Abstract

Background: It is unclear which anthropometric measure is most useful for assessment of the cardiovascular risk. We investigated the association between different anthropometric measures and risk of heart failure (HF) hospitalization. Methods: BMI, waist–hip ratio (WHR), waist circumference (WC), body fat percentage (BF%), weight and height were measured among 26 653 subjects (aged 45–73 years) without history of myocardial infarction (MI), stroke or HF from the Malmö Diet and Cancer cohort at baseline in 1991–96. Incidence of HF hospitalizations was monitored during a mean follow-up of 15 years. Results: Seven hundred and twenty-seven subjects were hospitalized with HF as primary diagnosis, of whom 157 had an MI before or concurrent with the HF. After adjustment for potential confounding factors, the hazard ratios of HF hospitalization (fourth vs. first sex-specific quartile) were 1.80 (95% CI: 1.45–2.24) for BMI, 1.87 (1.50–2.34) for WC, 1.77 (1.43–2.19) for WHR, 1.35 (1.09–1.68) for BF%, 1.93 (1.57–2.39) for weight and 1.18 (0.96–1.44) for height. Significant interactions between BMI and WC and WHR, respectively, were observed, and the joint exposure of high BMI and high WC or high WHR further increased the risk. The results were similar in secondary analyses, i.e. excluding incident HF with previous MI during the follow-up. Conclusion: Our results support the view that raised BMI, WC, WHR or BF% increases the risk of HF hospitalization. The joint exposure of high BMI and high WHR or high WC further increased the risk in an additive way.

Introduction

Obesity is a major risk factor for several cardiovascular diseases (CVD),1,5 including an increased risk of heart failure (HF).6,8 The underlying causal links between obesity and cardiac dysfunction are complex. Obesity is associated with a higher risk of hypertension,9 insulin resistance and diabetes mellitus,4 inflammation,10,12 socio-economic status and lifestyle,13,15 all of which could increase the cardiovascular risk.

It is still controversial which anthropometric measure is most useful for assessment of the cardiovascular risk. BMI, being the marker for general fat, is the most practical and commonly used. However, the INTERHEART study, a multi-national case–control study of myocardial infarction (MI), reported substantially stronger relationships for the waist–hip ratio (WHR) than for BMI.3 Because visceral fat is more metabolically active than other fat tissues, it has been proposed that WHR or waist circumference (WC) is preferable.2,16 However, few have studied the relationships with incidence of HF, and the results are not consistent. Some studies have shown that BMI, WC and WHR had similar prediction for incident HF.17 Others found that abdominal body fat distribution may be a stronger risk factor for HF than overall obesity.18,19

The aim of the study was to explore the relationship between risk of HF hospitalization and different anthropometric measures, i.e. BMI, WC, WHR, body fat percentage (BF%), weight and height, in a population-based cohort study. We also explored whether there is any combined effect of the different anthropometric measures on the risk of HF.

Methods

Study population

The ‘Malmö Diet and Cancer (MDC)’ cohort, from the city of Malmö in southern Sweden, was used for the present study. Detailed information for MDC has been described previously.20,22 In brief, all men and women, born between 1923 and 1950 in Malmö city were invited to the MDC study. During the period March 1991 to September 1996, 28 449 subjects (11 246 men and 17 203 women) underwent sampling of peripheral venous blood, measurement of blood pressure and anthropometric measures and filled out a self-administered questionnaire. Participation (rate of 41%) in the MDC study has been described in detail elsewhere.21 In short, mortality has been shown significantly higher in non-participants both during and following the recruitment period. The participants in the MDC study were also compared with participants in a mailed health survey in Malmö 1994, with regard to subjective health, lifestyle and socio-demographic characteristics. The proportion reporting good health was found higher in the MDC study than in the mailed health survey (where 75% participated); however, socio-economic structure and prevalence of smoking and overweight/obesity were similar.21

In the present study, subjects with history of cardiovascular events (MI or stroke, n = 970 subjects) or HF (n = 46 subjects) before the baseline examination were excluded. In addition, subjects (n = 780) were also excluded due to missing values of anthropometric measurements and other biological, life-style and socio-economic variables. Thus, the final study population in the analysis consisted of 26 653 (10 223, 38.4% men, and 16 430, 61.6% women) subjects, aged 45–73 years.

The ethics committee at Lund University Lund, Sweden, approved the study (LU 51/90), and all participants provided informed consent.

Baseline examinations

The examinations were performed by two trained nurses at the screening centre. Standing height was measured with a fixed stadiometer calibrated in centimetres. Weight was measured to the nearest 0.1 kg using balance-beam scale, with subjects wearing light clothing and no shoes. BMI was calculated as weight (kg) divided by the square of the height (m2). Waist was measured as the circumference (cm) between the lowest rib margin and iliac crest and hip circumference (cm) as the largest circumference between waist and thighs. WHR was defined as the ratio of circumference of waist to hip. Bioelectrical Impedance Analyzers (BIA) was used for estimating body composition, and BF% was calculated using an algorithm, according to procedures provided by the manufacturer (BIA 103, RJL systems, single-frequency analyser, Detroit, USA).23 Weight, height, BMI, WC, WHR and BF% were categorized into sex-specific quartiles Q1–4.

Information on current use of lipid, blood pressure-lowering and anti-diabetic medications, smoking habits, alcohol consumption, leisure-time physical activity, education level, civil status and immigrant status was obtained from a self-administered questionnaire. Blood pressure was measured using a mercury-column sphygmomanometer after 10 min of rest in the supine position. Leucocyte concentrations were analysed consecutively in fresh heparinized blood. Diabetes mellitus was defined as self-reported physician’s diagnosis of diabetes, or use of anti-diabetic medications. Low level of leisure-time physical activity was defined as the lowest tertile of a score revealed through 18 questions covering a range of activities in the four seasons. The evaluation of the questionnaire has been previously reported.24 Subjects were categorized into current smokers (i.e. those who smoked regularly or occasionally) or non-smokers (i.e. former smokers and never smokers). High alcohol consumption was defined as >40 g alcohol per day for men and >30 g per day for women. Educational level was classified into three categories. ‘Primary education’ included those who had <9 years of education, ‘Some/completed secondary education’ included those who had 9–12 years of education and ‘Education at college or university level’ included those who had >12 years of education. Civil status was categorized into married or not. Immigrant status was grouped as Swedish-born and foreign-born.

Ascertainment of cardiovascular events and HF

All subjects were followed from the baseline examination until a first hospitalization due to HF as primary diagnosis, emigration, death or 30 June 2009, whichever came first. The primary analysis included all HF hospitalization. Subjects were censored at first non-fatal MI in an additional analysis. HF was defined as code 427.00, 427.10, 428.99 (ICD-8); 428 (ICD-9); and I50, I11 (ICD-10) as the primary diagnosis. Non-fatal MI was defined as 410 (ICD-8 and -9) or I21 (ICD-10). The Swedish Hospital Discharge Register (SHDR) was used for case retrieval. Validation studies in the SHDR have shown a validity of 95% for a primary diagnosis of HF, and 94% for MI.25,26

Statistical analysis

Cox proportional hazards regression was used to examine the association between anthropometric measures (in sex-specific quartiles) and incidence of HF hospitalization. Time axis was follow-up time until death, emigration, incident HF or end of follow-up. Hazard ratios (HR), with 95% confidence interval (CI), were calculated. Age and sex were included as covariates in the basic model. Secondly, we also adjusted for systolic blood pressure, leucocyte counts, use of blood pressure or lipid-lowering medication, diabetes mellitus, current smoking, high alcohol consumption, low leisure physical activity, low education, marital status and immigrant status. In an additional analysis we also censored subjects with incident MI before or concurrent with HF during the follow-up period. The Harrell’s C statistics27 were calculated to assess the HF prediction efficiency. The log likelihood ratio was calculated to assess whether the model was improved by adding anthropometric measures to the explanatory variables. Possible interaction between anthropometric measures and age, sex and cardiovascular risk factors on incident HF was explored by introducing interaction terms in the multivariate model. All analyses were performed using PASW version 18 (SPSS Inc., Chicago, IL, USA).

Results

Study cohort

The study population characteristics are presented in table 1. Overall, mean age at baseline was 58 ± 7.6 years, and 61.6% were women. Men compared with women were older, taller and heavier, had higher BMI, WC and WHR, and had a lower BF%. Men were more often hypertensive, high alcohol consumers and had more diabetes. Men were also more often married and had lower education as compared with women.

Table 1

Characteristics of the ‘Malmö Diet and Cancer (MDC)’ cohort participants

MDC (n = 26 653) Men (n = 10 223) Women (n = 16 430) 
Hospitalization due to HF, n (per 1000 p-y) 398 (2.78) 329 (1.39) 
Age at screening (years) 59.0 ± 7.0 57.4 ± 7.9 
BMI (kg/m226.2 ± 3.4 25.4 ± 4.2 
WC (cm) 93.6 ± 12.6 77.7 ± 10.5 
WHR 0.94 ± 0.1 0.79 ± 0.1 
BF% 20.7 ± 4.9 30.7 ± 5.0 
Weight (kg) 81.7 ± 12.0 67.9 ± 11.6 
Height (cm) 176.5 ± 6.6 163.7 ± 6.0 
Leucocyte count (109/l) 6.4 ± 2.6 6.4 ± 2.3 
Systolic blood pressure (mmHg) 143.9 ± 19.2 139.2 ± 20.2 
Diastolic blood pressure (mmHg) 88.0 ± 9.8 84.0 ± 9.7 
Use of blood pressure-lowing medication (%) 17.9 15.5 
Use of lipid-lowing medication (%) 3.3 1.8 
Diabetes (%) 3.6 2.3 
Current smoking (%) 28.8 27.9 
High alcohol consumption (%) 7.5 2.4 
Low physical activity (%) 25.3 24.8 
Married (%) 72.6 60.8 
Primary education (%) 45.2 39.0 
Some/completed secondary school (%) 31.8 37.5 
Education at college or university level (%) 23.0 23.5 
Foreign-born (%) 12.0 11.8 
MDC (n = 26 653) Men (n = 10 223) Women (n = 16 430) 
Hospitalization due to HF, n (per 1000 p-y) 398 (2.78) 329 (1.39) 
Age at screening (years) 59.0 ± 7.0 57.4 ± 7.9 
BMI (kg/m226.2 ± 3.4 25.4 ± 4.2 
WC (cm) 93.6 ± 12.6 77.7 ± 10.5 
WHR 0.94 ± 0.1 0.79 ± 0.1 
BF% 20.7 ± 4.9 30.7 ± 5.0 
Weight (kg) 81.7 ± 12.0 67.9 ± 11.6 
Height (cm) 176.5 ± 6.6 163.7 ± 6.0 
Leucocyte count (109/l) 6.4 ± 2.6 6.4 ± 2.3 
Systolic blood pressure (mmHg) 143.9 ± 19.2 139.2 ± 20.2 
Diastolic blood pressure (mmHg) 88.0 ± 9.8 84.0 ± 9.7 
Use of blood pressure-lowing medication (%) 17.9 15.5 
Use of lipid-lowing medication (%) 3.3 1.8 
Diabetes (%) 3.6 2.3 
Current smoking (%) 28.8 27.9 
High alcohol consumption (%) 7.5 2.4 
Low physical activity (%) 25.3 24.8 
Married (%) 72.6 60.8 
Primary education (%) 45.2 39.0 
Some/completed secondary school (%) 31.8 37.5 
Education at college or university level (%) 23.0 23.5 
Foreign-born (%) 12.0 11.8 

p-y, person-years.

Values are means ± standard deviation, unless stated otherwise.

Risk of HF hospitalizations in relation to anthropometric measures

During a mean follow-up of 14 years, a total of 727 individuals (398 men and 329 women) were hospitalized with HF as primary diagnosis. Of them, 157 (91 men and 66 women) had an incident MI before or concurrent with HF hospitalization during follow-up. The latter group was censored at the time of the infarction in a secondary analysis.

The overall analysis showed that overweight and obesity increased the risk of HF hospitalization independently from several socio-demographic, lifestyle and biological factors. BMI, WC, WHR and BF% were significantly related to an increased risk of HF in both sexes (table 2). Taking potential confounding factors into account, the HRs of HF hospitalization (fourth vs. first sex-specific quartile) were 1.87 (95% CI: 1.50–2.34) for WC, 1.80 (1.45–2.24) for BMI, 1.77 (1.43–2.19) for WHR and 1.35 (1.09–1.68) for BF% (table 2, figure 1). Interaction terms between sex and age, respectively, and different anthropometric indictors were added in the final Cox’s proportional hazards model with adjustment for possible confounders. There was a statistically significant interaction between BMI with sex (P = 0.002); however, no interaction was observed between sex and other anthropometric measures (e.g. WC, WHR, BF%, weight or height). For BMI the HR for all HF hospitalization (fourth vs. first quartile) was 1.82 (1.37–2.42) among men and 1.80 (1.28–2.53) among women. There were no significant interactions between age and any of the anthropometric measures.

Figure 1

Adjusted HR (95% CI) for different anthropometric measures (in sex-specific quartiles, Q1–Q4) in relation to the risk of hospitalization due to heart failure

Figure 1

Adjusted HR (95% CI) for different anthropometric measures (in sex-specific quartiles, Q1–Q4) in relation to the risk of hospitalization due to heart failure

Table 2

Hospitalizations due to heart failure (HF) in relation to different anthropometric measures

Sex-specific quartiles Q1 Q2 Q3 Q4 P for trend 
BMI, n 6666 6661 6662 6664  
    Median, kg/m2 (men/women) 22.5/21.1 24.9/23.5 26.9/26.0 30.0/30.1  
HF, n (per 1000 p-y) 121 (1.27) 132 (1.38) 164 (1.72) 310 (3.34)  
    Adjusted HR (95% CI) 1.00 0.98 (0.76–1.25) 1.12 (0.88–1.42) 1.80 (1.45–2.24) <0.001 
WC, n 6766 6103 7229 6555  
    Median, cm (men/women) 82/67 90/73 96/79 105/90  
HF, n (per 1000 p-y) 112 (1.15) 112 (1.27) 174 (1.68) 329 (3.64)  
    Adjusted HR (95% CI) 1.00 0.92 (0.71–1.19) 1.15 (0.90–1.46) 1.87 (1.50–2.34) <0.001 
WHR, n 6661 6698 6650 6644  
    Median (men/women) 0.88/0.74 0.92/0.77 0.96/0.80 1.01/0.85  
HF, n (per 1000 p-y) 149 (1.34) 137 (1.42) 169 (1.78) 292 (3.18)  
    Adjusted HR (95% CI) 1.00 1.04 (0.82–1.32) 1.13 (0.90–1.42) 1.77 (1.43–2.19) <0.001 
BF%, n 6851 6228 7266 6308  
    Median, % (men/women) 15/25 19/29 22/32 26/37  
HF, n (per 1000 p-y) 133 (1.37) 135 (1.51) 205 (1.97) 254 (2.87)  
    Adjusted HR (95% CI) 1.00 0.98 (0.77–1.24) 1.18 (0.95–1.47) 1.35 (1.09–1.68) 0.001 
Weight, n 6946 6407 6702 6598  
    Median, kg (men/women) 69/56 77/64 84/70 95/81  
HF, n (per 1000 p-y) 134 (1.36) 145 (1.58) 166 (1.73) 282 (3.05)  
    Adjusted HR (95% CI) 1.00 1.13 (0.89–1.43) 1.26 (1.00–1.59) 1.93 (1.57–2.39) <0.001 
Height, n 6821 6375 6328 7129  
    Median, cm (men/women) 169/157 175/162 178/165 184/170  
HF, n (per 1000 p-y) 236 (2.48) 179 (1.98) 140 (1.54) 172 (1.67)  
    Adjusted HR (95% CI) 1.00 0.98 (0.81–1.19) 0.86 (0.70–1.07) 1.18 (0.96–1.44) 0.357 
Sex-specific quartiles Q1 Q2 Q3 Q4 P for trend 
BMI, n 6666 6661 6662 6664  
    Median, kg/m2 (men/women) 22.5/21.1 24.9/23.5 26.9/26.0 30.0/30.1  
HF, n (per 1000 p-y) 121 (1.27) 132 (1.38) 164 (1.72) 310 (3.34)  
    Adjusted HR (95% CI) 1.00 0.98 (0.76–1.25) 1.12 (0.88–1.42) 1.80 (1.45–2.24) <0.001 
WC, n 6766 6103 7229 6555  
    Median, cm (men/women) 82/67 90/73 96/79 105/90  
HF, n (per 1000 p-y) 112 (1.15) 112 (1.27) 174 (1.68) 329 (3.64)  
    Adjusted HR (95% CI) 1.00 0.92 (0.71–1.19) 1.15 (0.90–1.46) 1.87 (1.50–2.34) <0.001 
WHR, n 6661 6698 6650 6644  
    Median (men/women) 0.88/0.74 0.92/0.77 0.96/0.80 1.01/0.85  
HF, n (per 1000 p-y) 149 (1.34) 137 (1.42) 169 (1.78) 292 (3.18)  
    Adjusted HR (95% CI) 1.00 1.04 (0.82–1.32) 1.13 (0.90–1.42) 1.77 (1.43–2.19) <0.001 
BF%, n 6851 6228 7266 6308  
    Median, % (men/women) 15/25 19/29 22/32 26/37  
HF, n (per 1000 p-y) 133 (1.37) 135 (1.51) 205 (1.97) 254 (2.87)  
    Adjusted HR (95% CI) 1.00 0.98 (0.77–1.24) 1.18 (0.95–1.47) 1.35 (1.09–1.68) 0.001 
Weight, n 6946 6407 6702 6598  
    Median, kg (men/women) 69/56 77/64 84/70 95/81  
HF, n (per 1000 p-y) 134 (1.36) 145 (1.58) 166 (1.73) 282 (3.05)  
    Adjusted HR (95% CI) 1.00 1.13 (0.89–1.43) 1.26 (1.00–1.59) 1.93 (1.57–2.39) <0.001 
Height, n 6821 6375 6328 7129  
    Median, cm (men/women) 169/157 175/162 178/165 184/170  
HF, n (per 1000 p-y) 236 (2.48) 179 (1.98) 140 (1.54) 172 (1.67)  
    Adjusted HR (95% CI) 1.00 0.98 (0.81–1.19) 0.86 (0.70–1.07) 1.18 (0.96–1.44) 0.357 

Q, quartile.

HR adjusted for age, sex, civil status, education level, immigrant status, smoking habits, alcohol consumption, physical activities, blood pressure-lowering medication, lipid-lowering medication, systolic blood pressure, leucocyte count and diabetes mellitus.

If cases with MI before or concurrent with HF hospitalization (n = 157) were excluded in the analysis, the risk for HF hospitalization was only marginally changed (data not shown). Nor did the results change in an analysis including 28 cases without previous history of HF hospitalization, who had HF as underlying cause of death (data not shown).

C-statistics and P-value for model improvement was calculated (Supplementary table S1). In model 1 all anthropometric measures significantly added information above risk factors. WC and WHR, but not BF%, significantly added to the model on top of BMI. C-statistic results were marginally increased compared with a model including risk factors and BMI (Supplementary table S1).

Risk of HF hospitalization in relation to combined pattern of different anthropometric measures

Significant interaction was observed between BMI and WHR (P = 0.004), waist (P = 0.005), weight (P = 0.010) and height (P = 0.035), respectively, on hospitalization due to HF. The joint exposure of high BMI (the top quartile) and high WHR (the top quartile) further increased the risk in an additive way. Overweight subjects with high BMI combined with high WHR had a 2-fold higher risk compared with individuals with low or normal BMI (quartile 1–3) and low or normal WHR (quartile 1–3) (table 3). A similar additive effect was observed for BMI and WC (table 3).

Table 3

Risk of hospitalization due to heart failure (HF) in relation to combined pattern of different anthropometric measures

MDC (n = 26 653) HF, n (per 1000 p-y) Adjusted HR (95% CI) 
BMI Q1–3, WHR Q1–3 (n = 16 851) 316 (1.30) 1.00 
BMI Q1–3, WHR Q4 (n = 3138) 103 (2.34) 1.58 (1.26–1.98) 
BMI Q4, WHR Q1–3 (n = 3158) 119 (2.65) 1.66 (1.34–2.06) 
BMI Q4, WHR Q4 (n = 3506) 191 (3.99) 2.13 (1.77–2.58) 
BMI Q1–3, WC Q1–3 (n = 18 460) 351 (1.32) 1.00 
BMI Q1–3, WC Q4 (n = 1529) 66 (3.11) 1.62 (1.24–2.11) 
BMI Q4, WC Q1–3 (n = 1638) 47 (1.99) 1.44 (1.06–1.95) 
BMI Q4, WC Q4 (n = 5026) 263 (3.80) 1.97 (1.67–2.34) 
MDC (n = 26 653) HF, n (per 1000 p-y) Adjusted HR (95% CI) 
BMI Q1–3, WHR Q1–3 (n = 16 851) 316 (1.30) 1.00 
BMI Q1–3, WHR Q4 (n = 3138) 103 (2.34) 1.58 (1.26–1.98) 
BMI Q4, WHR Q1–3 (n = 3158) 119 (2.65) 1.66 (1.34–2.06) 
BMI Q4, WHR Q4 (n = 3506) 191 (3.99) 2.13 (1.77–2.58) 
BMI Q1–3, WC Q1–3 (n = 18 460) 351 (1.32) 1.00 
BMI Q1–3, WC Q4 (n = 1529) 66 (3.11) 1.62 (1.24–2.11) 
BMI Q4, WC Q1–3 (n = 1638) 47 (1.99) 1.44 (1.06–1.95) 
BMI Q4, WC Q4 (n = 5026) 263 (3.80) 1.97 (1.67–2.34) 

HR adjusted for age, sex, civil status, education level, immigrant status, smoking habits, alcohol consumption, physical activities, blood pressure-lowering medication, lipid-lowering medication, systolic blood pressure, leucocyte count and diabetes mellitus.

Discussion

Previous studies have established an association between overweight or obesity and an increased risk of HF. The present population-based cohort study extends these findings showing an independent association between overweight or obesity and risk of first HF hospitalization among middle-aged subjects. The relationships were independent of potential confounders, including multiple biological, lifestyle, and socio-economic factors. Overall weight, BMI, WC and WHR were anthropometric measures significantly associated to the risk of HF hospitalization with largely the same effect sizes for all measures. Although statistically significant, BF% showed weaker relationships with HF.

The underlying mechanism between obesity and HF is complex. Obesity is associated with a higher risk of hypertension,9 insulin resistance and diabetes mellitus,28,29 which result in neurohormonal change and MI; obesity can also cause renal sodium retention, higher leptin30 and inflammation oxidative stress.11,12 All of these circumstances can contribute to haemodynamic overload, leading to left ventricular hypertrophy, which per se increases the risk of HF.31–33 The joint exposure of high WHR or high WC and high BMI further increased the risk in an additive way, which indicates that the location of body fat add additional information about risk of HF.

The main risk factors for HF include age, male sex, diabetes, hypertension, high levels of blood lipids, inflammation, smoking, high alcohol consumption, low physical activity and socio-economic factors.7,10,34,39 In our study, overweight or obesity measured by BMI, WC, WHR or BF%, respectively, emerged as significant independent predictors of HF in multivariate models, taking these risk factors into account. This suggests that obesity by itself or its mediated mechanisms are responsible for HF. Ischaemic heart disease is another major cause of HF. In our additional analysis, the risk for HF hospitalization was only marginally changed after censoring 157 subjects with non-fatal MI during follow-up period.

Strength and limitations

The strength of the study was the large numbers of subjects and events during a long follow-up period. Further, the cardiovascular endpoints, e.g. HF and MI, were retrieved from national registers with high case validity.25,26

A main limitation of the present study is lack of information on type and cause of HF. All HF cases were treated in-hospital as a primary diagnosis, and we were unable to include less severe out-patient HF cases. Because we have no information on the less severe cases, we cannot make any conclusion about them. However, in the MESA (Multi-Ethnic Study of Atherosclerosis) study, including men and women aged 45–84 years without clinically apparent CVD, obesity based on different anthropometric measures was found associated with concentric left ventricular remodeling in men as in women.40 In that same cohort it has also been shown that overweight and obesity based on BMI were associated with differences in the right ventricular morphology, independent of left ventricular measures.41 This supports the view that obesity also is associated with less severe cases of HF.

Another question is whether the study cohort was representative for the background population, as the participation rate of MDC was 41%. A previous study found no substantial difference in terms of socio-demographic structure or in prevalence of smoking and overweight/obesity among participants in the MDC study compared with a mailed health survey (where 75% participated) from the city of Malmö.21 Additionally, we excluded 708 subjects owing to missing data of anthropometric, lifestyle or socio-economic circumstances. In that group there were 36 cases of HF hospitalization during the follow-up period. Mean BMI, waist circumference, body fat percentage was somewhat higher, and rate of hospitalization due to HF significantly higher, among cases with than without missing data (n = 727) (Supplementary table S2).

Change in exposure is an inherent problem in long-term cohort studies. We lack information on anthropometric measures during the follow-up period. It is possible that body fat distribution in some cases changed during the 15 years follow-up. However, this is usually a slow process and one study found that adipose tissue distribution is stable through the lifespan.42 Finally, although we adjusted our analysis for several biological, lifestyle and socio-demographic factors, and because of the observational nature of the study, we cannot exclude the possibility of residual confounding; however, our data may indicate a link between anthropometric measures and an increased risk of HF hospitalization.

This study add to the previous knowledge43 that both general adiposity and abdominal adiposity are associated with increased risk of CVD and death, and that use of waist circumference or waist-to-hip ratio in addition to BMI might be useful when assessing individual risk for CVD and its complications.

Conclusions

In conclusion, elevated BMI, WC, WHR and BF% increase the long-term risk of HF hospitalization with the similar effect. The joint exposure of high BMI and high WHR or high WC further increased the risk in an additive way.

Supplementary data

Supplementary data are available at EURPUB online.

Funding

This work and the Malmö Diet and Cancer study was supported by grants from the Swedish Cancer Society, the Swedish Research Council (Dnr 2011-3891) the Swedish Heart and Lung Foundation (20060294 and 20100244), the Swedish Heart and Lung Foundation, Faculties of medicine, Uppsala University and Lund University, the Malmö City Council and by funds from the Region Skåne, Skåne University Hospital, Malmö and the Lundströms Foundation.

Conflicts of interest: None declared.

Key points

  • There was an independent association between overweight or obesity and risk of first HF hospitalization among middle-aged subjects.

  • In conclusion, elevated BMI, WC, WHR and BF% increase the long-term risk of HF hospitalization.

  • The joint exposure of high BMI and high WHR or high WC further increased the risk in an additive way.

  • Additional information of WC and WHR to BMI might be useful when assessing individual risk for CVD and its complications.

References

1
Obesity: preventing and managing the global epidemic
Report of a WHO consultation
World Health Organ Tech Rep Ser
 , 
2000
, vol. 
894
  
i–xii, 1–253
2
Winter
Y
Rohrmann
S
Linseisen
J
, et al.  . 
Contribution of obesity and abdominal fat mass to risk of stroke and transient ischemic attacks
Stroke
 , 
2008
, vol. 
39
 (pg. 
3145
-
51
)
3
Yusuf
S
Hawken
S
Ounpuu
S
, et al.  . 
Obesity and the risk of myocardial infarction in 27,000 participants from 52 countries: a case-control study
Lancet
 , 
2005
, vol. 
366
 (pg. 
1640
-
9
)
4
Wilson
PW
D'Agostino
RB
Sullivan
L
, et al.  . 
Overweight and obesity as determinants of cardiovascular risk: the Framingham experience
Arch Intern Med
 , 
2002
, vol. 
162
 (pg. 
1867
-
72
)
5
Poirier
P
Giles
TD
Bray
GA
, et al.  . 
Obesity and cardiovascular disease: pathophysiology, evaluation, and effect of weight loss: an update of the 1997 American Heart Association Scientific Statement on Obesity and Heart Disease from the Obesity Committee of the Council on Nutrition, Physical Activity, and Metabolism
Circulation
 , 
2006
, vol. 
113
 (pg. 
898
-
918
)
6
Kenchaiah
S
Evans
JC
Levy
D
, et al.  . 
Obesity and the risk of heart failure
N Engl J Med
 , 
2002
, vol. 
347
 (pg. 
305
-
13
)
7
He
J
Ogden
LG
Bazzano
LA
, et al.  . 
Risk factors for congestive heart failure in US men and women: NHANES I epidemiologic follow-up study
Arch Intern Med
 , 
2001
, vol. 
161
 (pg. 
996
-
1002
)
8
Chen
YT
Vaccarino
V
Williams
CS
, et al.  . 
Risk factors for heart failure in the elderly: a prospective community-based study
Am J Med
 , 
1999
, vol. 
106
 (pg. 
605
-
12
)
9
Stamler
J
Epidemiologic findings on body mass and blood pressure in adults
Ann Epidemiol
 , 
1991
, vol. 
1
 (pg. 
347
-
62
)
10
Engstrom
G
Melander
O
Hedblad
B
Leukocyte count and incidence of hospitalizations due to heart failure
Circ Heart Fail
 , 
2009
, vol. 
2
 (pg. 
217
-
22
)
11
Bahrami
H
Bluemke
DA
Kronmal
R
, et al.  . 
Novel metabolic risk factors for incident heart failure and their relationship with obesity: the MESA (Multi-Ethnic Study of Atherosclerosis) study
J Am Coll Cardiol
 , 
2008
, vol. 
51
 (pg. 
1775
-
83
)
12
Engstrom
G
Hedblad
B
Tyden
P
, et al.  . 
Inflammation-sensitive plasma proteins are associated with increased incidence of heart failure: a population-based cohort study
Atherosclerosis
 , 
2009
, vol. 
202
 (pg. 
617
-
22
)
13
McLaren
L
Socioeconomic status and obesity
Epidemiol Rev
 , 
2007
, vol. 
29
 (pg. 
29
-
48
)
14
Rissanen
AM
Heliovaara
M
Knekt
P
, et al.  . 
Determinants of weight gain and overweight in adult Finns
Eur J Clin Nutr
 , 
1991
, vol. 
45
 (pg. 
419
-
30
)
15
Kahn
HS
Williamson
DF
Stevens
JA
Race and weight change in US women: the roles of socioeconomic and marital status
Am J Public Health
 , 
1991
, vol. 
81
 (pg. 
319
-
23
)
16
Canoy
D
Boekholdt
SM
Wareham
N
, et al.  . 
Body fat distribution and risk of coronary heart disease in men and women in the European Prospective Investigation Into Cancer and Nutrition in Norfolk cohort: a population-based prospective study
Circulation
 , 
2007
, vol. 
116
 (pg. 
2933
-
43
)
17
Loehr
LR
Rosamond
WD
Poole
C
, et al.  . 
Association of multiple anthropometrics of overweight and obesity with incident heart failure: the Atherosclerosis Risk in Communities study
Circ Heart Fail
 , 
2009
, vol. 
2
 (pg. 
18
-
24
)
18
Levitan
EB
Yang
AZ
Wolk
A
, et al.  . 
Adiposity and incidence of heart failure hospitalization and mortality: a population-based prospective study
Circ Heart Fail
 , 
2009
, vol. 
2
 (pg. 
202
-
8
)
19
Nicklas
BJ
Cesari
M
Penninx
BW
, et al.  . 
Abdominal obesity is an independent risk factor for chronic heart failure in older people
J Am Geriatr Soc
 , 
2006
, vol. 
54
 (pg. 
413
-
20
)
20
Berglund
G
Elmstahl
S
Janzon
L
, et al.  . 
The Malmo Diet and Cancer Study. Design and feasibility
J Intern Med
 , 
1993
, vol. 
233
 (pg. 
45
-
51
)
21
Manjer
J
Carlsson
S
Elmstahl
S
, et al.  . 
The Malmo Diet and Cancer Study: representativity, cancer incidence and mortality in participants and non-participants
Eur J Cancer Prev
 , 
2001
, vol. 
10
 (pg. 
489
-
99
)
22
Borne
Y
Smith
JG
Melander
O
, et al.  . 
Red cell distribution width and risk for first hospitalization due to heart failure: a population-based cohort study
Eur J Heart Fail
 , 
2011
, vol. 
13
 (pg. 
1355
-
61
)
23
Lahmann
PH
Lissner
L
Gullberg
B
, et al.  . 
Differences in body fat and central adiposity between Swedes and European immigrants: the Malmo diet and cancer study
Obes Res
 , 
2000
, vol. 
8
 (pg. 
620
-
31
)
24
Li
C
Aronsson
CA
Hedblad
B
, et al.  . 
Ability of physical activity measurements to assess health-related risks
Eur J Clin Nutr
 , 
2009
, vol. 
63
 (pg. 
1448
-
51
)
25
Ingelsson
E
Arnlöv
J
Sundström
J
, et al.  . 
The validity of a diagnosis of heart failure in a hospital discharge register
Eur J Heart Fail
 , 
2005
, vol. 
7
 (pg. 
787
-
91
)
26
Hammar
N
Alfredsson
L
Rosen
M
, et al.  . 
A national record linkage to study acute myocardial infarction incidence and case fatality in Sweden
Int J Epidemiol
 , 
2001
, vol. 
30
 
Suppl. 1
(pg. 
S30
-
4
)
27
Harrell
FE
Jr
Lee
KL
Mark
DB
Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors
Stat Med
 , 
1996
, vol. 
15
 (pg. 
361
-
87
)
28
Bombelli
M
Facchetti
R
Sega
R
, 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
, vol. 
58
 (pg. 
1029
-
35
)
29
Ingelsson
E
Sundstrom
J
Arnlov
J
, et al.  . 
Insulin resistance and risk of congestive heart failure
JAMA
 , 
2005
, vol. 
294
 (pg. 
334
-
41
)
30
Lieb
W
Sullivan
LM
Harris
TB
, et al.  . 
Plasma leptin levels and incidence of heart failure, cardiovascular disease, and total mortality in elderly individuals
Diabetes Care
 , 
2009
, vol. 
32
 (pg. 
612
-
6
)
31
Vasan
RS
Cardiac function and obesity
Heart
 , 
2003
, vol. 
89
 (pg. 
1127
-
9
)
32
Lauer
MS
Anderson
KM
Kannel
WB
, et al.  . 
The impact of obesity on left ventricular mass and geometry. The Framingham Heart Study
JAMA
 , 
1991
, vol. 
266
 (pg. 
231
-
6
)
33
Messerli
FH
Sundgaard-Riise
K
Reisin
ED
, et al.  . 
Dimorphic cardiac adaptation to obesity and arterial hypertension
Ann Intern Med
 , 
1983
, vol. 
99
 (pg. 
757
-
61
)
34
Pencina
MJ
D'Agostino
RB
Larson
MG
, et al.  . 
Predicting the Thirty-year Risk of Cardiovascular Disease: the Framingham Heart Study
Circulation
 , 
2009
, vol. 
119
 (pg. 
3078
-
84
)
35
Levy
D
Larson
MG
Vasan
RS
, et al.  . 
The progression from hypertension to congestive heart failure
JAMA
 , 
1996
, vol. 
275
 (pg. 
1557
-
62
)
36
Smith
JG
Newton-Cheh
C
Almgren
P
, et al.  . 
Assessment of conventional cardiovascular risk factors and multiple biomarkers for the prediction of incident heart failure and atrial fibrillation
J Am Coll Cardiol
 , 
2009
, vol. 
56
 (pg. 
1712
-
19
)
37
Hu
G
Jousilahti
P
Antikainen
R
, et al.  . 
Joint effects of physical activity, body mass index, waist circumference, and waist-to-hip ratio on the risk of heart failure
Circulation
 , 
2010
, vol. 
121
 (pg. 
237
-
44
)
38
Christensen
S
Mogelvang
R
Heitmann
M
, et al.  . 
Level of education and risk of heart failure: a prospective cohort study with echocardiography evaluation
Eur Heart J
 , 
2011
, vol. 
32
 (pg. 
450
-
8
)
39
Dunlay
SM
Weston
SA
Jacobsen
SJ
, et al.  . 
Risk factors for heart failure: a population-based case-control study
Am J Med
 , 
2009
, vol. 
122
 (pg. 
1023
-
8
)
40
Turkbey
EB
McClelland
RL
Kronmal
RA
, et al.  . 
The impact of obesity on the left ventricle: the Multi-Ethnic Study of Atherosclerosis (MESA)
JACC Cardiovasc Imaging
 , 
2010
, vol. 
3
 (pg. 
266
-
74
)
41
Chahal
H
McClelland
RL
Tandri
H
, et al.  . 
Obesity and right ventricular structure and function: the MESA-Right Ventricle Study
Chest
 , 
2012
, vol. 
141
 (pg. 
388
-
95
)
42
Katzmarzyk
PT
Perusse
L
Malina
RM
, et al.  . 
Seven-year stability of indicators of obesity and adipose tissue distribution in the Canadian population
Am J Clin Nutr
 , 
1999
, vol. 
69
 (pg. 
1123
-
9
)
43
Pischon
T
Boeing
H
Hoffmann
K
, et al.  . 
General and abdominal adiposity and risk of death in Europe
N Engl J Med
 , 
2008
, vol. 
359
 (pg. 
2105
-
20
)

Supplementary data

Comments

0 Comments