Introduction

The worldwide prevalence of obesity has nearly tripled during the last 50 years1, and the burden of obesity is expected to increase even further in the coming decade2,3. Heart failure (HF) is also an emerging epidemic, with lifetime risk estimates between 20 and 33% in both sexes4. Excessive adipose tissue accumulation increases the risk of developing HF5,6,7, and recent data indicate that among modifiable risk factors, obesity explained a substantial proportion of incident HF in the general population8,9.

The current definition of overweight and obesity is based on body-mass index (BMI), even though it is known that BMI may not accurately reflect fat mass10. Interestingly, anthropometric measures such as waist circumference (WC) and waist-to-hip ratio (WHR), that more strongly relate with abdominal fat distribution, were not found to be substantially better than BMI in predicting HF risk11,12. With obesity becoming more prevalent and likely becoming a major driver of HF risk, there is a growing need for more “adequate but easy-to-measure” surrogates of fat mass that also better associate with future cardiovascular risk.

Over the past 10 years, several indices of adiposity were developed, that more accurately correlated with body fat distribution and total fat mass. They include body shape index (BSI)13, body roundness index (BRI)14, weight-adjusted-weight index (WWI)15 and relative fat mass (RFM)16. In the current study, we sought to examine for the first time the association between newly developed indices of adiposity and the risk of incident HF in community-dwelling individuals.

Methods

The PREVEND (Prevention of Renal and Vascular End-Stage Disease) study (1997–1998) is a Dutch cohort taken from the general population of Groningen, the Netherlands17,18,19. This study was designed to prospectively evaluate whether increased urinary albumin excretion (UAE > 10 mg/L) in community-dwelling individuals was associated with cardiovascular and renal disease20. In brief, all inhabitants from Groningen, aged 28 to 75 years were asked to respond to a short questionnaire and provide early-morning urine samples (N = 85,421). The response rate was 48% (n = 40,856). Responders with UAE ≥ 10 mg/L (n = 7786) as well as a randomly selected control group with UAE < 10 mg/L (n = 3395) were invited to the outpatient clinic for comprehensive health assessment including responding to questionnaires, anthropometric measurements, fasting blood draw, and urine sampling19. Individuals with type-1 diabetes (defined as insulin requirement), pregnant women (self-reported), and unwilling subjects were excluded from the study19. A final total of 6000 individuals with UAE ≥ 10 mg/L and 2592 individuals with UAE < 10 mg/L constitute the baseline PREVEND cohort (N = 8592)19. From the baseline cohort, we excluded participants with HF at baseline according to hospital records (n = 23), BMI < 18.5 kg/m2 (n = 74), WC < 40 cm (n = 2) and missing covariates (n = 198), resulting in a total of 8295 participants available for analysis (Fig. 1). The current study conformed to the principles drafted in the Helsinki declaration and was approved by the medical ethical committee of the University Medical Center Groningen (UMCG). Written informed consent was received from all participants.

Figure 1
figure 1

Flow diagram: participant selection in the PREVEND study.

Full size image

Baseline measurements

Body weight, height, WC and hip circumference (HC) were measured in a standing position during the baseline visit. WC was measured midway between the lowest rib and the iliac crest at the end of expiration. HC was measured at the widest portion at the level of greater trochanters. BMI was calculated as the ratio between weight and height-squared11,12; overweight was defined as BMI between 25 and 30 kg/m2 and obesity as BMI ≥ 30 kg/m2; WHR was the ratio between WC and HC11,12; BSI was calculated as: [1000 × WC × Weight-2/3 × Height5/6]13; BRI was calculated as: [364.2–365.5 × (1 − ((0.5 × WC/π)2/(0.5 × Height)2))0.5]14; WWI was calculated as: [(WC × 100)/(Weight0.5)]15; and RFM was calculated as: [64 − (20 × Height/WC) + (12 × sex), with sex = 0 (men), and sex = 1 (women)]16.

Smoking was defined as self-reported current smoking or smoking cessation within the previous year. Hypertension was defined as systolic BP (SBP) ≥ 140 mm Hg, diastolic BP (DBP) ≥ 90 mm Hg or self-reported antihypertensive medication usage. Blood pressure (BP) was measured ten times during 10 min using an automatic Dinamap XL Model 9300 series; BP was calculated as the mean of the last two measurements19. Type-2 diabetes was defined as a fasting plasma glucose ≥ 7.0 mmol/L (126 mg/dL), random plasma glucose ≥ 11.1 mmol/L (200 mg/dL), self-reporting of a physician diagnosis or record of glucose-lowering medication use obtained from central pharmacy registry19. History of myocardial infarction and cerebrovascular accident were based on individuals’ medical history derived from a structured questionnaire i.e., hospitalization ≥ 3 days as a result of this condition; this was complemented by a review of the medical report19. Individuals with AF at baseline screening were considered to have prevalent AF. Total cholesterol and plasma glucose were measured by a dry chemistry method (Eastman Kodak, Rochester, New York).

Incident heart failure

Individuals were prospectively followed for the first occurrence of HF or death within 13.5 years of baseline examination. HF records including dates were retrieved from clinical charts. Individuals suspected of having HF were identified according to European Society of Cardiology (ESC) guidelines21. An endpoint adjudication committee of seven independent HF experts further evaluated these selected individuals, and two different experts validated each case. A joint decision was made within the committee in the case of disagreement. Based on left ventricular ejection fraction (LVEF) cutpoint of 50%22, HF was subcategorized into HF with reduced EF (HFrEF) or preserved EF (HFpEF); LVEF was available for all HF cases. Further details can be found elsewhere17.

Statistical analyses

Continuous data are presented as medians, Q1-Q3 (50th percentile, 25th–75th percentile) and categorical variables are represented as percentages. For further analyses, all continuous adiposity measures were standardized. In primary analyses, we examined associations of adiposity indices with incident HF in the total population using Cox regression models adjusting initially for age and sex, and subsequently also for smoking, glucose, cholesterol, systolic blood pressure23, history of myocardial infarction, stroke and atrial fibrillation. We tested for adiposity index × sex terms in the multivariable model. A multiple testing corrected P-value of 0.007 (0.05/7, Bonferroni adjustment) and an interaction P-value (Pint) of 0.1 denoted statistical significance. We also examined the shape of associations of adiposity indices with incident HF using multivariable fractional polynomial models. In secondary analyses, we examined associations of adiposity indices with incident HF according to pre-specified age categories (young < 55 years; middle aged 55–65 years; and old ≥ 65 years), and across BMI categories (lean < 25 kg/m2; overweight 25–30 kg/m2; and obese ≥ 30 kg/m2). Additionally, we evaluated associations of adiposity indices with HFrEF and HFpEF separately. Finally, we used Harrell’s C-statistic, Akaike information criteria (AIC)24,25 and P-values based on likelihood ratio (LHR) test22 to examine the incremental predictive value of adiposity indices (beyond a clinical model) for HF and its subtypes. A PLHR < 0.05 was considered as moderate evidence against the null hypothesis, and a PLHR < 0.01 was considered as strong evidence against the null hypothesis26. All statistical analyses were performed using STATA version-14.

Results

Mean age of the cohort was 49.8 ± 12.6 years and 49.8% (n = 4134) were women. Around 41% of the population was overweight (n = 3416) of which 42% were women, and 16% was obese (n = 1311) of which 54% were women (Table 1). Sex-specific distributions of adiposity indices are shown in Supplementary Table S1. Most adiposity indices showed moderate to strong correlations with each other, and with age in both sexes (Supplementary Table S2).

Table 1 Baseline characteristics of PREVEND participants.
Full size table

During a mean follow-up of 11.3 ± 3.1 years, 363 individuals (4.1%) developed HF. The overall incidence rate of HF was 3.88 per 1000 person-years (5.13 per 1000 person-years in men and 2.67 per 1000 person-years in women). In multivariable Cox regression models, all adiposity indices, except BSI, were significantly associated with incident HF (P < 0.001) (Table 2, Supplementary Fig. S1). While a unit change in standardized BMI was associated with a 28% increased risk of developing HF [hazard ratio per 1-SD increase (HR) 1.28; 95% confidence interval (CI) 1.15–1.43], an equivalent change in RFM was associated with a 67% increased risk [HR 1.67; 95% CI 1.37–2.04]. Sex did not significantly modify the association of any adiposity index with incident HF (Pint > 0.1) (Fig. 2). To check the robustness of results derived from primary analysis, we performed two sensitivity analyses: (i) we substituted glucose and systolic blood pressure with diabetes and hypertension in multivariable models, and (ii) we accounted for death as a competing risk. In both cases, patterns of associations between adiposity indices and incident HF were similar to that observed in the primary analysis (Supplementary Tables S3, S4).

Table 2 Associations of adiposity indices with incident heart failure and its subtypes.
Full size table
Figure 2
figure 2

Associations of adiposity indices with incident heart failure. Models were adjusted for age, sex, smoking, cholesterol, systolic blood pressure, glucose, and history of myocardial infarction, stroke and atrial fibrillation. HR represents the hazard ratio per standard deviation change in adiposity index; CI represents confidence interval; Pinteraction represents the P-value for sex × covariate interaction. AIC Akaike information criterion, BMI body-mass index, BRI body roundness index, BSI body shape index, HF heart failure, HFrEF HF with reduced ejection fraction, HFpEF HF with preserved ejection fraction, RFM relative fat mass, WC waist circumference, WHR waist-to-hip ratio, WWI weight-adjusted-waist index.

Full size image

Associations of adiposity indices with incident HF across age categories, and across BMI categories are shown in Tables 3 and 4. When multivariable models were further adjusted for BMI, all adiposity indices except WC remained significantly associated with incident HF (Supplementary Table S5). Largest effect sizes were again observed with RFM [HR 1.59, 95% CI (1.12–2.25)].

Table 3 Associations of adiposity indices with incident heart failure across age categories.
Full size table
Table 4 Associations of adiposity indices with incident heart failure across body-mass index categories.
Full size table

Among incident HF cases, a total of 120 individuals were classified as HFpEF and 243 were classified as HFrEF. The overall incidence rate of HFpEF was 1.27 per 1000 person-years (1.20 per 1000 person-years in men and 1.34 per 1000 person-years in women), and that of HFrEF was 2.58 per 1000 person-years (3.90 per 1000 person-years in men and 1.32 per 1000 person-years in women). The majority of adiposity indices were strongly associated with both HFpEF and HFrEF in the total population. Amongst adiposity indices, RFM showed strongest associations with both HFpEF [HR 1.76, 95% CI (1.26–2.45)] and HFrEF [HR 1.61, 95% CI (1.25–2.06)]. All adiposity indices were similarly associated with incident HFpEF in both sexes, but RFM, WHR and BMI displayed stronger associations with HFrEF in men than in women (Pint: 0.07, 0.06 and 0.09 respectively) (Table 2).

None of the adiposity indices improved discrimination of HF beyond the clinical model. However, addition of several adiposity indices, including BMI, substantially improved model fit by reducing the prediction error (Table 5, Fig. 2). Strongest improvements were observed after adding BRI and RFM (ΔAIC − 26.5 and − 24.4 respectively). Within HF subtypes, strongest improvement in model fit was observed after adding RFM and BRI for HFpEF, and after adding WHR, BRI and RFM for HFrEF (Table 5).

Table 5 Adiposity indices and improvement in model fit.
Full size table

Discussion

In the current study including 8295 community-dwelling adults, most of the novel as well as established indices of adiposity were strongly associated with incident HF. Amongst adiposity indices, we found that RFM displayed the strongest associations with incident HF.

RFM is a newly-developed parameter of adiposity that more accurately estimates whole-body fat percentage compared to traditional equations based on BMI or WHR16. From a practical standpoint, it is interesting to observe that fat percentage is higher in women than in men27,28, and average RFM values are also higher in women than in men; for instance, in the PREVEND cohort, mean RFM was around 35% in women and around 25% in men. By contrast, most existing measures of adiposity are, on average, higher in men than in women (Table 1). With regard to wide-spread applicability, RFM could potentially be used in routine clinical practice or public health surveillance programmes—even in resource poor settings. This is because RFM not only correlates strongly with HF risk, but can also be calculated using a relatively simple formula, requiring only height and waist circumference—both of which could be determined using a measuring tape.

In subgroup analyses, adiposity indices tended to be more strongly associated with incident HF in younger individuals (i.e., < 55 years). RFM, however, displayed the strongest relative risk across all age groups, including older individuals (i.e., ≥ 65 years). Across BMI categories, adiposity indices, in general, were more strongly associated with incident HF in overweight individuals. Again, RFM had the strongest relative risk across all BMI categories, including in those classified as obese. Most indices of adiposity were also associated with individual HF subtypes, and RFM consistently displayed the largest effect sizes for both HFrEF and HFpEF.

Sex-related differences in associations of BMI with incident HF has been previously reported; while some studies showed stronger associations in men than in women22, other studies showed neutral9 or even opposite trends8. In the current study, we report that both novel as well as established indices of adiposity were similarly associated with incident HF in both sexes. Among HF subtypes, we found that most measures of adiposity were strongly and similarly associated with incident HFpEF in both sexes, suggesting that excessive fat accumulation predisposes both men and women to HFpEF. On the other hand, RFM, BMI and WHR tended to be more strongly associated with HFrEF development in men. Similar trends have been observed for BMI and WC in a meta-analysis of four large cohorts including the PREVEND study29, highlighting that adiposity may be a stronger risk factor for HFrEF in men than in women.

Interestingly, we found that none of the adiposity indices substantially improved discrimination of the clinical HF model. One potential explanation could be that the C-statistic is a less sensitive tool to measure improvement in model fit, especially if the base model already has excellent discrimination22,30. Using more global and sensitive measures of model fit such as LHR χ2 statistic and AIC, we show that RFM and BRI may be more useful than BMI in improving HF risk prediction. Given that RFM and BRI also consistently improved prediction of both HFrEF and HFpEF, future studies should evaluate the potential of including these indices in HF risk estimation.

In summary, we report for the first time, associations between newly-developed adiposity indices (i.e., RFM, BRI and WWI) and incident HF in a large well-characterized cohort. A particular strength of our study was that the HF endpoint was adjudicated, and LVEF was available for all HF cases. The long term follow-up of participants and a 1:1 sex ratio further strengthen our analyses. Nevertheless, we acknowledge that PREVEND is a relatively young cohort with low overall event rates. Another limitation is that the PREVEND study, by design, included a higher proportion of individuals with UAE > 10 mg/L. Finally, the current study was conducted on a predominantly White population from the northern Netherlands, warranting validation of our findings in cohorts from other geographical locations and ethnicities.

Conclusion

Amongst indices of adiposity, relative fat mass displayed the strongest association with HF risk in community-dwelling adults. Future studies should examine the value of including relative fat mass in HF risk prediction models.