Numerous studies have documented an obesity paradox in which the overweight and obese elderly have a better prognosis than those with ideal body weight. Good prognosis among the overweight or obese elderly may reflect the relative safety of storing the harmful lipophilic chemicals, known as persistent organic pollutants (POPs), in adipose tissue rather than in other critical organs. Therefore, we hypothesized lower mortality among the obese elderly with a higher body burden of POPs, but this pattern may not exist among the obese elderly with a lower body burden of POPs.
Using the National Health and Nutrition Examination Survey (NHANES) 1999–2004 study with a mean 4.2-year follow-up, we tested whether the association between fat mass and total mortality in 635 (652 for organochlorine pesticides) elderly participants aged ⩾70 years differed depending on serum concentrations of 23 POPs.
There were statistically significant interactions between fat mass and POPs in predicting total mortality. In those with low POP concentrations, there was no obesity paradox; mortality increased with fat mass (hazard ratios about 2–3 in the highest vs lowest quintile of fat mass). However, consistent with an obesity paradox, these patterns completely disappeared in those with high POP concentrations. Compared with the lowest quintile of fat mass, statistically significantly lower mortality was observed in the elderly in the third to fifth quintiles of fat mass. In the case of polychlorinated biphenyls, the mortality in the highest quintile of fat mass was only one-fifth of that in the lowest quintile.
These findings are consistent with our hypothesis that adipose tissue provides relatively safe storage of toxic lipophilic chemicals, a phenomenon that could explain the obesity paradox. Although weight loss may be beneficial among the obese elderly with low POP concentrations, weight loss in the obese elderly with higher serum concentrations of POPs may carry some risk.
Although overweight and obesity are well-established independent risk factors for chronic disease development in the general population, numerous studies have documented an obesity paradox in which overweight and obese people with various conditions including old age, cardiovascular diseases, cancer or chronic kidney diseases have a better prognosis than do patients with ideal body weight.1, 2, 3, 4
Several mechanisms have been discussed as possible explanations.1, 2, 3, 4 For example, the inaccurate assessment of body fatness by the conventional body mass index has been most commonly suggested as a potential explanation. However, studies using directly measured body fat mass observed a similar obesity paradox.5 Other speculations include a survival effect, increased lean body mass, more aggressive treatment or better nutritional reservation among obese patients or the elderly,1, 2, 3, 4 but data specifically supporting these possibilities are not available.
Recently, background exposure in the general population to persistent organic pollutants (POPs), lipophilic chemicals that accumulate in adipose tissue, has been strongly linked to a variety of chronic diseases.6, 7, 8, 9, 10, 11, 12, 13 POPs, including polychlorinated dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans (PCDFs), polychlorinated biphenyls (PCBs) and organochlorines pesticides (OCPs) have been highlighted as being chemicals of concern.
POPs metabolize very slowly in humans, with half-lives of several years or decades. As the natural organ for storage of fat and fat-soluble substances, adipose tissue may be a relatively safer depot for POP storage than other critical organs. From this viewpoint, high body fat mass may not be harmful or could even provide the benefit of affording enhanced capacity for less harmful storage of POPs. Therefore, we hypothesized that the association between obesity and mortality would differ depending on POP levels. Specifically, we expected good prognosis among the obese elderly with a higher body burden of POPs, but a poorer prognosis among the obese elderly with a lower body burden of POPs. This hypothesis was tested in subsamples of the National Health and Examination Survey (NHANES) 1999–2004 mortality follow-up database. Fat mass derived from dual X-ray absorptiometry was used as indices of obesity.
Materials and methods
The continuous NHANES, conducted annually since 1999 by the Centers for Disease Control and Prevention, is an ongoing survey designed to measure the health and nutritional status of the civilian noninstitutionalized US population. In the NHANES 1999–2002, all POPs were measured in the same subsample, whereas in the NHANES 2003–2004 OCPs were measured in a subsample different from the one in which other POPs were measured. Thus, the final sample sizes were 635 for PCDDs, PCDFs and PCBs and 652 for OCPs, including subjects aged 70 years or older and followed up for mortality through the end of 2006.
Venous blood samples were collected and shipped weekly at −20 °C. PCDDs, PCDFs, PCBs and OCPs were all measured as individual chemicals by high-resolution gas chromatography/high-resolution mass spectrometry using isotope dilution for quantification. We selected 3 PCDDs (1,2,3,6,7,8,-hexachlorodibenzo-p-dioxin, 1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin and 1,2,3,4,6,7,8,9-octachlorodibenzo-p-dioxin), 3 PCDFs (2,3,4,7,8-pentachlorodibenzofuran, 1,2,3,4,7,8-hexachlorodibenzofuran and 1,2,3,4,6,7,8-heptachlorodibenzofuran), 12 PCBs (PCB028, PCB074, PCB099, PCB118, PCB126, PCB146, PCB153, PCB156, PCB169, PCB170, PCB180 and PCB187) and five OCPs (p,p′-DDE, trans-nonachlor, oxychlordane, heptachlor epoxide and β-hexachlorocyclohexane) for which at least 80% of the study subjects had concentrations more than the limit of detection. The fat mass was measured by dual X-ray absorptiometry using Hologic QDR 4500A fan-beam densitometers (Hologic, Bedford, MA, USA).
Probabilistic matching was used to link NHANES participants with the National Death Index to ascertain vital status. Matching was based on 12 identifiers for each participant (for example, Social Security number, sex and date of birth). Persons who survived the entire follow-up period were administratively censored on 31 December 2006. Follow-up time for each person was calculated as the difference between the NHANES examination date and the last known date alive or censored.
First, we checked whether there were statistically significant interactions of serum concentrations of individual POPs (tertiles) with fat mass (quintiles) in predicting total mortality using Cox proportional hazard models. Cutoff points of tertiles of individual POPs were presented in Supplementary Table S1. In addition to individual POPs, we also used the summary measures of subclasses of POPs by summing the rank orders of the individual POPs belonging to each subclass for subjects with detectable values of each POP, assigning rank 0 to values that were not detectable. Second, we calculated hazard ratios on the associations of total mortality with fat mass by stratifying subclasses of POPs, which showed statistically significant or marginally significant interactions. Adjusting covariates were age, gender, race ethnicity, smoking status and physical activity.
All statistical analyses were performed with the SAS 9.1 (SAS Institute Inc., Cary, NC, USA) and SUDAAN 9.0 (Research Triangle Institute, Research Triangle Park, NC, USA). Estimates of main results were calculated accounting for stratification and clustering,14 adjusting for age, gender, race ethnicity, smoking status and physical activity instead of using sample weights; this adjustment has been regarded as a good compromise between efficiency and bias.14, 15 As results were very similar with SAS 9.1 and SUDAAN 9.0, we present the results based on SAS 9.1.
Among 635 elderly with information on PCDDs, PCDFs and PCBs, 150 subjects (underlying causes: 51 cardiovascular, 39 cancer, 17 respiratory disease and 43 others) died during the mean follow-up time of 4. 2 years. Among 652 elderly with information on OCPs, the number of deaths was 135. Data of the 635 participants with regard to distributions of several factors including demographic variables according to fat mass are shown in Table 1. The elderly with more fat mass were likely to be younger, women, non-smokers and physically inactive. They also had a higher prevalence of diabetes or hypertension than the elderly with less fat mass. Results based on 652 participants with information on OCPs were similar to that shown in Table 1 (data not shown). Among all subjects, compared with those with the lowest fat mass, the adjusted hazard ratios were 1.0, 0.8, 0.6 (95% CI: 0.3–0.9), 0.9 and 0.9 across fat mass quintiles.
Table 2 shows P-values for interactions of fat mass with individual POPs in prediction of total mortality after adjusting for age, gender, race, cigarette smoking and physical activity. Among 23 individual POPs included in the analyses, all 3 PCDDs, 2 out of 3 PCDFs, 11 out of 12 PCBs and 1 out of 5 OCPs showed strong interactions with fat mass in the prediction of total mortality. The summary measure of PCDDs, PCDFs and PCBs also showed strong interactions with fat mass. In POPs with dioxin activity, there was no clear pattern according to toxic equivalent factor. Among five OCPs, only heptachlor epoxide showed a meaningful significant interaction.
Table 3 shows results of stratified analyses by summary measures of those POP subclasses, which showed statistically significant or marginally significant interactions in Table 2. Patterns of associations between fat mass and total mortality differed by POP concentrations. When the elderly had low levels of PCDDs, PCDFs or PCBs, those in the highest quintile of fat mass showed about two times higher mortality rates compared with those in the lowest quintile of fat mass. Adjusted hazard ratios of fat mass were 1.0, 1.8, 2.0, 3.0 (95% CI: 1.2–7.8) and 1.6 across fat mass quintiles among the elderly within the lowest tertile of heptachlor epoxide belonging to OCPs. However, these patterns completely disappeared when the elderly had high levels of POPs. Statistically significantly lower mortality rates in the elderly with high fat mass (third to fifth quintile) were observed. The inverse association between fat mass and mortality was particularly strong with PCBs; the mortality was one-fifth as great among the elderly in the highest quintile of fat mass compared with those in the lowest quintile of fat mass. The interactions between fat mass and POPs were displayed in Figure 1.
Among all subjects aged 70 years or older, compared with the lowest fat mass group, elderly persons who were in the third quintile of fat mass had a significantly lower mortality, and those in the fourth or fifth quintile of fat mass did not show increased risk of mortality, consistent with the well-known obesity paradox in the elderly. Although methodological problems such as survivor bias, selection bias, lead time bias or, in meta-analyses, publication bias and confounders have been discussed in previous studies,2 no explanation has yet provided data that clearly explained the paradox.
This study provides an avenue by which POPs and their storage in fat tissue could contribute to the obesity paradox in the elderly. For example, when we restricted analyses to subjects with relatively low POPs, there was no obesity paradox: compared with the lowest fat mass group, the risk of mortality among the elderly with high fat mass was two to three times higher than among those with low fat mass. On the other hand, the obesity paradox was strongly observed among the elderly with high levels of POPs. When the elderly had relatively high POPs, those with high fat mass showed a lower risk of mortality than those with low fat mass. This finding suggests a possible protective effect of obesity against harmful lipophilic chemicals such as POPs. Supporting the idea that harm could occur if POPs were not harbored in adipose tissue, POPs released into plasma during weight loss among obese persons were associated with a decrease in serum T3 concentration, resting metabolic rate and the activity of skeletal muscle oxidative enzymes.16, 17
Besides the elderly, the obesity paradox is widely observed among patients with a variety of chronic diseases including cardiovascular diseases or cancer. Recent epidemiological studies observed strong associations between serum concentrations of POPs and a variety of chronic diseases in the same NHANES data set,6, 7, 8, 9, 10, 11 as well as with incident diabetes and related variables in other studies.12, 13 Therefore, patients with these diseases tend to have higher levels of POPs, similar to the elderly with high POPs. This observation could help explain why the obesity paradox is observed among patients with a variety of chronic diseases, as well as the elderly.
Within the concept of the obesity paradox, efficacy and safety of weight loss can be questioned among the obese elderly or patients with chronic diseases. One study reported that intentional weight loss among overweight and obese patients with coronary artery disease showed marked improvements in obesity indices, lipid profiles and inflammation markers, but there was only a nonsignificant trend for lower mortality, suggesting that obesity might have some benefit that traded off against the beneficial effects of improving traditional CVD risk factors.5 If obese elderly or patients with chronic diseases had high serum concentrations of POPs, our findings may indicate that fat mass has such a beneficial effect as a relatively safe storage site of lipophilic chemicals such as POPs. If an elderly person was found to have high POP levels, our findings imply that careful consideration should be given before instituting intervention leading to weight loss. On the other hand, weight loss may be beneficial among the obese elderly with low POP concentrations.
Importantly, it is well known that spatial and temporal trends of POPs vary greatly among populations.18, 19 Therefore, the interactions between obesity and POPs in prediction of mortality may or may not be observed depending on the distribution of POPs in a specific population. For example, if a population was studied in which most people had high POP concentrations, for example, similar to or higher than the third tertile of POPs in this study subjects, it may be difficult to observe the interactions between obesity and POPs in the prediction of mortality.
In fact, the relationships between obesity and POPs themselves are not simple. Multidirectional associations between these two closely related factors are possible. Obesity itself can increase the half-lives of POPs.20 At the same time, exposure to low-level POPs can itself cause obesity even though high-dose POPs can lead to weight loss, likely due to cell toxicity.12, 21 In addition, weight loss leads to increased serum concentrations of POPs, whereas weight gain leads to decreased serum POPs.22 Among the elderly, the evolution of obesity over time during adulthood and aging-related body composition and fat distribution changes may affect the dynamics of POPs in the body. It would be desirable to disentangle these complicated relationships in studies that include repeated measurements of fat mass and concentrations of POPs with follow-up for incident disease and intermediate factors.
One limitation of this study was that, although power was adequate for the study of total mortality, we had restricted power for analysis of cause-specific mortality due to the small number of death cases. Nevertheless, when we applied the same analyses to the 51 cardiovascular deaths, we observed a similar pattern of interaction between fat mass and POPs. In addition, the elderly within the lowest tertile of POPs are not sufficient to evaluate the pure effect of obesity on mortality, because of strong correlations of POPs with age, so that even the lowest tertile had substantial POPs concentrations. The slight decrease of mortality in the second or third quintile of fat mass compared with the elderly within the lowest tertile of POPs might disappear if analyses could be restricted to the elderly with less POP contamination. However, in a modern chemically contaminated society, such a subpopulation may not exist. Finally, we could not directly test whether the same interaction between fat mass and POPs on mortality is observed among patients with chronic diseases. On the basis of previous findings on the associations between POPs and chronic diseases in the same NHANES data set, we could expect that patients with these chronic diseases had higher concentrations of POPs.
In conclusion, the different associations between fat mass and mortality by levels of POPs are consistent with our hypothesis that adipose tissue provides relatively safe storage of toxic lipophilic chemicals. This phenomenon may contribute to explaining the obesity paradox. Additional research on this topic would be valuable.
Janssen I, Mark AE . Elevated body mass index and mortality risk in the elderly. Obes Rev 2007; 8: 41–59.
Oreopoulos A, Kalantar-Zadeh K, Sharma AM, Fonarow GC . The obesity paradox in the elderly: potential mechanisms and clinical implications. Clin Geriatr Med 2009; 25: 643–659, viii.
Romero-Corral A, Montori VM, Somers VK, Korinek J, Thomas RJ, Allison TG et al. Association of bodyweight with total mortality and with cardiovascular events in coronary artery disease: a systematic review of cohort studies. Lancet 2006; 368: 666–678.
Zamboni M, Mazzali G, Zoico E, Harris TB, Meigs JB, Di Francesco V et al. Health consequences of obesity in the elderly: a review of four unresolved questions. Int J Obes (Lond) 2005; 29: 1011–1029.
Lavie CJ, Milani RV, Artham SM, Patel DA, Ventura HO . The obesity paradox, weight loss, and coronary disease. Am J Med 2009; 122: 1106–1114.
Ha MH, Lee DH, Jacobs DR . Association between serum concentrations of persistent organic pollutants and self-reported cardiovascular disease prevalence: results from the National Health and Nutrition Examination Survey, 1999–2002. Environ Health Perspect 2007; 115: 1204–1209.
Ha MH, Lee DH, Son HK, Park SK, Jacobs Jr DR . Association between serum concentrations of persistent organic pollutants and prevalence of newly diagnosed hypertension: results from the National Health and Nutrition Examination Survey 1999–2002. J Hum Hypertens 2009; 23: 274–286.
Lee DH, Jacobs DR, Kocher T . Associations of serum concentrations of persistent organic pollutants with the prevalence of periodontal disease and subpopulations of white blood cells. Environ Health Perspect 2008; 116: 1558–1562.
Lee DH, Lee IK, Porta M, Steffes M, Jacobs Jr DR . Relationship between serum concentrations of persistent organic pollutants and the prevalence of metabolic syndrome among non-diabetic adults: results from the National Health and Nutrition Examination Survey 1999–2002. Diabetologia 2007; 50: 1841–1851.
Lee DH, Lee IK, Song K, Steffes M, Toscano W, Baker BA et al. A strong dose-response relation between serum concentrations of persistent organic pollutants and diabetes: results from the National Health and Examination Survey 1999–2002. Diabetes Care 2006; 29: 1638–1644.
Lee DH, Steffes M, Jacobs DR . Positive associations of serum concentration of polychlorinated biphenyls or organochlorine pesticides with self-reported arthritis, especially rheumatoid type, in women. Environ Health Perspect 2007; 115: 883–888.
Lee D . Low dose organochlorine pesticides and polychlorinated biphenyls predict obesity, dyslipidemia, and insulin resistance among people free of diabetes. PLoS One 2011.
Lee DH, Steffes MW, Sjodin A, Jones RS, Needham LL, Jacobs Jr DR . Low dose of some persistent organic pollutants predicts type 2 diabetes: a nested case-control study. Environ Health Perspect 2010; 118: 1235–1242.
Korn EL, Graubard BI . Epidemiologic studies utilizing surveys: accounting for the sampling design. Am J Public Health 1991; 81: 1166–1173.
Graubard BI, Korn EL . Analyzing health surveys for cancer-related objectives. J Natl Cancer Inst 1999; 91: 1005–1016.
Imbeault P, Tremblay A, Simoneau JA, Joanisse DR . Weight loss-induced rise in plasma pollutant is associated with reduced skeletal muscle oxidative capacity. Am J Physiol Endocrinol Metab 2002; 282: E574–E579.
Pelletier C, Doucet E, Imbeault P, Tremblay A . Associations between weight loss-induced changes in plasma organochlorine concentrations, serum T concentration, and resting metabolic rate. Toxicol Sci 2002; 67: 46–51.
Aguilar A, Borrell A, Reijnders PJ . Geographical and temporal variation in levels of organochlorine contaminants in marine mammals. Mar Environ Res 2002; 53: 425–452.
Ueno D, Watanabe M, Subramanian A, Tanaka H, Fillmann G, Lam PK et al. Global pollution monitoring of polychlorinated dibenzo-p-dioxins (PCDDs), furans (PCDFs) and coplanar polychlorinated biphenyls (coplanar PCBs) using skipjack tuna as bioindicator. Environ Pollut 2005; 136: 303–313.
Longnecker MP . Pharmacokinetic variability and the miracle of modern analytical chemistry. Epidemiology 2006; 17: 350–351.
Ruzzin J, Petersen R, Meugnier E, Madsen L, Lock EJ, Lillefosse H et al. Persistent organic pollutant exposure leads to insulin resistance syndrome. Environ Health Perspect 2010; 118: 465–471.
Lim JS, Son HK, Park SK, Jacobs Jr DR, Lee DH . Inverse associations between long-term weight change and serum concentrations of persistent organic pollutants. Int J Obes (Lond) 2011; 35: 744–747.
This work was partly supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MEST; no. 2011-0004692).
The authors declare no conflict of interest.
Supplementary Information accompanies the paper on International Journal of Obesity website
About this article
Cite this article
Hong, NS., Kim, KS., Lee, IK. et al. The association between obesity and mortality in the elderly differs by serum concentrations of persistent organic pollutants: a possible explanation for the obesity paradox. Int J Obes 36, 1170–1175 (2012). https://doi.org/10.1038/ijo.2011.187
- obesity paradox
- persistent organic pollutants
- fat mass
Current Diabetes Reports (2017)
Environmental Science and Pollution Research (2016)
International Journal of Obesity (2015)
Adipose tissue compartments, muscle mass, muscle fat infiltration, and coronary calcium in institutionalized frail nonagenarians
European Radiology (2015)