39. 39.

    Fujita M, Hata A. Sex and age differences in the effect of obesity on incidence of hypertension in the Japanese population: A Large Historical Cohort Study. J Am Soc Hypertens. 2014;8:64–70.

  40. 40.

    Talaei M, Sadeghi M, Mohammadifard N, Shokouh P, Oveisgharan S, Sarrafzadegan N. Incident hypertension and its predictors: the Isfahan Cohort Study. J Hypertens. 2014;32:30–8.

  41. 41.

    Schmiegelow MD, Andersson C, Kober L, Andersen SS, Norgaard ML, Jensen TB, et al. Associations between body mass index and development of metabolic disorders in fertile women—a nationwide cohort study. J Am Heart Assoc. 2014;3:e000672.

  42. 42.

    Sun Jiayi ZD, Wei Wang, Jing Liu, Jun Cheng, Lanping Qin. Association between body mass index and ten-year-accumulative-risk of hypertension. Chin J Epidemiol. 2009;30:435–8.

  43. 43.

    Jackson C, Herber-Gast G-C, Brown W. Joint effects of physical activity and BMI on risk of hypertension in women: A Longitudinal Study. J Obes. 2014;2014:1–7.

  44. 44.

    Thawornchaisit P, de Looze F, Reid CM, Seubsman SA, Sleigh AC, Thai Cohort Study T. Health risk factors and the incidence of hypertension: 4-year prospective findings from a national cohort of 60 569 Thai Open University students. BMJ Open. 2013;3:e002826.

  45. 45.

    Tsujimoto T, Sairenchi T, Iso H, Irie F, Yamagishi K, Tanaka K, et al. Impact of obesity on incident hypertension independent of weight gain among nonhypertensive Japanese: the Ibaraki Prefectural Health Study (IPHS). J Hypertens. 2012;30:1122–8.

  46. 46.

    Chien KL, Hsu HC, Su TC, Chang WT, Sung FC, Chen MF, et al. Prediction models for the risk of new-onset hypertension in ethnic Chinese in Taiwan. J Hum Hypertens. 2011;25:294–303.

  47. 47.

    Tirosh A, Afek A, Rudich A, Percik R, Gordon B, Ayalon N, et al. Progression of normotensive adolescents to hypertensive adults: a study of 26,980 teenagers. Hypertension. 2010;56:203–9.

  48. 48.

    Forman JP, Stampfer MJ, Curhan GC. Diet and lifestyle risk factors associated with incident hypertension in women. JAMA. 2009;302:401–11.

  49. 49.

    Hi Choy-LyeChei, Yamagishi Kazumasa, Tanigawa Takeshi, Cui Renzhe, Imano Hironori, Kiyama Masahiko, Kitamura Akihiko, Sato Shinichi, Shimamoto Takashi. Body fat distribution and the risk of hypertension and diabetes among Japanese men and women. Hypertens Res. 2008;31:851–7.

  50. 50.

    Nemesure B, Wu SY, Hennis A, Leske MC, BESS Group. The relationship of body mass index and waist-hip ratio on the 9-year incidence of diabetes and hypertension in a predominantly African-origin population. Ann Epidemiol. 2008;18:657–63.

  51. 51.

    Takase H, Dohi Y, Toriyama T, Okado T, Tanaka S, Sato K, Kimura G. Metabolic disorders predict development of hypertension in normotensive Japanese subjects. Hypertens Res. 2008;31:665–71.

  52. 52.

    Li Y, Zhai F, Wang H, Wang Z. A four-year prospective study of the relationship between body mass index and waist cirumstances and hypertension in Chinese adults. J Hyg Res. 2007;36:478–80.

  53. 53.

    Gu D, Wildman RP, Wu X, Reynolds K, Huang J, Chen CS, et al. Incidence and predictors of hypertension over 8 years among Chinese men and women. J Hypertens. 2007;25:517–23.

  54. 54.

    Gelber RP, Gaziano JM, Manson JE, Buring JE, Sesso HD. A prospective study of body mass index and the risk of developing hypertension in men. Am J Hypertens. 2007;20:370–7.

  55. 55.

    Lee SH, Kim YS, Sunwoo S, Huh BY. A retrospective cohort study on obesity and hypertension risk among Korean adults. J Korean Med Sci. 2005;20:188–95.

  56. 56.

    Radi S, Lang T, Lauwers-Cances V, Chatellier G, Fauvel JP, Larabi L, et al. One-year hypertension incidence and its predictors in a working population: the IHPAF study. J Hum Hypertens. 2004;18:487–94.

  57. 57.

    Hu G, Barengo NC, Tuomilehto J, Lakka TA, Nissinen A, Jousilahti P. Relationship of physical activity and body mass index to the risk of hypertension: a prospective study in Finland. Hypertension. 2004;43:25–30.

  58. 58.

    Folsom AR, Kushi LH, Anderson KE, Mink PJ, Olson JE, Hong CP, et al. Associations of general and abdominal obesity with multiple health outcomes in older women: the Iowa Women’s Health Study. Arch Intern Med. 2000;160:2117–28.

  59. 59.

    He J, Klag MJ, Appel LJ, Charleston J, Whelton PK. Seven-year incidence of hypertension in a cohort of middle-aged African Americans and whites. Hypertension. 1998;31:1130–5.

  60. 60.

    Huang Z, Willett WC, Manson JE, Rosner B, Stampfer MJ, Speizer FE, et al. Body weight, weight change, and risk for hypertension in women. Ann Intern Med. 1998;128:81–8.

  61. 61.

    Hunt SC, Stephenson SH, Hopkins PN, Williams RR. Predictors of an increased risk of future hypertension in Utah. A screening analysis. Hypertension. 1991;17(6 Pt 2):969–76.

  62. 62.

    Ukawa S, Tamakoshi A, Wakai K, Ando M, Kawamura T. Body mass index is associated with hypertension in Japanese young elderly individuals: findings of the new integrated suburban seniority investigation. Intern Med. 2015;54:3121–5.

  63. 63.

    Kubo T, Fujino Y, Nakamura T, Kunimoto M, Tabata H, Tsuchiya T, et al. An industry-based cohort study of the association between weight gain and hypertension risk among rotating shift workers. J Occup Environ Med. 2013;55:1041–5.

  64. 64.

    Bombelli M, Facchetti R, Sega R, Carugo S, Fodri D, Brambilla G, et al. Impact of body mass index and waist circumference on the long-term risk of diabetes mellitus, hypertension, and cardiac organ damage. Hypertension. 2011;58:1029–35.

  65. 65.

    Matsuo T, Sairenchi T, Suzuki K, Tanaka K, Muto T. Long-term stable obesity increases risk of hypertension. Int J Obes. 2011;35:1056–62.

  66. 66.

    Panagiotakos DB, Chrysohoou C, Pitsavos C, Skoumas J, Lentzas Y, Katinioti A, et al. Hierarchical analysis of anthropometric indices in the prediction of 5-year incidence of hypertension in apparently healthy adults: The ATTICA Study. Atherosclerosis. 2009;206:314–20.

  67. 67.

    Chuang SY, Chou P, Hsu PF, Cheng HM, Tsai ST, Lin IF, et al. Presence and progression of abdominal obesity are predictors of future high blood pressure and hypertension. Am J Hypertens. 2006;19:788–95.

  68. 68.

    Hayashi T, Boyko EJ, Leonetti DL, McNeely MJ, Newell-Morris L, Kahn SE, et al. Visceral adiposity is an independent predictor of incident hypertension in Japanese Americans. Ann Intern Med. 2004;140:992–1000.

  69. 69.

    Ishikawa-Takata K, Ohta T, Moritaki K, Gotou T, Inoue S. Obesity, weight change and risks for hypertension, diabetes and hypercholesterolemia in Japanese men. Eur J Clin Nutr. 2002;56:601–7.

  70. 70.

    Nyamdorj R, Qiao Q, Soderberg S, Pitkaniemi J, Zimmet P, Shaw J, et al. Comparison of body mass index with waist circumference, waist-to-hip ratio, and waist-to-stature ratio as a predictor of hypertension incidence in Mauritius. J Hypertens. 2008;26:866–70.

  71. 71.

    Li M, McDermott R. Obesity, albuminuria, and gamma-glutamyl transferase predict incidence of hypertension in indigenous Australians in rural and remote communities in northern Australia. J Hypertens. 2015;33:704–9.

  72. 72.

    DeMarco VG, Aroor AR, Sowers JR. The pathophysiology of hypertension in patients with obesity. Nat Rev Endocrinol. 2014;10:364–76.

  73. 73.

    Zhou L, Li Y, Guo M, Wu Y, Zhao L. 61-Relations of body weight status in early adulthood and weight changes until middle age with hypertension in the Chinese population. Hypertens Res. 2016;39:913–8.

  74. 74.

    Power C, Kuh D, Morton S. From developmental origins of adult disease to life course research on adult disease and aging: insights from birth cohort studies. Annu Rev Public Health. 2013;34:7–28.

Download references

Author information

Author notes

  1. These authors contributed equally: Wen Zhou, Yuanyuan Shi.


  1. Department of Epidemiology and Health Statistics, College of Public Health, Zhengzhou University, 100 Kexue Avenue, Zhengzhou, Henan, 450001, China

    • Wen Zhou
    • , Yuanyuan Shi
    • , Zhiguang Ping
    • , Chongjian Wang
    • , Xuejiao Liu
    • , Jie Lu
    • , Zhen-xing Mao
    • , Dongdong Zhang
    • , Zhongyan Tian
    • , Lulu Zhang
    •  & Linlin Li
  2. Department of Clinical Pharmacology, School of Pharmaceutical Science, Zhengzhou University, 100 Kexue Avenue, Zhengzhou, Henan, 450001, China

    • Yu-qian Li
  3. Department of Endocrinology, Military Hospital of Henan Province, Zhengzhou, 18 Jinshui Road, Henan, 450003, China

    • Jingzhi Zhao
    •  & Lei Yin


  1. Search for Wen Zhou in:

  2. Search for Yuanyuan Shi in:

  3. Search for Yu-qian Li in:

  4. Search for Zhiguang Ping in:

  5. Search for Chongjian Wang in:

  6. Search for Xuejiao Liu in:

  7. Search for Jie Lu in:

  8. Search for Zhen-xing Mao in:

  9. Search for Jingzhi Zhao in:

  10. Search for Lei Yin in:

  11. Search for Dongdong Zhang in:

  12. Search for Zhongyan Tian in:

  13. Search for Lulu Zhang in:

  14. Search for Linlin Li in:

Conflict of interest

The authors declare that they have no conflict of interest.

Corresponding author

Correspondence to Linlin Li.

Electronic supplementary material

About this article

Publication history