Portal de Periódicos da CAPES

Carta Acesso aberto Produção Nacional Revisado por pares

BIBTEX RIS

The “obesity paradox” in an elderly population with a high prevalence of Chagas disease: The 10-year follow-up of the Bambuí (Brazil) Cohort Study of Aging

2012; Elsevier BV; Volume: 166; Issue: 2 Linguagem: Inglês

10.1016/j.ijcard.2012.09.126

1874-1754

Alline Beleigoli, Antônio Luiz Pinho Ribeiro, Maria de Fátima Haueisen Sander Diniz, Maria Fernanda Lima‐Costa, Eric Boersma,

Cardiovascular Function and Risk Factors

Chagas disease (ChD) affects approximately 10 million individuals in Latin America and, due to immigration it is also of increasing importance in North America and Europe. Chronic cardiomyopathy, observed in 20–40% of the cases, is the most important and lethal complication of ChD [[1]Biolo A. Ribeiro A.L. Clausell N. Chagas cardiomyopathy—where do we stand after a hundred years?.Prog Cardiovasc Dis. 2010; 52: 300-316Abstract Full Text Full Text PDF PubMed Scopus (96) Google Scholar]. Control of the transmission by the use of insecticides and aging of the individuals infected in early adulthood are making ChD a health burden in the elderly in old endemic areas. As the prevalence of overweight/obesity in this age group has been increasing, the two conditions are likely to co-exist in older individuals [[2]WHO_CountryProfiles.https://apps.who.int/infobase/CountryProfiles.aspxGoogle Scholar]. The “obesity paradox” (i.e., longer survival of overweight/obese individuals in comparison to lean ones) has been described in older adults with and without cardiovascular diseases (CVD) [[3]Dorner T.E. Rieder A. Obesity paradox in elderly patients with cardiovascular diseases.Int J Cardiol. 2012; 155: 56-65Abstract Full Text Full Text PDF PubMed Scopus (80) Google Scholar]. The etiology of heart disease (HD) might have influence on the phenomenon [[4]Zamora E, Lupon J, deAntonioM, et al. The obesity paradox in heart failure: is etiology a key factor? Int J Cardiol in press. http://dx.doi.org/10.1016/j.ijcard.2011.11.022.Google Scholar]. Whether overweight/obesity are protective determinants of mortality in subjects with Chagas disease (ChD) is still unknown. Our aim was to investigate the relationship between body mass index (BMI), waist circumference (WC) and death, in relation to heart disease (HD), among elderly participants with a high prevalence of ChD in the Bambuí (Brazil) Cohort Study of Aging (BHAS). The BHAS, a cohort study of elderly residents in the Bambuí City (Minas Gerais, southeast of Brazil) is described in detail elsewhere [[5]Lima-Costa M.F. Firmo J.O. Uchoa E. Cohort profile: the Bambui (Brazil) Cohort Study of Ageing.Int J Epidemiol. 2011; 40: 862-867Crossref PubMed Scopus (66) Google Scholar], and was approved by the ethics board of the Fundação Oswaldo Cruz, Belo Horizonte, Brazil. An informed consent form was obtained from all participants. The authors of this manuscript have certified that they comply with the Principles of Ethical Publishing. The outcome of the present analysis was overall death from baseline (1997) to 2007. Trypanosoma cruzi infection, anthropometric measurements (AM) assessment, B-type natriuretic (BNP) test, and definitions of other measurements performed were detailed previously [[5]Lima-Costa M.F. Firmo J.O. Uchoa E. Cohort profile: the Bambui (Brazil) Cohort Study of Ageing.Int J Epidemiol. 2011; 40: 862-867Crossref PubMed Scopus (66) Google Scholar]. AM were repeated in the surviving participants in 2000 and 2002. Underweight (BMI<18.5 kg/m2; n=104; 7.2%) subjects were excluded. ECGs were codified according to the Minnesota code (MC) [[6]Prineas R.J. Crow RS H.B. The Minnesota code manual of electrocardiographic findings. John Wright-PSG, Littleton, MA1982Google Scholar] and classified as abnormal in the presence of major abnormalities [[7]Denes P. Larson J.C. Lloyd-Jones D.M. Prineas R.J. Greenland P. Major and minor ECG abnormalities in asymptomatic women and risk of cardiovascular events and mortality.JAMA. 2007; 297: 978-985Crossref PubMed Scopus (93) Google Scholar] or of frequent supraventricular and ventricular premature beats (MC 8.1.1, 8.1.2 or 8.1.3). HD was defined by the combination of an abnormal ECG and augmented BNP levels. As BNP levels are inversely related to BMI and WC levels in the BHAS [[8]Beleigoli A.M. Lima-Costa M.F. Diniz Mde F. Ribeiro A.L. B-type natriuretic peptide and anthropometric measures in a Brazilian elderly population with a high prevalence of Trypanosoma cruzi infection.Peptides. 2011; 32: 1787-1792Crossref PubMed Scopus (5) Google Scholar], we used distinct cut-off points according to BMI classification: 106 pg/mL in the normal and 128 pg/mL in the high BMI group. Survival rates were compared across the groups formed according to HD status, and to normal (18.5≤BMI<25 kg/m2) or high BMI (BMI≥25 kg/m2), and low (<88 cm for women, <102 cm for men) or high WC (≥88 cm for women, ≥102 cm for men) by Kaplan–Meier (KM) curves and log-rank tests. Overall, 7.2% of all values were missing. We performed multiple imputation of missing values with generation of five complete datasets [[9]Sterne J.A. White I.R. Carlin J.B. et al.Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls.BMJ. 2009; 338: b2393Crossref PubMed Scopus (3264) Google Scholar]. Hazard ratios (HR) and 95% confidence intervals (CI) of death according to BMI/WC (continuous) at various time-points were estimated by extended Cox regression models [[10]Kleinbaum D.G. Klein M. Survival analysis. A self-learning text.2nd ed. Springer, Atlanta, GA2005Google Scholar]. Each model was additionally adjusted for a set of demographic, clinical, socioeconomic and behavioral determinants of death, as well as for a product term of interaction between BMI/WC and HD status. Subsequently, we stratified by ChD, and excluded subjects with probable cachexia (≥10% weight loss weight and death within the first five years of follow-up). Absolute rates of death per unit of BMI were estimated by KM curves. After exclusions and losses to follow-up (78; 5.8%), 1271 participants entered the analysis, 208 (16.4%) of whom had HD. These were older (70.4, SD: 6.7 versus 68.1, SD: 7.2 years; p=0.003) and had a higher prevalence of ChD (137, 65.9% versus 320, 30.0%; p<0.001). Differences between the groups with and without HD in relation to BMI status are depicted in Table 1.Table 1Baseline characteristics, according to groups with and without HD and with normal or high BMI levels.CharacteristicsWithout HD(n 1063; 83.6%)With HD(n 208; 16.4%)Normal BMI(478; 37.6%)Normal BMI(585; 46.0%)Differences*(95% CI)Normal BMI(107; 8.4%)High BMI(101; 7.9%)Differences*(95% CI)Age† (years)68.767.71.071.669.02.6(7.0)(6.5)(0.2, 1.8)(7.6)(6.5)(0.7, 4.5)Female sex‡252403−16.5%5575−22.9%(51.6)(68.1)(−22.2, −10.6)(51.4)(74.3)(−35.6, −10.1)BMI (kg/m2)§22.628.2–21.927.7–(20.9, 23.8)(26.5, 30.5)(20.6, 23.4)(26.4, 29.5)BMI (kg/m2)†22.329.1−6.722.028.7−6.7(1.7)(3.8)(−7.1, −6.5)(1.8)(3.7)(−7.5, −5.9)WC (cm)8699−13.085.298.9−13.7(7.1)(9.1)(−14.0, −12.1)(6.6)(8.4)(−15.7, −11.8)Chagas disease‡1611576.9%795816.4%(33.7)(26.8)(1.3, 12.4)(73.8)(57.4)(3.7, 29.1)Smoking‡106 (22.2)5812.3%261014.4(9.9)(7.8, 16.7)(24.3)(9.9)(4.4, 24.4)BNP (pg/dL)§66.059.0–225.0189.0–(38, 113)(32, 97)(162, 321)(139, 265)BNP (pg/dL)†100.888.212.6305.8258.547.3(137.4)(118.0)(−3.0, 28.2)(268.5)(234.5)(−21.1, 115.7)CRP (mg/dL)§2.313.94–3.154.94–(1.06, 4.92)(1.84, 6.88)(1.27, 7.29)(2.30, 8.66)CRP (pg/dL)†4.736.01−1.35.828.09−2.3(9.39)(10.12)(−2.5, −0.1)(8.33)(1.20)(−5.0, 0.4)Systolic blood pressure (mm Hg)†136138−2.4141144−3.8(23)(20)(−5.0, 0.3)(26)(26)(−10.8, 3.2)Diabetes mellitus‡47117−10.1%1119−8.5%(9.9)(20.0)(−14.3, −5.9)(10.3)(18.8)(−18.1, 1.0)Digoxin use‡5375−1.7%27232.4%(11.1)(12.8)(−5.6, 2.2)(25.2)(22.8)(−9.1, 14.1)Anti-hypertensive medication use‡176360−24.7%5969−13.2%(36.8)(61.5)(−30.6, −18.9)(55.1)(68.3)(−26.3, −0.1)Serum creatinine (mg/dL)§0.850.83–0.900.90–(0.75, 0.97)(0.73, 0.97)(0.79, 1.09)(0.79, 1.08)Serum creatinine (mg/dL)†0.900.86−0.020.990.950.00(0.35)(0.20)(−0.07, 0.00)(0.49)(0.29)(−0.08, 0.14)Total cholesterol (mg/dL)†231239−7.9233234−0.8(49)(49)(−13.8, −1.9)(53)(47)(−14.3, 12.8)Physically active‡98156−6.2%1921−3.0%(20.2)(26.4)(−11.3, −1.2)(17.8)(20.8)(−13.8, 7.7)Family income‡Lower33734211.5%84781.3%(69.9)(58.4)(5.8, 17.3)(78.5)(77.2)(−10.0, 12.6)Intermediate110176−7.2%1919−1.1%(22.8)(30.0)(−12.5, −1.9)(17.8)(18.8)(−11.6, 9.5)Higher3568−4.3%44−0.2%(7.3)(11.6)(−7.8, −0.9)(3.7)(4.0)(−5.5, 5.0)Education‡Lower16012811.3%50416.1%(32.9)(21.6)(6.0, 16.6)(46.7)(40.6)(−7.3, 19.6)Intermediate278368−5.0%5255−5.9%(57.2)(62.2)(−10.8, 0.9)(48.6)(54.5)(−19.4, 7.7)Higher4896−6.3%55−0.3%(9.9)(16.2)(−10.3, −2.4)(4.7)(5.0)(−6.1, 5.5)HD heart disease; BMI body mass index; WC waist circumference; BNP B-type natriuretic peptide; CRP C-reactive protein.*Differences between means and proportions; continuous variables are described by means (SD)† or median (IQR)§ and categorical variables by frequencies (%)‡. Open table in a new tab HD heart disease; BMI body mass index; WC waist circumference; BNP B-type natriuretic peptide; CRP C-reactive protein. *Differences between means and proportions; continuous variables are described by means (SD)† or median (IQR)§ and categorical variables by frequencies (%)‡. Mean follow-up time was 9.0 years. Deaths occurred in 128 (61.5%) and 310 (29.2%) subjects with and without HD, respectively. High BMI/WC were associated with the lowest survival rates at 10-year follow-up regardless of HD status (Fig. 1, Fig. 2). After full adjustment, the relationship between mortality and BMI and WC, was U-shaped and non-significant, respectively. These results were similar after exclusion of participants with probable cachexia, and regardless of HD and ChD status (Table 2). BMI between 30 and 32 kg/m2 was associated with the lowest absolute mortality rates at 10-year follow-up in participants with (46–47%) and without HD (20–21%).Fig. 2Comparison of survival rates among groups with and without heart disease (HD), according to low/high waist circumference (WC).View Large Image Figure ViewerDownload Hi-res image Download (PPT)Table 2Hazard ratios (HR) and 95% confidence intervals (95% CI) for mortality related to BMI and WC in the overall population and in the subgroups of interest.Anthropometric measuresHR (95% CI)p value for interaction between HD and the anthropometric measureUnadjustedaUnadjusted: the BMI quadratic term is included in BMI models (HR not shown).Fully adjustedbFully-adjusted: BMI and BMI square/WC plus age (continuous), sex, HD (no, yes), ChD (no, yes), current smoking (no, yes), diabetes (no, yes), total cholesterol (continuous), log-transformed creatinine (continuous), log-transformed C-reactive protein (continuous), physical activity (no, yes), household income, education (low, intermediate, high).All participantsBMI0.770.820.48(0.73–0.82)(0.76–0.88)WC1.001.000.60(0.99–1.01)(0.99–1.01)Participants with ChDBMI0.650.700.13(0.52–0.83)(0.54–0.89)WC0.991.000.94(0.98–1.01)(0.99–1.01)Participants without ChDBMI0.770.790.22(0.70–0.84)(0.72–0.87)WC1.001.000.22(0.99–1.01)(0.99–1.01)Participants without weight loss or early deathBMI0.820.880.98(0.76–0.89)(0.81–0.96)WC1.001.010.99(0.99–1.02)(0.99–1.02)BMI body mass index; WC waist circumference; HD heart disease; ChD Chagas disease.a Unadjusted: the BMI quadratic term is included in BMI models (HR not shown).b Fully-adjusted: BMI and BMI square/WC plus age (continuous), sex, HD (no, yes), ChD (no, yes), current smoking (no, yes), diabetes (no, yes), total cholesterol (continuous), log-transformed creatinine (continuous), log-transformed C-reactive protein (continuous), physical activity (no, yes), household income, education (low, intermediate, high). Open table in a new tab BMI body mass index; WC waist circumference; HD heart disease; ChD Chagas disease. Our results are similar to studies which found a protective role of overweight/obesity in the prognosis of subjects with HD [4Zamora E, Lupon J, deAntonioM, et al. The obesity paradox in heart failure: is etiology a key factor? Int J Cardiol in press. http://dx.doi.org/10.1016/j.ijcard.2011.11.022.Google Scholar, 11Oreopoulos A. Padwal R. Norris C.M. Mullen J.C. Pretorius V. Kalantar-Zadeh K. Effect of obesity on short- and long-term mortality postcoronary revascularization: a meta-analysis.Obesity (Silver Spring). 2008; 16: 442-450Crossref PubMed Scopus (185) Google Scholar]. Regarding the association between WC and mortality, previous findings in populations with HF are heterogeneous [12Testa G. Cacciatore F. Galizia G. et al.Waist circumference but not body mass index predicts long-term mortality in elderly subjects with chronic heart failure.J Am Geriatr Soc. 2010; 58: 1433-1440Crossref PubMed Scopus (39) Google Scholar, 13Clark AL, Fonarow GC, Horwich TB. Waist circumference, body mass index, and survival in systolic heart failure: the obesity paradox revisited. J Card Fail 17(5):374–80.Google Scholar]. Not only high BMI being a marker of greater muscle mass, but also benefits associated with high fat mass, such as an increased strength capacity can explain these findings [[14]Zavin A, Daniels K, Arena R, et al. Adiposity facilitates increased strength capacity in heart failure patients with reduced ejection fraction. Int J Cardiol 2012 in press. http://dx.doi.org/10.1016/j.ijcard.2012.06.007.Google Scholar]. Neither reverse causation due to cachexia in elderly with normal BMI nor neutralization of the inflammatory effects of tumor necrosis factor (TNF)-alpha by soluble receptors in the adipose tissue [[3]Dorner T.E. Rieder A. Obesity paradox in elderly patients with cardiovascular diseases.Int J Cardiol. 2012; 155: 56-65Abstract Full Text Full Text PDF PubMed Scopus (80) Google Scholar] seems plausible explanations to our results, as suggested by the sensitivity analysis and by the highest CRP levels in subjects with both high BMI and HD, respectively. A healthier status of overweight/obese in comparison to lean subjects was not observed either. Our study is unique in investigating the “obesity paradox” in older adults with ChD. The use of both BMI and WC directly measured at various time-points, the exclusion of subjects with underweight and probable cachexia, the long-term follow-up with minimal number of losses, and the high rate of events are major strengths. Limitations due to the small number of subjects with BMI≥40 kg/m2 (12; 1.1%), and to the lack of more accurate measurements of fat mass and left ventricle ejection function warrant mention. In conclusion, high BMI levels are associated with higher survival regardless of HD status in an elderly population with a high prevalence of ChD, whereas high WC values do not influence on mortality. The “obesity paradox” should be taken into account when weight control is planned for elderly subjects with HD and ChD. This work was supported by Financiadora de Estudos e Projetos, Rio de Janeiro, Brazil ; the Ministério da Saúde, Brasília, Brazil ; and the Fundação de Amparo à Pesquisa do Estado de Minas Gerais, Belo Horizonte, Brazil . M.F. Lima-Costa and A.L. Ribeiro are fellows of the Conselho Nacional de Desenvolvimento Científico e Tecnológico. Beleigoli, AM was supported by the Programa de Doutorando com Estágio no Exterior (PDEE) do Conselho de Aperfeiçoamento de Pessoal Superior (CAPES), Brazil .