Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

General and abdominal obesity parameters and their combination in relation to mortality: a systematic review and meta-regression analysis


Epidemiological studies assessing general and abdominal obesity measures or their combination for mortality prediction have shown inconsistent results. We aimed to systematically review the associations of body mass index (BMI), waist-to-hip ratio (WHR), waist circumference (WC) and waist-to-height ratio (WHtR) with all-cause mortality in prospective cohort studies. In this systematic review, which includes a meta-regression analysis, we analysed the associations with all-cause mortality of BMI, WHR, WC and WHtR in prospective cohort studies available in Medline, Embase, the Cochrane Database of Systematic Reviews and Esbiobase from inception through 7 May 2010. A total of 18 studies met the inclusion criteria, comprising 689 465 participants and 48 421 deaths during 5–24 years of follow-up. The studies were heterogeneous, mainly due to differences in categorization of anthropometric parameters (AP) and different approaches to statistical analysis. Both general and abdominal obesity measures were significantly associated with mortality. In analyses using categorical variables, BMI and WC showed predominantly U- or J-shaped associations with mortality, whereas WHR and WHtR demonstrated positive relationships with mortality. All measures showed similar risk patterns for upper quantiles in comparison to reference quantiles. The parameters of general and abdominal obesity each remained significantly associated with mortality when adjusted for the other. This evidence suggests that abdominal obesity measures such as WC or WHR, show information independent to measures of general obesity and should be used in clinical practice, in addition to BMI, to assess obesity-related mortality in adults.


Obesity is a major public health concern worldwide.1 According to a recent estimate, there are 1.46 billion overweight adults (body mass index (BMI) 25 kg/m2) and 503 million of them are obese (BMI 30 kg/m2).2 As obesity is a major risk factor for chronic diseases and mortality,1, 3, 4, 5 accurate and simple measures are essential to refine the descriptive and analytical epidemiology of this condition. BMI, which has been shown to be associated with mortality in large meta- and pooled analyses,6, 7, 8, 9 is widely used as the current standard measure of obesity in clinical practice. However, BMI does not discriminate between fat and lean mass. Nor does it address body fat distribution.10, 11 Recently, abdominal obesity and fat mass were shown to be associated with cardiovascular morbidity and mortality.5, 11, 12 Abdominal obesity measures such as waist circumference (WC) and waist-to-hip ratio (WHR) were shown to more accurately measure body fat distribution.11 Both WC and WHR are closely related to adverse metabolic profiles and their resulting diseases and both are independent predictors of mortality.13, 14, 15, 16, 17 Some have argued that WC, WHR and the recently proposed waist-to-height ratio (WHtR) may be more precise than BMI in assessing obesity-related health burden, including total mortality, risk of type 2 diabetes and cardiovascular disease.5, 18, 19, 20 However, a recent meta-analysis of individual data showed that BMI and abdominal parameters had similar strengths of association with cardiovascular disease risk.21 For all-cause mortality, large observational studies reported controversial results whether abdominal adiposity measures are more strongly predictive of mortality than BMI.12, 19, 22, 23, 24, 25, 26, 27 Therefore, the objective of this review was to systematically compare the associations of both single and combined anthropometric parameters (APs) of obesity with all-cause mortality. Whereas former systematic reviews on this topic have mostly been limited to specific diseases, we focused on population-based cohorts to increase generalizability. As body distribution is likely to be a relevant factor in the obesity–mortality relationship,16, 28 we systematically reviewed combined effects of two APs on mortality to broaden current knowledge and literature. To achieve this, we analysed and compared cohort studies investigating BMI and at least one abdominal obesity measure, such as WC, WHR or WHtR.


Data sources and searches

We searched Medline, Embase, the Cochrane Database of Systematic Reviews and Esbiobase from inception through 7 May 2010 (search strategy: Supplementary Table 1). References from all relevant articles were hand-searched to identify additional relevant studies. We performed our systematic review according to the MOOSE reporting guidelines.29

Study selection

Two reviewers (SC and MHF) independently analysed citations at the levels of title, abstract and full text. We included both (i) prospective cohort studies and observational follow-up studies of non-intervention groups of community-based randomised controlled trials, which analysed (ii) all-cause mortality as an outcome, (iii) reported BMI and at least one abdominal AP (WC, WHR or WHtR) as exposure variables at baseline, (iv) had a sample size of at least 4000 participants, (v) a follow-up of at least 3 years and (vi) participants with a baseline age of at least 18 years. We excluded articles examining pregnant women, terminally ill or frail patients, groups with specific diseases or residents in nursing homes. Due to differences in the AP–mortality relationships between the various ethnic groups in both general and abdominal obesity parameters,30, 31, 32 we excluded studies with <75% Caucasian participants in order to reduce complexity of the analyses. We also excluded cross-sectional analyses, case reports, comments, letters, reviews and study samples that combined analyses with another risk factor of mortality. For each cohort study, we selected the publication with the longest follow-up. We calculated reviewer agreement on citation selection on the abstract and full-text levels using kappa statistics.33

Data extraction and quality assessment

For each article, we determined the study design; year of publication; country; study population; inclusion and exclusion criteria; assessment methods for the AP and outcome; follow-up time; number and percentage of patients lost to follow-up; methods used for statistical analysis, including whether AP was analysed as a categorical or continuous variable, and whether participants with early mortality were excluded. We extracted total numbers of participants, deaths and person-years for the total population and for each category of AP (if available); sex and age range; and relative risk (RR) measures, such as hazard ratios or RRs. For simplicity and ease of presentation, hazard ratios will also be referred to as RR. We also show 95% confidence intervals (CIs) for risk categories within each AP (if analysed categorically), or for each AP as a whole (if analysed continuously). Some studies provided cross-classifications of two APs, stratifying their sample into predefined or population-based quantiles of BMI and these into quantiles of WC or WHR. These cross-classifications are used to describe different body configuration (for example, low BMI and high WC). We collected all variables for which RRs were adjusted. If multiple statistical models were provided, we used the model with the highest degree of adjustment. We differentiated between multivariate models with and without adjustment for another AP, and included both models in the analysis, if available. All results are based on a 5% significance level.

Two reviewers independently assessed the methodological quality of reporting of all included articles (SC and MHF) using the checklist provided by Downs and Black34 modified by the STROBE criteria.35 We excluded publications with <12 out of 20 (60%) questions answered positively. This is a commonly used cutoff for assessing methodological quality.36

Data synthesis and analysis

We show the results of individual studies in the tables. To assess publication bias, we transformed category-specific RR into estimates of the total RR using the generalized least-squares for trend estimation method.37 For these calculations, we assumed a linear relationship between the natural logarithm of risk ratio and increasing AP. Because relationships between logRR and AP are not necessarily linear, we initially looked at quadratic relationships, which could allow, for example, for J-shaped associations. Using model reduction strategies in order to obtain a parsimonious model lead to the linear relationship to fulfil the above named assumption. We also assessed publication bias using the regression method by Egger38 with estimated regression coefficients and corresponding standard errors.

We calculated summary RR estimates and 95% CIs using maximum-likelihood methods based on linear mixed models, with the log of RR as dependent variable. Our meta-regression aims to relate effect size to one or more characteristics of the studies involved. It uses study-specific covariates and must be distinguished from regression analyses where individual data on outcomes and covariates are available. One purpose of meta-regression is to explain heterogeneity between studies using study-specific covariates. In practice, linear mixed effects models are often used for analysis in order to allow for variability between studies. We performed meta-regressions for every AP, using them as a covariate, including a quadratic term in order to allow for possible nonlinear associations. In order to estimate association strength in meta-regressions, we included only studies reporting information on several categories of AP. We standardized covariates for ease of interpretation with BMI at 24.9 kg/m2, and WC and WHR at the 50th percentile of participants in the EPIC cohort39 (WC: 80 cm for females, 95 cm for males; WHR: 0.8 for females, 0.95 for males). This large, population-based cohort allows us to compare relative mortality risks to the mean of a representative European sample. As this is a meta-regression, the mean BMI was chosen in order to centre the covariate BMI. Doing so, leaves an unbiased estimate and interpretation of the regression’s model intercept. We used a zero intercept model in which the standardized covariate, a BMI of 24.9 kg/m2, for example, corresponds to a RR of one. We based these regressions on mixed effects models to account for the correlation between different levels of AP. In order to obtain parsimonious models, we evaluated and selected models using the likelihood-ratio test and Bayesian information criterion. The complexity of data, including the variety of statistical analyses, prevented presentation of forest plots. All statistical analyses were conducted with the statistical package SAS version 9.2 (SAS Institute, Cary, NC) using a two-sided significance level of 0.05.


Out of 2575 citations from four databases, we identified 18 studies meeting the inclusion criteria in a systematic literature search (Figure 1). Reviewer agreement on selection of abstracts for full-text evaluation was 68% (κ=0.58, P<0.001) and for inclusion of articles in the final review was 94% (κ=0.80, P<0.001). All studies passed the predefined cutoff of at least 60% of the reporting quality criteria (Supplementary Table 2). Performing the regression method by Egger, the estimated regression coefficient for BMI was −3.98 (P=0.246), 0.39 (P=0.80) for WC and 2.15 (P=0.249) for WHR. Thus, we found no statistically significant publication bias.

Figure 1

Flow-chart for the literature search and results.

Description of included studies

Table 1 summarizes the 18 studies, including a total of 693 739 participants and follow-up times between 5 and 24 years. Of these, 13 studies were population-based prospective cohort studies,10, 16, 19, 24, 25, 26, 27, 40, 41, 42, 43, 44, 45 3 were follow-ups of the non-intervention group of community-based randomized controlled trials,46, 47, 48 and 2 studies reported pooled analyses of multiple cohort studies.12, 49 Four publications reported complementary data from two different cohorts (Rotterdam-Cohort42, 43 and Nurses Health Study41, 44). Two studies provided only data stratified by smoking status42, 47 and one study was limited to stratified data by age.26 We only included these studies in subgroup analyses, thus leaving 15 studies (14 cohorts) for the main analyses.10, 12, 16, 19, 24, 25, 27, 40, 41, 43, 44, 45, 46, 48, 49 The studies reported higher correlations between BMI and WC (Pearson correlation coefficient r, range 0.77–0.80)12, 26 and between BMI and WHtR (range 0.81–0.90)12, 40 than between BMI and WHR (range 0.31–0.55).12, 16

Table 1 Design of included cohorts

Associations of APs with all-cause mortality: qualitative analysis

We distinguished between analyses with and without adjustment for a second AP. Studies in our main analysis (n=15) chose various statistical approaches to assess the AP–mortality relationship. Nine studies analysed APs as categorical variables (Table 2), four as continuous variables10, 25, 45, 49 and two used both approaches12, 44 (Supplementary Table 3). Category cutoffs, including the reference category of AP, differed substantially between studies (Table 2).

Table 2 Associations of APs with all-cause mortality: analyses without adjustment for a second AP

Models without adjustment for a second AP

All studies that analysed BMI as a categorical variable showed a U-shape association with mortality.16, 19, 24, 27, 40, 41, 43, 44, 46, 48 In comparison to the individual reference category (Table 2) of each study, high BMI categories were significantly associated with higher mortality in four studies16, 27, 40, 41 and showed no significant association in five studies.19, 24, 43, 46, 48 Within studies analysing participants’ AP as continuous variables (Supplementary Table 3), there was no significant association of BMI with mortality in three studies12, 25, 49 or the male group of the fourth study. Within the female group, there was a significant positive relationship.10 One study showed significantly decreased mortality risk with increasing BMI.45 WHR had significant positive association with all-cause mortality in both categorical and continuous analyses.10, 16, 19, 24, 25, 44, 48, 49 As an exception, a male subpopulation10 and the Health Professional Follow-up Study showed insignificant positive correlation after multivariate adjustment.27 WC demonstrated a significant positive association with mortality in two studies44, 49 and two female subgroups10, 25 analysing WC as a continuous variable and showed a J-shaped association when analysed as a categorical variable.16, 24, 27, 40, 44 However, the WC–mortality relationship was not significant in four studies and in two gender-specific subgroups for both categorical and continuous analyses of AP10, 12, 19, 25, 43, 46 and was negative in one study45 (Table 2, Supplementary Table 3). WHtR was analysed for mortality relationship in four studies.12, 25, 40, 49 Of these, three studies and the female group of the fourth study demonstrated a monotonic positive association between WHtR and mortality,12, 25, 40, 49 whereas the male group of the fourth study showed a nonsignificant positive tendency25 (Table 2, Supplementary Table 3).

Comparison of selected APs

The studies chose different strategies to check selected APs for mortality prediction. Five studies calculated RR for each s.d. increase in BMI and abdominal AP10, 12, 25, 45, 49 and one study did so for WHR and WC.44 Predominantly, these results showed either no significant or a significant lower mortality risk per s.d. increase of BMI, whereas abdominal parameters were associated with a significant increased mortality risk per s.d. increase (Supplementary Table 3). However, one female subgroup reported a significant positive association with mortality for BMI,10 so generalization of these results is limited. Furthermore, with WC as the abdominal AP, one study45 reported a significant negative association and two male subgroups10, 25 showed no significant association with mortality. In addition, the males of the NFPS cohort showed no significant association between WHtR and mortality.25

Two other studies compared the predictive values for all-cause mortality. They performed receiver operating curve (ROC) analyses and calculated respective areas under the ROC curve (AUC) for prediction of all-cause mortality.12, 25 Both reported significantly better discrimination for abdominal obesity parameters compared with BMI. There was no significant difference between WHtR, WC and WHR in the AUC12, 25 (Supplementary Table 3).

The seven studies analysing AP as categorical variables grouped their samples into quantiles by AP for both BMI and abdominal AP.12, 19, 24, 40, 42, 46, 47 However, two of them reported only data stratified by smoking status42, 47 and one adjusted for another AP.12 Thus, we used the remaining four studies to compare mortality risks in each AP quintile.19, 24, 40, 46 In these, the fifth quintile of the abdominal obesity parameter predominantly showed significant increased mortality risk compared with the individual reference category,19, 24, 40 whereas BMI showed no significant difference between the fifth quintile and the reference category19, 24, 46 (Table 2). However, there was little difference between the RR estimates of the fifth quintile and the reference of any APs.

To summarize, models without adjustment for a second AP showed heterogeneous results, especially for BMI and WC when analysed as categorical or continuous variables. However, they reported similar values for the RR estimates in higher categories of AP and elevation of risk per s.d. increase of AP.

Combined associations of two APs with mortality

To describe combined associations of two AP with mortality, (i) six studies adjusted their risk estimates to include another AP (Table 3a) by controlling for BMI in abdominal obesity measures and vice versa and (ii) five studies analysed interaction of BMI and abdominal AP on mortality by cross-classification of BMI with WC or WHR, respectively (Table 3b). In studies using the first approach, and reporting higher mortality risks for high BMI without controlling for a second AP, values decreased after adjusting for abdominal obesity.12, 45 In contrast, WC, WHR and WHtR showed a significant increase in association with mortality when adjusted for BMI.12, 16, 19, 44, 45, 48 Studies providing both WC and WHR quintiles, adjusted for BMI, showed similar RRs for each abdominal obesity parameter.12, 16, 44

Table 3a Association of APs with all-cause mortality: analyses with adjustment for a second AP
Table 3b Association of APs with all-cause mortality: cross-classification of BMI and one abdominal AP for mortality risk

In cross-classification analysis for all age groups, the highest relative mortality risks were shown for the combination of low BMI with large WC or WHR.16, 19 In studies not undertaking age-stratification, participants with normal BMI and high WC or WHR had a significantly increased risk compared with participants of the reference category, but were not those at highest risk.16, 19 They had even a lower or equal mortality risk compared with those participants with high BMI and high WC or WHR. Participants in the middle or upper third of BMI values with a small WC or WHR had the lowest relative mortality risk. Two cohorts analysed either older participants (65 years old)45 or stratified the cohort according to age (cutoff at 65 years) before performing the cross-classification.26 As for age <65 years, participants in the lowest BMI category and a high WHR had the highest mortality risk.26 The lowest mortality risk in younger people had the group with normal BMI and low WHR. In older participants, the highest death rate was seen for participants with a normal-to-moderately elevated BMI and a low-to-moderately high WC.26, 45 In summary, the combination of BMI with one abdominal parameter improved the prediction of mortality based on RR (Tables 3a and b). However, as categorization varied between the studies, the comparability of these results is limited.16, 19, 26, 45

Meta-regression analyses

We performed meta-regression analysis on the nine cohorts that provided both RR and 95% CIs and defined category boundaries for each AP.16, 19, 24, 26, 27, 41, 43, 44, 46, 48 We chose random effects models to calculate effects in pooled data. The between-study variance shows heterogeneity between the studies (τ=0.04 for BMI, τ=0.04 for WHR and τ=0.03 for WC, Table 4). BMI and WC showed a U- or J-shaped positive association with mortality, whereas WHR demonstrated a nearly monotonic-positive association (Figure 2). The model-based estimates for RR in the included studies were 1.05 (95% CI: 1.03–1.08) for a BMI of 30 kg/m2 in comparison to one of 25 kg/m2. Comparing a BMI of 35 kg/m2 to a BMI of 25 kg/m2 gave a RR estimate of 1.27 (95% CI: 1.21–1.33; Table 5). For a WC of 100 cm compared with 80 cm in women and WC of 115 cm compared with 95 cm in men, the model-based RR estimate was 1.32 (95% CI: 1.22–1.43). For a WHR of 0.85 compared with 0.8 in females and a WHR of 1.0 compared with 0.95 in males, the model-based RR estimate was 1.13 (95% CI: 1.11–1.59, Supplementary Table 4).

Table 4 Global model fit of used mixed-models
Figure 2

(a–c) Pooled relative mortality risks of dose-response data of BMI (a), WHR (b) and WC (c) by meta-regression analysis from included observational studies. BMI, body mass index, RR, relative mortality risk; WC, waist circumference; WHR, waist-to-hip ratio; ♀ for women; ♂ for men.

Table 5 Pooled risk estimates from the random effects models

Stratified analyses

We conducted separate analyses of studies, stratifying by age, gender and/or smoking status to address the heterogeneity of study results. However, we did not analyse WHtR this way, as only one study investigating WHtR provided stratified analyses.40

Six studies analysed the AP–mortality relationships by age groups (<65 and 65 years old).16, 26, 27, 41, 44, 45, 46 However, because the EPIC cohort adjusted WHR and WC for BMI in its stratified analyses, we considered only analyses of BMI from this cohort for our stratified analyses. Qualitatively, in younger participants (<65 years old), BMI, WC and WHR showed significant positive association with mortality in three of four cohorts.16, 26, 27, 41, 44 Reis and colleagues reported a marginal significant positive association of WHR with mortality, but found no significant relationship with mortality for BMI or WC.26 In older persons ( 65 years), four26, 27, 45, 46 of five studies16, 26, 27, 45, 46 reported either a negative or statistically insignificant relationships for BMI, WC and WHR with mortality (Supplementary Table 5). Only the EPIC cohort reported significantly increased mortality associated with a high BMI in older adults.16 Quantitatively, a meta-regression analysis of five qualifying cohorts16, 26, 27, 41, 44, 46 showed no significant age effect on the BMI–mortality relationship (RR: 0.99 (95% CI: 0.94–1.03)). The relative mortality risk for both WC (RR: 0.77 (95% CI: 0.65–0.90)) and WHR (RR: 0.47 (95% CI: 0.35–0.63)) was lower in older than in younger people.

In both sexes, BMI, WHR and WC exhibited similar patterns of association with mortality.10, 16, 19, 24, 25, 26, 27, 40, 41, 44, 45, 46, 48 A meta-regression analysis of eight qualifying cohorts showed no statistically significant difference in the AP–mortality relationship between males and females (Supplementary Table 4).16, 19, 24, 26, 27, 41, 44, 46, 48 Qualitative analyses, stratified by smoking status, showed heterogeneous results (Supplementary Table 6).16, 24, 41, 42, 44, 46, 47 In one16 of four cohorts analysing current smokers,16, 24, 42, 47 the relative mortality risk associated with a high BMI as compared with reference BMI category was significantly higher among people who had never smoked than among current smokers. But this was not the case in other study samples. All studies16, 24, 42, 47 investigating current smokers reported the highest RR for current smokers with low BMI compared with the reference category, although this was only significant in one study16 and a male subcohort.24 For abdominal AP, studies reported heterogeneous associations with mortality in never smokers and current smokers.42, 46, 47


This systematic review analyzes and compares the associations of general and abdominal obesity measures with all-cause mortality for adults of all ages. The difference between the associations of BMI, WC, WHR and WHtR with total mortality was minor and may not be clinically relevant. No single AP was shown to be clearly superior to the others.

Comparison with current literature

Although other joint efforts such as the Prospective Studies Collaboration and the National Cancer Institute Cohort Consortium6, 9 were limited to the associations of mortality risk with BMI only, this systematic review additionally investigated abdominal obesity. Contrarily to our findings, Seidell et al.23 concluded in a previous non-systematic review that WC and WHR may be better predictors of all-cause mortality compared with BMI on the basis of a narrative approach. In addition, Czernichow et al.17 reported in a recent meta-analysis of individual-participant data of nine British cohort studies a more consistent association for WC and WHR with total mortality, whereas BMI showed a J-shaped association in multivariate analyses. However, in their sub-analysis of the AUC and the relative integrated discrimination improvement no major differences in predictive values for mortality between BMI, WC and WHR were shown,17 which corresponds to our results.

Other systematic reviews investigated the AP relationship with cardiovascular outcomes. The Emerging Risk Factors Collaboration showed similar strengths of association for general and abdominal obesity parameters for cardiovascular disease.21 Although systematic reviews on outcomes such as hypertension, dyslipidemia or diabetes detected modestly higher relative cardiovascular morbidity risks for WC, WHR and WHtR than for BMI, the difference was either not statistically significant50, 51 or did not seem clinically relevant.52

Different statistical approaches for analysing mortality prediction

Contradictory conclusions across different studies and reviews could be attributed to heterogeneity in chosen (i) study population, (ii) statistical analysis and (iii) measurement of AP (refer to section limitations). The studies included in our review analysed APs either (i) as continuous variables (Supplementary Table 3),10, 12, 25, 44, 45, 49 or (ii) via ROC analyses12, 25 or (iii) by categorizing APs by their distribution within the study sample.19, 24, 40, 46 For approaches (i) and (ii), studies tended to confirm a better mortality prediction for abdominal AP than for BMI. However, for (i), all selected studies10, 12, 25, 44, 45, 49 included BMI as a linear variable in their continuous model. This approach assumes a linear association between abdominal obesity parameters and mortality, but does not take nonlinear (for example, U-shaped) associations into account. The ROC analysis methods of (ii) do not consider the generally low-risk distribution of population-based cohorts.53, 54 Regardless, the AUC differences found might be too small to be clinically relevant. Therefore, we treated results of both statistical approaches (i and ii) with caution. Only four studies provided data for the third approach (iii). All four reported comparable positive relative mortality risks for WHR, WC and WHtR in the highest quantile, but only one of them40 for BMI.19, 24, 40, 46 However, the differences in relative mortality risks between abdominal AP and BMI seemed to be minor, with BMI results just narrowly missing significance. This indicates that the difference between BMI and abdominal AP may not be clinically relevant. Although here, it should be noted that our study was not designed to identify optimal BMI or other AP values associated with the lowest mortality risk.

Combined associations of different APs

We evaluated the impact of the association with total mortality risk for two APs combined. All eligible studies reported that, when adjusted for abdominal obesity, BMI showed either a significant negative19, 45, 48 or an insignificant association12 with all-cause mortality, whereas WHR,12, 16, 19, 44, 48 WC12, 16, 44, 45 and WHtR12, 44 showed a significant positive association with mortality when controlled for BMI. This may indicate that combining abdominal obesity measures with BMI may more accurately assess obesity-related mortality risk. A recently published Canadian study confirmed this interpretation.55 Interaction between BMI and abdominal AP with mortality risk was analysed by several studies performing cross-classification of BMI quantiles and quantiles of WC or WHR, respectively. The highest risk was shown for participants in the lowest BMI category having a median-to-large WC or WHR.16, 19, 26, 45 Combination of BMI and an abdominal AP may provide insight into risk-types of body fat distribution for research and clinical practice. However, evidence is limited by inconsistent classification of examined APs and too few studies investigating the correlation of combined APs with mortality risk. Clearly, further research on the relationship between multiple APs and obesity-related mortality risk is needed.


We performed a meta-regression analysis to quantify the effect of age on the association between AP and mortality. As only three of the studies selected provided stratified analyses adjusted for a second AP, we restricted the meta-regression analysis to studies without relative mortality risk control for a second AP. We detected no significant age effect on the BMI–mortality relationship in our meta-regression analysis comparing older (65 years) to younger adults (<65 years). Other studies, which only assessed BMI for mortality prediction, and a recent systematic review by Chang et al., showed a lower relative mortality risk for high BMI in older adults compared with younger adults.28, 56, 57, 58 In our qualitative analysis, we detected either nonsignificant or significant negative mortality relationships for BMI, WC and WHR in all16, 26, 27, 45, 46 except one16 studies investigating older adults. Contrarily to these studies and our results, a recent systematic review by Donini et al.59 reported a higher mortality risk for higher BMI levels in older adults (65 years old). The authors considered WC as potentially equal to BMI for mortality risk prediction with a tendency to prefer WC over BMI for higher grades of obesity in older age. However, in agreement with Donini et al., our results showed a higher mortality risk for older adults with low BMI. For abdominal obesity parameters, in our meta-regression analyses, the relative mortality risk associated with WC or WHR was lower in older (65 years old) than younger adults (<65 years old). Chang et al. argued that inconsistent conclusion in literature about AP–mortality associations may be due to, among other factors, different methods of categorization.28

There are several possible explanations for differing associations of obesity with mortality between younger and older adults. Undiagnosed subclinical chronic diseases are associated with decreasing body weight, which may indirectly decrease, and therefore, undervalue the RR of higher BMI levels. Conversely, obesity might provide energy stores during periods of stress such as diseases or trauma.60, 61 Findings might also be influenced by a ceiling effect due to the higher absolute mortality of older adults.


Our meta-regression analysis, using gender-specific classifications for WC and WHR, detected no statistically significant gender difference in the associations of any AP with mortality. But as only 916, 19, 24, 26, 27, 41, 43, 44, 46, 48 of 18 studies were eligible for meta-regression analyses, these results should be treated with caution. In qualitative analyses, two studies ineligible for meta-regression analyses, but analysing APs as continuous variables, reported gender differences.10, 25 To our knowledge, this is the first review investigating this topic.


Smoking is associated with higher mortality57, 62 and a lower BMI,63 but also with higher WC and WHR.62, 64, 65, 66, 67 Our review, therefore, investigated the impact of smoking on the AP–mortality relationship. However, heterogeneous study results and the small number of stratified studies make interpretation difficult. Our analyses showed that three16, 24, 47 of five16, 24, 41, 42, 47 cohorts found the highest relative mortality risk for smokers in the lowest BMI category. One explanation for this higher mortality risk among persons with a low BMI is reverse causation by the higher prevalence of chronic diseases in smokers.68 Interestingly, all three studies16, 23, 45 that showed an increased mortality risk for low BMI in smokers excluded the first years of follow-up to minimize confounding by underlying illnesses.

Interpretation of APs

Our review demonstrated that BMI shows either negative or insignificant association at the 5% level with all-cause mortality, after adjustment for WC.12, 19, 48 However, when adjusted for BMI, WC and WHR showed stronger positive and independent associations with all-cause mortality than in analyses that did not control for BMI.12, 16, 44, 45 This may be because WC is a good measure of visceral fat mass69, 70 and is closely related to adverse metabolic profiles and resulting diseases,13, 14, 15 whereas BMI does not reflect body fat distribution. Studies in this review showed high correlations between BMI and WC,12, 26 whereas WHR correlated less well with BMI.16 However, WHR, representing visceral fat and gluteal muscle, may provide additional information regarding regional fat distribution. The WHtR appeared to be an attractive alternative to WHR, providing a correction for body frame size by using height, which is more commonly and more conveniently measured than hip circumference. Our review supports the relevance of WHtR as a mortality predictor, although few prospective studies have examined this variable. Ratio measures, such as WHR and WHtR, have been criticized for their potential for complex biological interpretations71 and poor test-retest reliability72 and because successful weight loss interventions commonly reduce both the numerator and denominator, resulting in no observed change in the ratio.73 Recently, a Swedish cohort showed a new AP, waist-to-hip-to-height-ratio, to be a better mortality predictor than BMI, WHR, WC and WHtR.74 This promising measure should be investigated in further analyses.


Our review has several strengths. Systematic reviews that have analysed obesity and its associations with mortality are often limited to patients with specific diseases.75, 76 Because we chose a systematic search strategy and analysed population-based cohort studies, our results could be generalizable to a broader patient population. By analysing combined associations of both general and abdominal AP with all-cause mortality, this systematic review extends current literature and may help to identify body distribution types at risk. Furthermore, it gives arguments for assessing two AP for mortality risk assessment in spite of limited time and economic resources.


The included studies and their results were heterogeneous, which limits our ability to generalize our results. The heterogeneity between studies resulted from differing categorization of APs, heterogeneous choice of reference categories and the variety of statistical analyses, including different variables adjusting relative mortality risk and may lead to different study results in the literature. Some confirmed confounders of the obesity–mortality relationship, such as cigarette smoking25, 27, 40, 48 and alcohol use,25, 27, 40, 45, 46 were not adjusted for in all studies. Adjustment for physical activity and possible mediators of obesity effects such as diabetes, hypertension and high cholesterol; may result in overadjustment.77 Only eight studies performed sensitivity analyses excluding deaths during the first years of follow-up to control for the effects of severe or terminal illnesses that may present with low body weight.10, 16, 19, 24, 26, 27, 46, 47 However, our sensitivity analyses addressing physical activity, drinking, obesity mediators and exclusion of early years of follow-up did not change the results significantly. The high collinearity between APs may have caused mutual adjustment in the combined analyses.

The literature search was performed for publications until 7 May 2010 and studies since then could not be included in the systematic review. In order to address this issue, we compared our findings with a recent meta-analysis of British cohort studies by Czernichow et al. and with a recent systematic review by Donini et al. for the obesity–mortality association in older persons (>65 years) as described above.17, 59 Patterns of association between AP and total mortality shown by Czernichow et al. were similar to our results. The results by Donini et al. confirmed the complexity of AP–mortality associations in older persons and we agree that age-specific categories should be examined in further studies.

Although our study could not provide a pooled analysis of individual patient data, we were able to perform a meta-regression analysis with nine eligible studies. Given the small number of studies in our stratified qualitative and meta-regression analyses, further research is needed to evaluate our results. We could not address changes in weight or body shape as only baseline measurements were taken into account by the studies included.

There is an ongoing discussion whether the kind of measurement of the abdominal AP influences the association with morbidity and mortality and whether this can be accurately measured in clinical practice.78 Indeed, assessment of the AP differed between the studies: either self-reported19, 27, 41, 44 or measured.10, 12, 16, 24, 25, 40, 42, 43, 45, 47, 48, 49 However, validation studies showed a high correlation between self-reported and measured AP.79, 80 A systematic review revealed that the measurement protocol had no substantial influence on the association between WC and mortality.81

We restricted our review to the Caucasian population. Because there may be ethnic differences in the associations between AP and obesity,30, 32, 82 our results should not be generalized to other ethnicities. Conversely, controlling for this variable increases the power of our results by eliminating some potentially confounding effects. Addressing mortality-associated risk of BMI, WC and WHtR in ethnic minority populations would be an interesting future research topic.


None of the examined individual obesity measures was shown to be clearly superior to the others in mortality prediction. However, all included studies investigating WHR and WHtR showed positive associations with all-cause mortality, whereas studies analysing BMI and WC as categorical variables were more heterogeneous and predominantly U- or J-shaped (Figure 2 of our meta-regression). Mortality prediction may be improved by combining BMI and an abdominal obesity measure such as WHR or WC. Participants in the lowest BMI quintile having a moderate-to-large WC or WHR had the highest relative mortality risk. WHR is less correlated with BMI than WC is and study results were less heterogeneous for its associations with all-cause mortality. In our meta-regression analysis, older age (65 years old) compared with younger age (<65 years old) was associated with a lower relative mortality risk for high WHR and WC, but not high BMI.

Further investigations are required to examine the combined relationships of APs representing general and abdominal obesity, with all-cause mortality. These should use standardized AP categories. Meta-analyses of individual data such as The Pooling Project of Prospective Studies of Diet and Cancer83 will further clarify associations and help to inform clinical practice.


  1. 1

    World Health Organisation. Obesity: preventing and managing the global epidemic. Report of a WHO consultation on obesity. WHO: Geneva: Switzerland, 1998.

  2. 2

    Finucane MM, Stevens GA, Cowan MJ, Danaei G, Lin JK, Paciorek CJ et al. National, regional, and global trends in body-mass index since 1980: systematic analysis of health examination surveys and epidemiological studies with 960 country-years and 9.1 million participants. Lancet 2011; 377: 557–567.

    Article  PubMed  PubMed Central  Google Scholar 

  3. 3

    Flegal KM, Graubard BI, Williamson DF, Gail MH . Excess deaths associated with underweight, overweight, and obesity. JAMA 2005; 293: 1861–1867.

    Article  CAS  PubMed  Google Scholar 

  4. 4

    Mokdad AH, Marks JS, Stroup DF, Gerberding JL . Actual causes of death in the United States. JAMA 2004; 291: 1238–1245.

    Article  PubMed  Google Scholar 

  5. 5

    Yusuf S, Hawken S, Ounpuu S, Bautista L, Franzosi MG, Commerford P et al. Obesity and the risk of myocardial infarction in 27,000 participants from 52 countries: a case-control study. Lancet 2005; 366: 1640–1649.

    Article  PubMed  Google Scholar 

  6. 6

    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: 1083–1096.

    Article  PubMed  Google Scholar 

  7. 7

    McGee DL, Diverse Populations Collaboration. Body mass index and mortality: a meta-analysis based on person-level data from twenty-six observational studies. Ann Epidemiol 2005; 15: 87–97.

    Article  PubMed  Google Scholar 

  8. 8

    Romero-Corral A, Montori VM, Somers VK, Korinek J, Thomas RJ, Allison TG et al. Association of body weight with total mortality and with cardiovascular events in coronary artery disease: a systematic review of cohort studies. Lancet 2006; 368: 666–678.

    Article  PubMed  Google Scholar 

  9. 9

    Berrington de Gonzalez A, Hartge P, Cerhan JR, Flint AJ, Hannan L, MacInnis RJ et al. Body-mass index and mortality among 1.46 million white adults. N Engl J Med 2010; 363: 2211–2219.

    Article  CAS  PubMed  Google Scholar 

  10. 10

    Mason C, Craig CL, Katzmarzyk PT . Influence of central and extremity circumferences on all-cause mortality in men and women. Obesity 2008; 16: 2690–2695.

    Article  PubMed  Google Scholar 

  11. 11

    Snijder MB, van Dam RM, Visser M, Seidell JC . What aspects of body fat are particularly hazardous and how do we measure them? Int J Epidemiol 2006; 35: 83–92.

    Article  CAS  PubMed  Google Scholar 

  12. 12

    Schneider HJ, Friedrich N, Klotsche J, Pieper L, Nauck M, John U et al. The predictive value of different measures of obesity for incident cardiovascular events and mortality. J Clin Endocrinol Metab 2010; 95: 1777–1785.

    Article  CAS  PubMed  Google Scholar 

  13. 13

    Berg AH, Scherer PE . Adipose tissue, inflammation, and cardiovascular disease. Circ Res 2005; 96: 939–949.

    Article  CAS  PubMed  Google Scholar 

  14. 14

    Calle EE, Kaaks R . Overweight, obesity and cancer: epidemiological evidence and proposed mechanisms. Nat Rev Cancer 2004; 4: 579–591.

    Article  CAS  PubMed  Google Scholar 

  15. 15

    Despres JP, Lemieux I, Prud'homme D . Treatment of obesity: need to focus on high risk abdominally obese patients. BMJ 2001; 322: 716–720.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. 16

    Pischon T, Boeing H, Hoffmann K, Bergmann M, Schulze MB, Overvad K et al. General and abdominal adiposity and risk of death in Europe. N Engl J Med 2008; 359: 2105–2120.

    Article  CAS  PubMed  Google Scholar 

  17. 17

    Czernichow S, Kengne AP, Stamatakis E, Hamer M, Batty GD . Body mass index, waist circumference and waist-hip ratio: which is the better discriminator of cardiovascular disease mortality risk? Evidence from an individual-participant meta-analysis of 82864 participants from nine cohort studies. Obes Rev 2011; 12: 680–687.

    CAS  PubMed  PubMed Central  Google Scholar 

  18. 18

    Welborn TA, Dhaliwal SS, Bennett SA . Waist-hip ratio is the dominant risk factor predicting cardiovascular death in Australia. Med J Aust 2003; 179: 580–585.

    PubMed  Google Scholar 

  19. 19

    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: 2117–2128.

    Article  CAS  PubMed  Google Scholar 

  20. 20

    Ashwell M, Hsieh SD . Six reasons why the waist-to-height ratio is a rapid and effective global indicator for health risks of obesity and how its use could simplify the international public health message on obesity. Int J Food Sci Nutr 2005; 56: 303–307.

    Article  PubMed  Google Scholar 

  21. 21

    Emerging Risk Factors Collaboration, Wormser D, Kaptoge S, Di Angelantonio E, Wood A, Pennells L et al. Separate and combined associations of body-mass index and abdominal adiposity with cardiovascular disease: collaborative analysis of 58 prospective studies. Lancet 2011; 377: 1085–1095.

    Article  Google Scholar 

  22. 22

    Katzmarzyk PT, Janssen I, Ardern CI . Physical inactivity, excess adiposity and premature mortality. Obes Rev 2003; 4: 257–290.

    Article  CAS  PubMed  Google Scholar 

  23. 23

    Seidell JC . Waist circumference and waist/hip ratio in relation to all-cause mortality, cancer and sleep apnea. Eur J Clin Nutr 2010; 64: 35–41.

    Article  CAS  PubMed  Google Scholar 

  24. 24

    Simpson JA, MacInnis RJ, Peeters A, Hopper JL, Giles GG, English DR . A comparison of adiposity measures as predictors of all-cause mortality: the Melbourne Collaborative Cohort Study. Obesity (Silver Spring) 2007; 15: 994–1003.

    Article  Google Scholar 

  25. 25

    Welborn TA, Dhaliwal SS . Preferred clinical measures of central obesity for predicting mortality. Eur J Clin Nutr 2007; 61: 1373–1379.

    Article  CAS  PubMed  Google Scholar 

  26. 26

    Reis JP, Macera Caroline A, Araneta Maria R, Lindsay Suzanne P, Marshall Simon J, Wingard Deborah L . Comparison of overall obesity and body fat distribution in predicting risk of mortality. Obesity 2009; 17: 1232–1239.

    PubMed  Google Scholar 

  27. 27

    Baik I, Ascherio A, Rimm EB, Giovannucci E, Spiegelman D, Stampfer MJ et al. Adiposity and mortality in men. Am J Epidemiol 2000; 152: 264–271.

    Article  CAS  PubMed  Google Scholar 

  28. 28

    Chang SH, Beason TS, Hunleth JM, Colditz GA . A systematic review of body fat distribution and mortality in older people. Maturitas 2012; 72: 175–191.

    Article  PubMed  PubMed Central  Google Scholar 

  29. 29

    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: 2008–2012.

    Article  CAS  PubMed  Google Scholar 

  30. 30

    Wen CP, David Cheng TY, Tsai SP, Chan HT, Hsu HL, Hsu CC et al. Are Asians at greater mortality risks for being overweight than Caucasians? Redefining obesity for Asians. Public Health Nutr 2009; 12: 497–506.

    Article  PubMed  Google Scholar 

  31. 31

    Lear SA, Chen MM, Frohlich JJ, Birmingham CL . The relationship between waist circumference and metabolic risk factors: cohorts of European and Chinese descent. Metabolism 2002; 51: 1427–1432.

    Article  CAS  PubMed  Google Scholar 

  32. 32

    Misra A, Khurana L . Obesity-related non-communicable diseases: South Asians vs White Caucasians. Int J Obes (Lond) 2011; 35: 167–187.

    Article  CAS  Google Scholar 

  33. 33

    Orwin RG . Evaluating coding decisions. In: Cooper H, Hedges LV, (editors). The Handbook of Research Synthesis. Russell Sage Foundation: New York (NY), 1994.

    Google Scholar 

  34. 34

    Downs SH, Black N . The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. J Epidemiol Community Health 1998; 52: 377–384.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. 35

    von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007; 370: 1453–1457.

    Article  PubMed  Google Scholar 

  36. 36

    May S, Littlewood C, Bishop A . Reliability of procedures used in the physical examination of non-specific low back pain: a systematic review. Aust J Physiother 2006; 52: 91–102.

    Article  PubMed  Google Scholar 

  37. 37

    Orsini N, Bellocco R, Greenland S . Generalized least squares for trend estimation of summarized dose-response data. Stata J 2007; 6: 40–57.

    Article  Google Scholar 

  38. 38

    Egger M, Davey Smith G, Schneider M, Minder C . Bias in meta-analysis detected by a simple, graphical test. BMJ 1997; 315: 629–634.

    CAS  Article  Google Scholar 

  39. 39

    Haftenberger M, Lahmann PH, Panico S, Gonzalez CA, Seidell JC, Boeing H et al. Overweight, obesity and fat distribution in 50- to 64-year-old participants in the European Prospective Investigation into Cancer and Nutrition (EPIC). Public Health Nutr 2002; 5: 1147–1162.

    Article  CAS  PubMed  Google Scholar 

  40. 40

    Higgins M, Kannel W, Garrison R, Pinsky J, Stokes J . Hazards of obesity–the Framingham experience. Acta Med Scand Suppl 1988; 723: 23–36.

    CAS  PubMed  Google Scholar 

  41. 41

    Hu FB, Willett WC, Li T, Stampfer MJ, Colditz GA, Manson JE . Adiposity as compared with physical activity in predicting mortality among women. N Engl J Med 2004; 351: 2694–2703.

    Article  CAS  PubMed  Google Scholar 

  42. 42

    Visscher TL, Seidell JC, Molarius A, van der Kuip D, Hofman A, Witteman JC . A comparison of body mass index, waist - hip ratio and waist. Int J Obes Relat Metab Disord 2001; 25: 1730–1735.

    Article  CAS  PubMed  Google Scholar 

  43. 43

    Walter S, Kunst A, Mackenbach J, Hofman A, Tiemeier H . Mortality and disability: the effect of overweight and obesity. Int J Obes (Lond) 2009; 33: 1410–1418.

    Article  CAS  Google Scholar 

  44. 44

    Zhang C, Rexrode Kathryn M, van Dam Rob M, Li Tricia Y, Hu Frank B . Abdominal obesity and the risk of all-cause, cardiovascular, and cancer. Circulation 2008; 117: 1658–1667.

    Article  PubMed  Google Scholar 

  45. 45

    Janssen I, Katzmarzyk PT, Ross R . Body mass index is inversely related to mortality in older people after adjustment for waist circumference. J Am Geriatr Soc 2005; 53: 2112–2118.

    Article  PubMed  Google Scholar 

  46. 46

    Dolan CM, Kraemer H, Browner W, Ensrud K, Kelsey JL . Associations between body composition, anthropometry, and mortality in women aged 65 years and older. Am J Public Health 2007; 97: 913–918.

    Article  PubMed  PubMed Central  Google Scholar 

  47. 47

    Price GM, Uauy R, Breeze E, Bulpitt CJ, Fletcher AE . Weight, shape, and mortality risk in older persons: elevated greater risk of death. Am J Clin Nutr 2006; 84: 449–460.

    Article  CAS  PubMed  Google Scholar 

  48. 48

    Tice JA, Kanaya A, Hue T, Rubin S, Buist DS, Lacroix A et al. Risk factors for mortality in middle-aged women. Arch Intern Med 2006; 166: 2469–2477.

    Article  PubMed  Google Scholar 

  49. 49

    Taylor AE, Ebrahim S, Ben-Shlomo Y, Martin RM, Whincup PH, Yarnell JW et al. Comparison of the associations of body mass index and measures of central adiposity and fat mass with coronary heart disease, diabetes, and all-cause mortality: a study using data from 4 UK cohorts. Am J Clin Nutr 2010; 91: 547–556.

    Article  CAS  PubMed  Google Scholar 

  50. 50

    Vazquez G, Duval S, Jacobs DR, Silventoinen K . Comparison of body mass index, waist circumference, and waist/hip ratio in predicting incident diabetes: a meta-analysis. Epidemiol Rev 2007; 29: 115–128.

    Article  PubMed  Google Scholar 

  51. 51

    Huxley R, Mendis S, Zheleznyakov E, Reddy S, Chan J . Body mass index, waist circumference and waist/hip ratio as predictors of cardiovascular risk--a review of the literature. Eur J Clin Nutr 2010; 64: 16–22.

    Article  CAS  PubMed  Google Scholar 

  52. 52

    Lee CM, Huxley RR, Wildman RP, Woodward M . Indices of abdominal obesity are better discriminators of cardiovascular risk factors than BMI: a meta-analysis. J Clin Epidemiol 2008; 61: 646–653.

    Article  PubMed  Google Scholar 

  53. 53

    Cook NR . Use and misuse of the receiver operating characteristic curve in risk prediction. Circulation 2007; 115: 928–935.

    Article  PubMed  Google Scholar 

  54. 54

    Pepe MS, Janes H, Longton G, Leisenring W, Newcomb P . Limitations of the odds ratio in gauging the performance of a diagnostic, prognostic, or screening marker. Am J Epidemiol 2004; 159: 882–890.

    Article  PubMed  Google Scholar 

  55. 55

    Staiano AE, Reeder BA, Elliott S, Joffres MR, Pahwa P, Kirkland SA et al. Body mass index versus waist circumference as predictors of mortality in Canadian adults. Int J Obes (Lond) 2012; 36: 1450–1454.

    Article  CAS  Google Scholar 

  56. 56

    Heiat A, Vaccarino V, Krumholz HM . An evidence-based assessment of federal guidelines for overweight and obesity as they apply to elderly persons. Arch Intern Med 2001; 161: 1194–1203.

    Article  CAS  PubMed  Google Scholar 

  57. 57

    Lindsted KD, Singh PN . Body mass and 26-year risk of mortality among women who never smoked: findings from the Adventist Mortality Study. Am J Epidemiol 1997; 146: 1–11.

    Article  CAS  PubMed  Google Scholar 

  58. 58

    Stevens J, Cai J, Pamuk ER, Williamson DF, Thun MJ, Wood JL . The effect of age on the association between body-mass index and mortality. N Engl J Med 1998; 338: 1–7.

    Article  CAS  PubMed  Google Scholar 

  59. 59

    Donini LM, Savina C, Gennaro E, De Felice MR, Rosano A, Pandolfo MM et al. A systematic review of the literature concerning the relationship between obesity and mortality in the elderly. J Nutr Health Aging 2012; 16: 89–98.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  60. 60

    Hedlund J . Community-acquired pneumonia requiring hospitalisation. Factors of importance for the short-and long term prognosis. Scand J Infect Dis 1995; 97: 1–60.

    CAS  Google Scholar 

  61. 61

    Potter J, Klipstein K, Reilly JJ, Roberts M . The nutritional status and clinical course of acute admissions to a geriatric unit. Age Ageing 1995; 24: 131–136.

    Article  CAS  PubMed  Google Scholar 

  62. 62

    Chiolero A, Faeh D, Paccaud F, Cornuz J . Consequences of smoking for body weight, body fat distribution, and insulin resistance. Am J Clin Nutr 2008; 87: 801–809.

    Article  CAS  PubMed  Google Scholar 

  63. 63

    Berlin I . Smoking-induced metabolic disorders: a review. Diabetes Metab 2008; 34: 307–314.

    Article  PubMed  Google Scholar 

  64. 64

    Barrett-Connor E, Khaw KT . Cigarette smoking and increased central adiposity. Ann Intern Med 1989; 111: 783–787.

    Article  CAS  PubMed  Google Scholar 

  65. 65

    Canoy D, Wareham N, Luben R, Welch A, Bingham S, Day N et al. Cigarette smoking and fat distribution in 21,828 British men and women: a population-based study. Obes Res 2005; 13: 1466–1475.

    Article  PubMed  Google Scholar 

  66. 66

    Komiya H, Mori Y, Yokose T, Tajima N . Smoking as a risk factor for visceral fat accumulation in Japanese men. Tohoku J Exp Med 2006; 208: 123–132.

    Article  PubMed  Google Scholar 

  67. 67

    Shimokata H, Muller DC, Andres R . Studies in the distribution of body fat. III. Effects of cigarette smoking. JAMA 1989; 261: 1169–1173.

    Article  CAS  PubMed  Google Scholar 

  68. 68

    Adams KF, Schatzkin A, Harris TB, Kipnis V, Mouw T, Ballard-Barbash R et al. Overweight, obesity, and mortality in a large prospective cohort of persons 50 to 71 years old. N Engl J Med 2006; 355: 763–778.

    Article  CAS  PubMed  Google Scholar 

  69. 69

    Despres JP, Pascot A, Lemieux I . Facteurs de risque associes a l'obesite: le point de vue du metabolicien. Ann Endocrinol 2000; 61: 31–38.

    CAS  Google Scholar 

  70. 70

    Lean ME, Han TS, Morrison CE . Waist circumference as a measure for indicating need for weight management. BMJ 1995; 311: 158–161.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  71. 71

    Molarius A, Seidell JC . Selection of anthropometric indicators for classification of abdominal fatness--a critical review. Int J Obes Relat Metab Disord 1998; 22: 719–727.

    Article  CAS  PubMed  Google Scholar 

  72. 72

    Chen MM, Lear SA, Gao M, Frohlich JJ, Birmingham CL . Intraobserver and interobserver reliability of waist circumference and the waist-to-hip ratio. Obes Res 2001; 9: 651.

    Article  CAS  PubMed  Google Scholar 

  73. 73

    Kay SJ, Fiatarone Singh MA . The influence of physical activity on abdominal fat: a systematic review of the literature. Obes Rev 2006; 7: 183–200.

    Article  CAS  PubMed  Google Scholar 

  74. 74

    Rosenblad A, Leppert J, Nilsson G . A novel anthropometric measure for predicting all-cause mortality. Eur J Cardiovasc Prev Rehabil 2010; 17: S1–S2.

    Google Scholar 

  75. 75

    Coutinho T, Goel K, Correa de Sa D, Kragelund C, Kanaya AM, Zeller M et al. Central obesity and survival in subjects with coronary artery disease: a systematic review of the literature and collaborative analysis with individual subject data. J Am Coll Cardiol 2011; 57: 1877–1886.

    Article  PubMed  Google Scholar 

  76. 76

    Protani M, Coory M, Martin JH . Effect of obesity on survival of women with breast cancer: systematic review and meta-analysis. Breast Cancer Res Treat 2010; 123: 627–635.

    Article  PubMed  Google Scholar 

  77. 77

    Manson JE, Stampfer MJ, Hennekens CH, Willett WC . Body weight and longevity. A reassessment. JAMA 1987; 257: 353–358.

    Article  CAS  PubMed  Google Scholar 

  78. 78

    Wang J, Thornton JC, Bari S, Williamson B, Gallagher D, Heymsfield SB et al. Comparisons of waist circumferences measured at 4 sites. Am J Clin Nutr 2003; 77: 379–384.

    Article  PubMed  Google Scholar 

  79. 79

    Rimm EB, Stampfer MJ, Giovannucci E, Ascherio A, Spiegelman D, Colditz GA et al. Body size and fat distribution as predictors of coronary heart disease among middle-aged and older US men. Am J Epidemiol 1995; 141: 1117–1127.

    Article  CAS  PubMed  Google Scholar 

  80. 80

    Willett W, Stampfer MJ, Bain C, Lipnick R, Speizer FE, Rosner B et al. Cigarette smoking, relative weight, and menopause. Am J Epidemiol 1983; 117: 651–658.

    Article  CAS  PubMed  Google Scholar 

  81. 81

    Ross R, Berentzen T, Bradshaw AJ, Janssen I, Kahn HS, Katzmarzyk PT et al. Does the relationship between waist circumference, morbidity and mortality depend on measurement protocol for waist circumference? Obes Rev 2008; 9: 312–325.

    Article  CAS  PubMed  Google Scholar 

  82. 82

    WHO Expert Consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet 2004; 363: 157–163.

    Article  Google Scholar 

  83. 83

    Smith-Warner SA, Spiegelman D, Ritz J, Albanes D, Beeson WL, Bernstein L et al. Methods for pooling results of epidemiologic studies: the Pooling Project of Prospective Studies of Diet and Cancer. Am J Epidemiol 2006; 163: 1053–1064.

    Article  PubMed  Google Scholar 

Download references


We are very thankful to Ute Troitzsch, Thuringian University and State Library, Jena, Germany, for her professional support in our literature search. We thank Dipl.-Biol. Nico Schneider, Institute of General Practice and Family Medicine, Jena, for his competent editing support. We thank Norbert Krause, MA, Scientific Coordinator Graduate Academy, Jena, for support in data extraction and synthesis for meta-regression analyses. We are very thankful to Douglas von Korff for the careful revision of our manuscript.

Author information



Corresponding author

Correspondence to M H Freitag.

Ethics declarations

Competing interests

Dr Schlattmann reported having received royalties from Springer and payment for development of educational presentations from Bayer. The remaining authors declare no conflict of interest.

Additional information

Supplementary Information accompanies this paper on European Journal of Clinical Nutrition website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Carmienke, S., Freitag, M., Pischon, T. et al. General and abdominal obesity parameters and their combination in relation to mortality: a systematic review and meta-regression analysis. Eur J Clin Nutr 67, 573–585 (2013).

Download citation


  • body mass index
  • waist circumference
  • waist-to-hip ratio
  • mortality

Further reading


Quick links