Abstract
Background
The obesity paradigm has been a health concern globally for many years, its meaning is controversial. In this study, we assess the characteristics and causes of obesity paradigm and detail the mediation of obesity and inflammation on survival.
Methods
The original cohort included participants from the US National Health and Nutrition Examination Survey (NHANES) from 1999 to 2018, a prospective cohort of a nationally representative sample of adult participants; the oncology validation cohort included patients from the Investigation on Nutrition Status and Clinical Outcome of Common Cancers (INSCOC) from 2013 to 2021, a prospective cohort of Chinese patients with cancer. Survival analysis was performed using weighted (NHANES) or unweighted (INSCOC) Cox survival analyses. The normal BMI group was used as a reference for all comparisons. Systemic inflammation was defined as neutrophil‐to‐lymphocyte ratio (NLR) > 3. Model‐based causal mediation analysis was used to identify the mediators.
Results
A total of 52 270 (weighted population: 528506229) participants of the NHANES [mean follow‐up times: 10.2 years; mean age (SD): 47 (19.16) years] were included in the original cohort; and a total of 17 418 patients with cancer of INSCOC [mean follow‐up times: 2.9 years; mean age (SD): 57.37 (11.66) years] were included in the validation cohort. In the subgroups of all the participants, the obesity paradigm was more apparent in older participants and participants with disease [HR (95% CI): age ≥ 65 years, 0.84 (0.76, 0.93); with cancer, 0.84 (0.71, 0.99); with CVD, 0.74 (0.65, 0.85)]. As aged, the protective effect of a high BMI on survival gradually increased and a high BMI showed the effect of a protective factor on older participants [for obese II, HR (95% CI): young adults, 1.91 (1.40, 2.62); middle age, 1.56 (1.28, 1.91); old adults, 0.85 (0.76, 0.96]). The aged‐related obesity paradigm in patients with cancer from the NHANES was verified in the INSCOC cohorts [for obese, HR (95%CI): 0.65 (0.52, 0.81)]. The NLR is an important mediator of the effect of BMI on survival (proportion of mediation = 15.4%).
Conclusions
The obesity paradigm has a strong correlation with age. Relative to normal weight, obese in young people was association with higher all‐cause mortality, and obese in elderly people was not association with higher mortality. The protection of obesity is association with systemic inflammation.
Keywords: Aging, Cancer, Inflammation, Obesity paradigm
Introduction
Global obesity rates have been rising over the past half‐century, increasing the burden of obesity‐related diseases. The World Health Organization (WHO) estimates that obesity among adults increased tripled from 1975 to 2021. 1 Obesity increases the risk of diseases, including cancer, cardiovascular diseases (CVDs), and kidney diseases. 2 , 3 However, studies have shown that BMI‐related death curves are generally U‐shaped. Overweight or even obesity is associated with improved survival in patients with CVDs, or cancer, and in critically ill patients. 4 , 5 This phenomenon that obesity contributes to the development of chronic diseases and also seems to improve survival in these patients was called the obesity paradigm. Sex, age, nutritional status, and disease, which affected the survival of obese people, might lead to the obesity paradigm. 6 A study on ICU patients showed that high BMI is a risk factor for co‐morbidities and critically ill‐related deaths but was a protective factor for 90‐day and 2‐year survival, which could be due to age. 7 , 8 Weight‐related survival increases with age and obesity may be a protective factor for the long‐term prognosis of the chronic wasting disease. 9 Some researchers regard obesity as a pathology that can cause low‐grade chronic inflammatory and pathological diseases and increase the risk of other diseases and believe that the obesity paradigm is due to methodological limitations. 10 , 11 To support their argument, many researchers have made various adjustments to the obesity paradigm, but some biases are not fully explained by methodological limitations. 12 , 13 Disease and obesity have a huge influence in patients, among which the effect on inflammation has been widely concerned. Previous studies have shown that the differences in metabolic caused by obesity affect anti‐tumour immunity. 14
Therefore, this study aims to find the association of obesity paradigm with age.
Methods
Study population
This prospective study used two cohorts from the United States and China, respectively. The US National Center for Health Statistics National Health and Nutrition Examination Survey (NHANES) has been conducting 2‐year cycles since 1999 to monitor the health and nutritional status of the US population. All NHANES protocols were approved by the National Center for Health Statistics Ethics Review Board and written informed consent was obtained from all participants. Participants in 10 cycles of the NHANES from 1999 to 2018 were included in this study. The Investigation on Nutrition Status and Clinical Outcome of Common Cancers (INSCOC) is a multicenter, large‐scale, long‐term follow‐up prospective study aimed to investigate malnutrition in patients with cancer and identify risk factors associated with negative outcomes. The details of INSCOC had been published elsewhere. 15
The inclusion criteria for this study were as follows: age ≥ 18 years, having a measured BMI, and having neutrophil and lymphocyte ratio. The INSCOC was approved by the medical ethics review committee of registered hospitals and conducted by the Declaration of Helsinki.
Patient characteristics
For the NHANES, data on cancer and CVDs, the diagnoses, and their types were collected during in‐person interviews. Participants were recorded as having cancer if they answered ‘yes’ to the following question: ‘Have you ever been told by a doctor or other health professional that you had cancer or a malignancy of any kind?’ Individuals who responded ‘yes’ were defined as cancer survivors and were then asked, ‘What kind of cancer was it?’. Participants were recorded as having CVDs if they answered ‘yes’ to the following question: ‘Has a doctor or other health professional ever told you that you had congestive heart failure/coronary heart disease/angina pectoris/stroke?’. The BMI was recorded conventionally as weight (kg) /height2 (m2), and was categorized as low (<20 kg/m2 for the United States and <18.5 kg/m2 for the Chinese), normal (20–25 kg/m2 for the United States and 18.5 kg/m2 for the Chinese), overweight (25–30 kg/m2 for the United States and 24–28 kg/m2 for the Chinese), and obese (≥30 kg/m2 for the United States and ≥28 kg/m2 for the Chinese). Only the obese (≥30 kg/m2) US population was further categorized, due to the weight distribution of the participants, into obesity I (30–35 kg/m2) and obesity II (≥35 kg/m2). The neutrophil‐to‐lymphocyte ratio (NLR) is measured as the neutrophil count (×109)/lymphocyte count (×109). Systemic inflammation was defined as an NLR > 3. 16 Tumour stages were categorized according to the 8th edition of the American Joint Committee on Cancer TNM staging system. 17 , 18 All data were collected or measured on the first or second day of enrollment.
Follow‐up
Causes of death were obtained using the NHANES Public Use Linked Mortality Files (Table S1). The file contains mortality follow‐up data on NHANES participants obtained through the National Death Index (NDI) link from 31 December 2019. The underlying cause of death was coded using the UCOD LEADING variable and classified as all‐cause death. 19 The follow‐up data of INSCOC included the last follow‐up date and survival status of the participants.
Statistical analyses
All NHANES analyses were weighted, as recommended, to represent the US population. The R software (version 4.0.2, http://www.rproject.org) was used for the data analysis. Continuous variables were presented as mean (SD), and categorical variables were presented as frequency (weighted percentage or percentage). Comparisons between different BMI groups were performed using the chi‐square test for categorical variables and the Student t test for continuous variables. The hazard ratios (HR) and the 95% confidence intervals (CI) of the patients were estimated by modelling the risk factors using multivariable Cox regression models, weighted for the NHANES by the survey package, and for the other survival analysis by the survival package. The normal BMI group was used as a reference for all comparisons. The matched variables included age, sex, race, ethnicity, and NLR, for the NHANES; and sex, age, tumour type, tumour stage, surgery, chemotherapy, radiotherapy, and NLR for the INSCOC. Kaplan–Meier curves were used to perform a survival analysis of the combined analysis of BMI and NLR. Model‐based causal mediation analysis was used to identify mediation factors, by the mediation package.
Results
Characteristics of the study population
Overall, 52 270 participants (weighted population, 528 506 229) from the NHANES, and 17 418 patients with cancer from the INSCOC were included (Figure S1). The participants in NHANES cohort were young [mean age (SD): 47 (19.16) year], overweight [mean BMI (SD): 29 (6.83) kg/m2], had normal NLR [mean (SD): 2 (1.22)], and an equal sex distribution [man, n (weighted %): 25244 (44.9)]. A total of 5360 participants had CVDs and 4393 participants had cancer in NHANES (Table 1). The patients with cancer in INSCOC were older [mean (SD): 57.37 (11.66) year], had a normal BMI [mean (SD):22.61 (3.63) kg/m2], a high NLR [mean (SD):3.62 (3.63)], and an equal sex distribution [man, n (%):9486 (54.5)] (Table S2).
Table 1.
Sample size a and characteristics for association of men and women, NHANES 1999 to 2018
| Characteristic | Total | Low BMI (<20 kg/m2) | Normal BMI (20–25 kg/m2) | Overweight (25–30 kg/m2) | Obesity (>30 kg/m2) | P value |
|---|---|---|---|---|---|---|
| 52 270 | 2746 | 13 693 | 17 305 | 18 526 | ||
| Sex, n (weighted %) | <0.001 | |||||
| Men | 25 244 | 1052 (33.4) | 6629 (44.4) | 9561 (56.0) | 8002 (45.8) | |
| Women | 27 026 | 1694 (66.6) | 7064 (55.6) | 7744 (44.0) | 10 524 (54.2) | |
| Age, years, mean (SD) | 47 (19.16) | 39 (18.50) | 44 (18.16) | 50 (17.19) | 49 (16.19) | <0.001 |
| Age, n (weighted %) | <0.001 | |||||
| <25 | 7921 | 989 (29.3) | 3065 (17.8) | 1937 (9.2) | 1930 (8.6) | |
| 25–35 | 8296 | 485 (20.7) | 2421 (20.6) | 2591 (16.6) | 2799 (16.4) | |
| 35–45 | 8274 | 303 (14.4) | 1996 (17.7) | 2772 (19.0) | 3203 (19.7) | |
| 45–55 | 8042 | 274 (13.8) | 1720 (16.3) | 2794 (20.4) | 3254 (20.6) | |
| 55–65 | 7891 | 238 (9.1) | 1597 (12.2) | 2710 (15.2) | 3346 (17.8) | |
| ≥65 | 11 846 | 457 (12.7) | 2894 (15.5) | 4501 (19.6) | 3994 (16.9) | |
| BMI, kg/m2, mean (SD) | 29 (6.83) | 19 (1.05) | 23 (1.37) | 27 (1.42) | 36 (5.72) | <0.001 |
| NLR, mean (SD) | 2. (1.22) | 2.21 (1.22) | 2.22 (1.24) | 2.19 (1.14) | 2.20 (1.08) | 0.039 |
| Marital, n (weighted %) | <0.001 | |||||
| Divorced | 4950 | 214 (9.0) | 1120 (9.0) | 1594 (9.5) | 2022 (11.0) | |
| Living with partner | 3727 | 221 (9.5) | 927 (7.7) | 1285 (7.7) | 1294 (6.9) | |
| Married | 25 737 | 874 (41.7) | 6226 (52.1) | 9278 (59.6) | 9359 (57.0) | |
| Never married | 10 153 | 918 (32.1) | 3342 (23.4) | 2684 (14.8) | 3209 (16.3) | |
| Separated | 1623 | 65 (1.9) | 385 (2.5) | 497 (2.1) | 676 (2.8) | |
| Widowed | 4053 | 205 (5.9) | 944 (5.2) | 1425 (6.2) | 1479 (6.0) | |
| Race and ethnicity, n (weighted %) | <0.001 | |||||
| Mexican American | 9703 | 327 (4.6) | 2070 (6.4) | 3607 (9.2) | 3699 (9.7) | |
| Non‐Hispanic Black | 10 894 | 591 (10.1) | 2448 (8.8) | 3089 (9.3) | 4766 (14.0) | |
| Non‐Hispanic White | 22 616 | 1244 (69.6) | 6260 (70.2) | 7570 (68.6) | 7542 (66.1) | |
| Other Hispanic | 4280 | 159 (4.3) | 973 (4.9) | 1593 (6.5) | 1555 (5.6) | |
| Other race ‐ including multi‐racial | 4777 | 425 (11.3) | 1942 (9.7) | 1446 (6.4) | 964 (4.6) | |
| Family poverty income ratio, n (weighted %) | <0.001 | |||||
| <1.3 | 15 287 | 972 (28.6) | 3939 (21.6) | 4745 (19.9) | 5631 (22.8) | |
| 1.3 to 3.5 | 14 438 | 636 (35.5) | 4022 (44.5) | 5055 (45.2) | 4725 (39.3) | |
| ≥3.5 | 17 988 | 872 (35.9) | 4566 (33.9) | 5945 (34.9) | 6605 (38.0) | |
| Educational attainment, n (weighted %) | <0.001 | |||||
| < High school | 6009 | 191 (4.2) | 1346 (5.0) | 2347 (6.9) | 2125 (5.7) | |
| High school | 25 147 | 1323 (58.0) | 6873 (60.4) | 8255 (57.7) | 8696 (55.5) | |
| > High school | 21 055 | 1226 (37.7) | 5450 (34.4) | 6687 (35.3) | 7692 (38.7) | |
| Cardiovascular disease, n (weighted %) | <0.001 | |||||
| No | 43 308 | 2034 (93.7) | 11 144 (93.9) | 14 669 (91.3) | 15 461 (89.0) | |
| Yes | 5360 | 191 (6.3) | 1031 (6.1) | 1811 (8.7) | 2327 (11.0) | |
| Patients with cancer, n (weighted %) | 0.077 | |||||
| No | 47 877 | 2558 (92.0) | 12 614 (91.5) | 15 746 (90.6) | 16 959 (90.8) | |
| Yes | 4393 | 188 (8.0) | 1079 (8.5) | 1559 (9.4) | 1567 (9.2) | |
Weighted to be nationally representative. Weighted percentage may not sum to 100% because of missing data.
For continuous variables, values are mean [standard deviation (SD)] for categorical variables, values are expressed as n (weighted %). Differences in baseline characteristics were compared using χ 2 test for categorical variables and t Wilcoxon ranked sum test for continuous variables.
NLR, neutrophil‐to‐lymphocyte ratio.
Association of body mass index and survival in all participants of the National Health and Nutrition Examination Survey
Among the NHANES participants, a BMI between 25 and 34.5 had a protective effect on BMI (Figure 1). In the subgroup analysis of NHANES, the BMI showed a U‐shaped curve in most subgroups except for participants with cancer or CVD. Overweight showed a protective factor in all subgroups, except in young adults (age < 45 years). The obesity paradigm was observed in older participants [HR (95% CI): 0.84 (0.76, 0.93)] (Figure 2), participants with cancer [HR (95% CI): 0.84 (0.71, 0.99)], and participants with CVDs [HR (95% CI): 0.74 (0.65, 0.85)] (Figure 3). In young adults, a normal BMI had the best protective effect; for middle age, overweight had the best protective effect [HR (95% CI): 0.80 (0.68, 0.95)]; for old participants, high BMI had a similar protective effect [overweight, HR (95% CI): 0.82 (0.75, 0.89); obese I, HR (95% CI): 0.83 (0.75, 0.93); obese II, HR (95% CI): 0.85 (0.76, 0.96)] (Figure 4).
Figure 1.
Restricted spline curve (weighted) examining the association of BMI and all‐cause mortality in all participants. NHANES 1999 to 2018. Adjusted by age, sex, race and ethnicity, NLR. NLR, neutrophil‐to‐lymphocyte ratio.
Figure 2.
Association of body mass index (BMI) and all‐cause mortality in different age subgroup, NHANES 1999 to 2018. Weighted to be nationally representative, values are expressed as n (weighted %). Adjusted by age, sex, race and ethnicity, NLR. NLR, neutrophil‐to‐lymphocyte ratio.
Figure 3.
Association of body mass index (BMI) and all‐cause mortality in different subgroup (weighted), NHANES 1999 to 2018. Weighted to be nationally representative, values are expressed as n (weighted %). Adjusted by age, sex, race and ethnicity, NLR. CVD, cardiovascular disease; NLR, neutrophil‐to‐lymphocyte ratio.
Figure 4.
The obesity paradigm across the spectrum of age. (Left) distribution of body mass index (BMI) (weighted number). (Right) the hazard ratios (HR) for all‐cause mortality with 95% confidence intervals (CI) (weighted) according to the BMI strata are shown for age groups, NHANES 1999 to 2018. Adjusted by sex, race and ethnicity, NLR. NLR, neutrophil‐to‐lymphocyte ratio.
Association of body mass index and survival in participants with cancer of the National Health and Nutrition Examination Survey or Investigation on the Nutrition Status and Clinical Outcome of Common Cancers
In participants with cancer, the BMI between 25 to 35 had a protective effect on NHANES, and a higher BMI had a protective effect on INSCOC (Figure S2). For the NHANES participants with cancer, the obesity paradigm was observed only in older participants [HR (95% CI): 0.79 (0.65, 0.94)]. In the INSCOC cohort, the obesity paradigm was observed in all subgroups, and the protective effect of obesity increased with age [young, HR (95% CI): 0.61 (0.40, 0.94); middle age, HR (95% CI): 0.72 (0.61, 0.85); older, HR (95%CI): 0.65 (0.52, 0.81)] (Figure 5). In a further special subgroup analysis, it was found that the obesity paradigm in cancer patients was associated with age [young patients with cancer, HR (95% CI): 1.12 (0.77, 1.63); older participants without cancer, HR (95% CI): 0.87 (0.77, 0.97)] (Figure S3).
Figure 5.
(Left) association of body mass index (BMI) and all‐cause mortality in different age (cut into three) of patients with cancer, NHANES 1999 to 2018. Weighted to be nationally representative, values are expressed as n (weighted %). Adjusted by sex, race and ethnicity, and NLR. (Right) association of body mass index (BMI) and all‐cause mortality in different age subgroups. INSCOC 2013 to 2021. Adjusted by sex, tumour type, tumour stage, surgery, chemotherapy, radiotherapy, smoking, drinking and NLR. NLR, neutrophil‐to‐lymphocyte ratio.
Association of body mass index and survival in participants with cardiovascular disease of the National Health and Nutrition Examination Survey
For CVD NHANES participants, the BMI between 25 and 36 had a protective effect (Figure S4). However, in subgroups for different ages of CVD participants, only a trend of a protective effect was observed in older participants [HR (95% CI): 0.88 (0.75, 1.03)] (Figure S5).
Association of body mass index and survival in participants of different sexes of the National Health and Nutrition Examination Survey
Both men and women showed trends similar to the overall results, and the effect of obesity between different sexes was no difference [Figure S6 for all participants, women, HR (95% CI): 0.82 (0.71, 0.95); men, HR(95% CI): 0.86 (0.75, 0.99); Figure S7 for patients with cancer, women, HR (95% CI): 0.79 (0.61, 1.04); men, HR(95% CI): 0.73 (0.55, 0.97); Figure S8 for patients with CVD, women, HR(95% CI): 0.84 (0.66, 1.06); men, HR(95% CI): 0.92 (0.75, 1.13)].
The mediation of obesity and inflammation on survival
To identify mediation factors, we made a mediation analysis of participant of NHANES. A mediating effect of the NLR was observed (proportion of mediation = 15.4%). High NLR was a risk factor for survival. In the Kaplan–Meier curves, participants with a high NLR and a low BMI had the worst survival. In older participants, the subgroup with a normal BMI and high NLR showed similar or worse survival than that in the subgroup with obesity and a high NLR (Figure 6).
Figure 6.
(Left) Hazard ratios (HR) for all‐cause mortality with 95% confidence intervals (weighted) are shown for body mass index (BMI) in relation to low and high neutrophil‐to‐lymphocyte ratio (NLR). (Right) Kaplan–Maier curves (weighted) showing all‐cause mortality for prespecified BMI groups stratified by inflammation status. The P‐value of a log‐rank test for trend is shown for each plot, NHANES 1999 to 2018. HR spectrum adjusted by sex, race and ethnicity, NLR. NLR, neutrophil‐to‐lymphocyte ratio.
Discussion
This study, for the first time, observe the obesity paradox in the same population but with different characteristics and performed validation in another prospective cohort. Identifying the association factors of the obesity paradigm could help us better understand the association between obesity and survival. In the US population, the obesity paradigm has been observed in older participants and participants with diseases (CVDs and cancer). Further analysis showed that the protective effect of obesity in the population with diseases increased with age. The Chinese patients with cancer in the INSCOC study also showed a similar trend of an increasingly protective effect with age. In the China Stroke Primary Prevention Trial (CSPPT), the same trend of CVD participants from NHANES that obesity had an increasingly protective effect with age had been observed. 20 The Chinese population had a lower BMI than the US population. People classified as obese in the Chinese population were in the overweight range in the US population, and only a few Chinese participants were classified as obese II, which may the reason for the L‐shaped BMI‐related death curves of the Chinese population. But we still observed the same age‐related protective effect of obesity across different ethnic groups. Collectively, we propose that the obesity paradigm may association with the age‐related disease burden. The disease exerts a profound influence on the process of inflammation. In mediation analysis, the benefits of obesity are associated with inflammation. High inflammation and being underweight are serious adverse factors associated with patient survival. With increasing age, obese elderly people can better tolerate the negative effects of age and disease compared with their non‐obese elderly people.
Previous studies also found that the older population has a more pronounced obesity paradigm. A meta‐analysis conducted by Katherine M Flegal showed that relative to normal weight, both obesity (all grades) and grades 2 and 3 obesity were associated with significantly higher all‐cause mortality (95% CI: 1.12–1.25), grade 1 obesity overall was not associated with higher mortality (95% CI: 0.88–1.01), and overweight was associated with significantly lower all‐cause mortality (95% CI: 0.91–0.96). 21 To estimate BMI associations with mortality, incident type 2 diabetes, and coronary heart disease in older people, a group of researchers used a cohort of 955 000 older adults and observed that after excluding specific confounders, mortality nadirs were at a modestly higher BMI, and risk slopes at a higher BMI were less marked. 22 Among older patients who undergo high‐risk emergency general surgery (EGS), overweight and obese patients had a lower risk of death. 23 Compared with young age, older age and diseases (CVDs and tumours) were associated with multimorbidity, such as augmented CVD and fracture risk or worse inflammatory and metabolic status. 24 A Korean study of 9 278 433 participants with no CVD showed that the associations between BMI and outcomes were significantly modified by age, and were independent of sex, smoking, physical activity, and co‐morbidities. 25 Aging negatively affects body metabolism and increases aging‐related co‐morbid diseases, 6 but the high metabolic reserves in patients with a high BMI help against the nutritional insult of disease recovery or withstand this nutritional insult of the disease. 26 Obesity may have survival benefits by protecting against inefficient energy use and protein‐energy wasting. 27 Higher metabolic reserves result in better endurance to disease burden, which may be one of the reasons for the benefits in older people with high BMIs. The average age of most CVD and tumour cohorts is greater than 65 years (Prausmüller, et. CVD, 28 mean age: 70 years), and the burden of disease and age is the highest in this group of patients. Compared with the normal population and the disease population, we can see that obesity in middle‐aged cancer population has shown a protective trend. The protective effect of disease on obesity has a younger age.
A study of cancer patients showed that obesity was a protective factor only for older women who had experienced minimal weight loss (<5% weight) over 6 months. 29 In our study, we cannot completely exclude the effect of sex on prognosis related to BMI, although the obesity paradigm was not observed in different sexes according to the US population parameters because of the design of obesity. The cutoff point for BMI is not classified according to sex, which may explain why some studies showed different prognoses between men and women. Fat tissue is an independent factor that affects the prognosis of women with colorectal cancer (CRC), but its presence is of limited use in men with CRC. 30 Women have more fat tissue than men, and different sexes also have different cutoff points, which may be the reason for the difference in protective effects. These results suggest that protective factors for obesity directly relate to older age and diseases, with sex also making a difference. Sex might play a role in the association between obesity and survival.
Obesity and disease were association with systemic inflammation, which is a major risk factor for survival. 16 In our study, we found that systemic inflammation has a significant impact on patient survival and is a major mediator of the protective effect of obesity. The protective effect of obesity can lower the risk of death in people with elevated inflammation as obesity reduces the risk of inflammation in older patients. At the same time, participants could benefit from controlling systemic inflammation. The obesity paradigm was also observed in patients with advanced chronic kidney diseases, but the biased analysis from the authors of confounders indicated that inflammation did not fully explain the obesity paradigm. 31 The sex distribution of this study might explain the difference (female: 5%), which also showed that sex makes a difference in inflammation and obesity of survival.
Whether the obesity paradigm is called the high BMI paradigm has been raised by researchers. 32 In our study, and previous studies, overweight as defined by the BMI, was not associated with a worse survival or development of disease. 21 , 33 It was suggested that the definition of obesity may need to be revised because the fat content in the body is not equivalent. 34 A recent review proposed that during aging, adipose inflammation leads to fatty infiltration into skeletal muscles, resulting in decreased strength and functionality and older people with a normal BMI may experience sarcopenic obesity. 35 A high BMI may not indicate an increased amount of adipose tissue but may include a high skeletal muscle mass. 12 BMI is not considered as a good indicator of obesity recently. Even though it has been discredited as an unreliable measurement for nearly a decade, BMI is still the most convenient anthropometric index, and hence is still relevant for studies on body weight. Recent studies have begun to focus on the effect of different body compositions and their distribution. 4 , 11 Distributions of body composition measured by the waist‐to‐hip ratio, dual‐energy X‐ray absorptiometry (DXA), bioelectrical impedance analysis (BIA), ratios of muscle to fat mass and visceral to subcutaneous fat, and the function of cellular water distribution and grip strength may be better risk indicators than the BMI. 36 , 37 , 38
Treatment options for cancer patients can be affected by obesity, and the efficacy of some treatments can be diminished in obese patients. 10 Exercise training can prevent age‐associated decline in skeletal muscle mass. 39 Even with the obesity paradigm, losing weight and exercising are recommended for obese people. Exercising to gain muscle mass without losing weight is beneficial for improving survival. 40
Our study demonstrates that the obesity paradigm should be better called the high BMI paradox, and the BMI paradox is closely related to age. Relative to normal weight, obese in young people was association with higher all‐cause mortality, and obese in elderly people was not association with higher mortality. The protection of obesity is related to systemic inflammation. Physical exercise is recommended for obese individuals. Physical exercise to increase skeletal muscle mass can better protect people suffering from chronic diseases, even if the goal is not weight loss.
The major strength of this study is that we observed the obesity paradigm both in the normal and the disease populations in the same cohort. We used different cohorts and got the similar conclusions. We also observed that the average survival time of NHANES participants was better than that of patients with the disease, as its study design did not include patients with severe disease, which might have affected the study results, hence, we evaluated it in the Chinese tumour population which was also corroborated by previous studies in the disease population. A limitation of this study is that we did not have data on disease occurrence and hence could not observe the disease occurring during the follow‐up period. Future studies are warranted to investigate the disease occurrence and prognosis of obesity in the same cohort. However, notwithstanding the limitation, our study demonstrates that the high BMI paradox is associated with aging and the burden of disease. In chronic diseases and older age, patients with high BMI have better survival.
Funding
This work was supported by the National Key Research and Development Program (No. 2022YFC2009600) and the Beijing Municipal Science and Technology Commission Program (SCW2018‐06) to Dr. Han‐ping Shi.
Conflict of interest
The authors declare no conflicts of interest.
Supporting information
Table S1. Sample Sizea of patients with cancer and death causes, NHANES 1999 to 2018.
Table S2. Sample size and characteristics for different BMI, INSCOC 2013 to 2021.
Figure S1. (Left) Restricted spline curves (weighted) examining the association of BMI and all‐cause mortality in patients with cancer. NHANES 1999 to 2018. Adjusted by age, sex, race and ethnicity, NLR. (Right) Restricted spline curves examining the association of BMI and all‐cause mortality in cancer patients, INSCOC 2013 to 2021. Adjusted by sex, age, tumour type, tumour stage, surgery, chemotherapy, radiotherapy, smoking, drinking and NLR.
Figure S2. Sensitivity analysis of the association of BMI and all‐cause mortality in young patients with cancer and old participants without cancer, NHANES 1999 to 2018. Weighted to be nationally representative, values are expressed as n (weighted %). Adjusted by, sex, race and ethnicity, NLR.
Figure S3. Restricted spline curves (weighted) examining the association of BMI and all‐cause mortality in patients with CVD. NHANES 1999 to 2018. Adjusted by age, sex, race and ethnicity, NLR.
Figure S4. Association of BMI and all‐cause mortality in different age (cut into three) of patients with CVD, NHANES 1999 to 2018. Weighted to be nationally representative, values are expressed as n (weighted %). Adjusted by sex, race and ethnicity, and NLR.
Figure S5. Association of BMI and all‐cause mortality in different age (cut into three) of different sex participants, NHANES 1999 to 2018. Weighted to be nationally representative, values are expressed as n (weighted %). Adjusted by sex, race and ethnicity, and NLR.
Figure S6. Association of body mass index (BMI) and all‐cause mortality in different age (cut into three) of different sex patients with cancer, NHANES 1999 to 2018. Weighted to be nationally representative, values are expressed as n (weighted %). Adjusted by sex, race and ethnicity, and NLR.
Figure S7. Association of BMI and all‐cause mortality in different age (cut into three) of different sex patients with CVD, NHANES 1999 to 2018. Weighted to be nationally representative, values are expressed as n (weighted %). Adjusted by sex, race and ethnicity, and NLR.
Acknowledgements
The authors wish to express thanks to the participants and workers of the NHANES and INSCOC study for providing the research data.
Ge Y.‐Z., Liu T., Deng L., Zhang Q., Liu C.‐A., Ruan G.‐T., et al (2023) The age‐related obesity paradigm: results from two large prospective cohort studies, Journal of Cachexia, Sarcopenia and Muscle, doi: 10.1002/jcsm.13415
Yi‐Zhong Ge, Tong Liu and Li Deng contributed equally to this work and therefore share first authorship.
The INSCOC was approved by the medical ethics review committee of registered hospitals and conducted by the Declaration of Helsinki.
Trial registration: The Investigation on Nutrition Status and Clinical Outcome of Common Cancers (INSCOC) (Chinese Clinical Trial Registry: ChiCTR1800020329, URL of registration: http://www.chictr.org.cn/showprojen.aspx?proj=31813).
Data availability statement
The data described in the manuscript, code book, and analytic code will be made available upon request pending application and approval.
References
- 1. Mozaffarian D. Perspective: obesity‐an unexplained epidemic. Am J Clin Nutr 2022;115:1445–1450. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Avgerinos KI, Spyrou N, Mantzoros CS, Dalamaga M. Obesity and cancer risk: emerging biological mechanisms and perspectives. Metabolism 2019;92:121–135. [DOI] [PubMed] [Google Scholar]
- 3. Jayedi A, Soltani S, Motlagh SZ, Emadi A, Shahinfar H, Moosavi H, et al. Anthropometric and adiposity indicators and risk of type 2 diabetes: systematic review and dose‐response meta‐analysis of cohort studies. BMJ 2022;376:e067516. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Trestini I, Carbognin L, Bonaiuto C, Tortora G, Bria E. The obesity paradox in cancer: clinical insights and perspectives. Eat Weight Disord 2018;23:185–193. [DOI] [PubMed] [Google Scholar]
- 5. Karampela I, Chrysanthopoulou E, Christodoulatos GS, Dalamaga M. Is there an obesity paradox in critical illness? Epidemiologic and metabolic considerations. Curr Obes Rep 2020;9:231–244. [DOI] [PubMed] [Google Scholar]
- 6. Wang S, Ren J. Obesity paradox in aging: from prevalence to pathophysiology. Prog Cardiovasc Dis 2018;61:182–189. [DOI] [PubMed] [Google Scholar]
- 7. Zhou D, Li Z, Shi G, Zhou J. Obesity paradox for critically ill patients may be modified by age: a retrospective observational study from one large database. Crit Care 2020;24:425. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Laupland KB, Edwards F, Ramanan M, Shekar K, Tabah A. Reconciling the obesity paradox: obese patients suffer the highest critical illness associated mortality rates. J Crit Care 2021;66:75–77. [DOI] [PubMed] [Google Scholar]
- 9. Chapman IM. Obesity paradox during aging. Interdiscip Top Gerontol 2010;37:20–36. [DOI] [PubMed] [Google Scholar]
- 10. O'Connell F, O'Sullivan J. Help or hindrance: the obesity paradox in cancer treatment response. Cancer Lett 2021;522:269–280. [DOI] [PubMed] [Google Scholar]
- 11. Lee DH, Giovannucci EL. The obesity paradox in cancer: epidemiologic insights and perspectives. Curr Nutr Rep 2019;8:175–181. [DOI] [PubMed] [Google Scholar]
- 12. Cespedes Feliciano EM, Kroenke CH, Caan BJ. The obesity paradox in cancer: how important is muscle? Annu Rev Nutr 2018;38:357–379. [DOI] [PubMed] [Google Scholar]
- 13. Boukhris M, Aboyans V. The ‘obesity paradox’ in cardiovascular diseases: age matters! Eur J Prev Cardiol 2022;29:1476–1478. [DOI] [PubMed] [Google Scholar]
- 14. Ringel AE, Drijvers JM, Baker GJ, Catozzi A, García‐Cañaveras JC, Gassaway BM, et al. Obesity shapes metabolism in the tumor microenvironment to suppress anti‐tumor immunity. Cell 2020;183:1848–1866.e1826. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Hong Xia X, Song CH, Yin LY, Wang C, Zhen Ming F, Guo ZQ, et al. Extension protocol for the Investigation on Nutrition Status and Clinical Outcome of Patients with Common Cancers in China (INSCOC) study: 2021 update. Precis Nutr 2022;1:e000014. [Google Scholar]
- 16. Zhang Q, Song MM, Zhang X, Ding JS, Ruan GT, Zhang XW, et al. Association of systemic inflammation with survival in patients with cancer cachexia: results from a multicentre cohort study. J Cachexia Sarcopenia Muscle 2021;12:1466–1476. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Amin MB, Greene FL, Edge SB, Compton CC, Gershenwald JE, Brookland RK, et al. The eighth edition AJCC cancer staging manual: continuing to build a bridge from a population‐based to a more ‘personalized’ approach to cancer staging. CA Cancer J Clin 2017;67:93–99. [DOI] [PubMed] [Google Scholar]
- 18. Centers for Disease Control and Prevention (CDC) . National Center for Health Statistics (NCHS). National Health and Nutrition Examination Survey Questionnaire (or Examination Protocol, or Laboratory Protocol). Hyattsville, MD: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, [1999‐2018] [https://wwwn.cdc.gov/nchs/nhanes/default.aspx].
- 19. Continuous NHANES Public‐use Linked Mortality Files. Hyattsville, Maryland; 2019. [https://www.cdc.gov/nchs/data‐linkage/mortality‐public.htm]. [Google Scholar]
- 20. Yang W, Li JP, Zhang Y, Fan FF, Xu XP, Wang BY, et al. Association between body mass index and all‐cause mortality in hypertensive adults: results from the China Stroke Primary Prevention Trial (CSPPT). Nutrients 2016;8:384. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Flegal KM, Kit BK, Orpana H, Graubard BI. Association of all‐cause mortality with overweight and obesity using standard body mass index categories: a systematic review and meta‐analysis. JAMA 2013;309:71–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Bowman K, Delgado J, Henley WE, Masoli JA, Kos K, Brayne C, et al. Obesity in older people with and without conditions associated with weight loss: follow‐up of 955,000 primary care patients. J Gerontol A Biol Sci Med Sci 2017;72:203–209. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. El Moheb M, Jia Z, Qin H, El Hechi MW, Nordestgaard AT, Lee JM, et al. The obesity paradox in elderly patients undergoing emergency surgery: a nationwide analysis. J Surg Res 2021;265:195–203. [DOI] [PubMed] [Google Scholar]
- 24. Perna S, Spadaccini D, Nichetti M, Avanzato I, Faliva MA, Rondanelli M. Osteosarcopenic visceral obesity and osteosarcopenic subcutaneous obesity, two new phenotypes of sarcopenia: prevalence, metabolic profile, and risk factors. J Aging Res 2018;2018:6147426. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Lee HJ, Kim HK, Han KD, Lee KN, Park JB, Lee H, et al. Age‐dependent associations of body mass index with myocardial infarction, heart failure, and mortality in over 9 million Koreans. Eur J Prev Cardiol 2022;29:1479–1488. [DOI] [PubMed] [Google Scholar]
- 26. Liu Z, Sanossian N, Starkman S, Avila‐Rinek G, Eckstein M, Sharma LK, et al. Adiposity and outcome after ischemic stroke: obesity paradox for mortality and obesity parabola for favorable functional outcomes. Stroke 2021;52:144–151. [DOI] [PubMed] [Google Scholar]
- 27. Naderi N, Kleine CE, Park C, Hsiung JT, Soohoo M, Tantisattamo E, et al. Obesity paradox in advanced kidney disease: from bedside to the bench. Prog Cardiovasc Dis 2018;61:168–181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Prausmüller S, Heitzinger G, Pavo N, Spinka G, Goliasch G, Arfsten H, et al. Malnutrition outweighs the effect of the obesity paradox. J Cachexia Sarcopenia Muscle 2022;13:1477–1486. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Martinez‐Tapia C, Diot T, Oubaya N, Paillaud E, Poisson J, Gisselbrecht M, et al. The obesity paradox for mid‐ and long‐term mortality in older cancer patients: a prospective multicenter cohort study. Am J Clin Nutr 2020;113:129–141. [DOI] [PubMed] [Google Scholar]
- 30. Xie H, Ruan G, Zhang H, Ge Y, Zhang Q, Lin S, et al. Sex differences in the association of triceps skinfold thickness and colorectal cancer mortality. Nutrition 2022;103‐104:111811. [DOI] [PubMed] [Google Scholar]
- 31. Soohoo M, Streja E, Hsiung JT, Kovesdy CP, Kalantar‐Zadeh K, Arah OA. Cohort study and bias analysis of the obesity paradox across stages of chronic kidney disease. J Ren Nutr 2021;32:529–536. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Donini LM, Pinto A, Giusti AM, Lenzi A, Poggiogalle E. Obesity or BMI paradox? Beneath the tip of the iceberg. Front Nutr 2020;7:53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Smith LK, Arabi S, Lelliott EJ, McArthur GA, Sheppard KE. Obesity and the impact on cutaneous melanoma: friend or foe? Cancers (Basel) 2020;12:1583. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Amundson DE, Djurkovic S, Matwiyoff GN. The obesity paradox. Crit Care Clin 2010;26:583–596. [DOI] [PubMed] [Google Scholar]
- 35. Li CW, Yu K, Shyh‐Chang N, Jiang Z, Liu T, Ma S, et al. Pathogenesis of sarcopenia and the relationship with fat mass: descriptive review. J Cachexia Sarcopenia Muscle 2022;13:781–794. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Chrysant SG, Chrysant GS. The single use of body mass index for the obesity paradox is misleading and should be used in conjunction with other obesity indices. Postgrad Med 2019;131:96–102. [DOI] [PubMed] [Google Scholar]
- 37. Sanchez A, Furberg H. Understanding the obesity paradox in cancer: looking to body composition and tumor‐specific biology to understand this complex association. Am J Clin Nutr 2020;113:1–2. [DOI] [PubMed] [Google Scholar]
- 38. Ge YZ, Ruan GT, Zhang KP, Tang M, Zhang Q, Zhang X, et al. Which anthropometric measurement is better for predicting survival of patients with cancer cachexia? Br J Nutr 2021;127:1–9. [DOI] [PubMed] [Google Scholar]
- 39. Ringholm S, Gudiksen A, Halling JF, Qoqaj A, Rasmussen PM, Prats C, et al. Impact of aging and lifelong exercise training on mitochondrial function and network connectivity in human skeletal muscle. J Gerontol A Biol Sci Med Sci 2022;78:373–383. [DOI] [PubMed] [Google Scholar]
- 40. Kunutsor SK, Whitehouse MR, Blom AW. Obesity paradox in joint replacement for osteoarthritis ‐ truth or paradox? Geroscience 2022;44:651–659. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1. Sample Sizea of patients with cancer and death causes, NHANES 1999 to 2018.
Table S2. Sample size and characteristics for different BMI, INSCOC 2013 to 2021.
Figure S1. (Left) Restricted spline curves (weighted) examining the association of BMI and all‐cause mortality in patients with cancer. NHANES 1999 to 2018. Adjusted by age, sex, race and ethnicity, NLR. (Right) Restricted spline curves examining the association of BMI and all‐cause mortality in cancer patients, INSCOC 2013 to 2021. Adjusted by sex, age, tumour type, tumour stage, surgery, chemotherapy, radiotherapy, smoking, drinking and NLR.
Figure S2. Sensitivity analysis of the association of BMI and all‐cause mortality in young patients with cancer and old participants without cancer, NHANES 1999 to 2018. Weighted to be nationally representative, values are expressed as n (weighted %). Adjusted by, sex, race and ethnicity, NLR.
Figure S3. Restricted spline curves (weighted) examining the association of BMI and all‐cause mortality in patients with CVD. NHANES 1999 to 2018. Adjusted by age, sex, race and ethnicity, NLR.
Figure S4. Association of BMI and all‐cause mortality in different age (cut into three) of patients with CVD, NHANES 1999 to 2018. Weighted to be nationally representative, values are expressed as n (weighted %). Adjusted by sex, race and ethnicity, and NLR.
Figure S5. Association of BMI and all‐cause mortality in different age (cut into three) of different sex participants, NHANES 1999 to 2018. Weighted to be nationally representative, values are expressed as n (weighted %). Adjusted by sex, race and ethnicity, and NLR.
Figure S6. Association of body mass index (BMI) and all‐cause mortality in different age (cut into three) of different sex patients with cancer, NHANES 1999 to 2018. Weighted to be nationally representative, values are expressed as n (weighted %). Adjusted by sex, race and ethnicity, and NLR.
Figure S7. Association of BMI and all‐cause mortality in different age (cut into three) of different sex patients with CVD, NHANES 1999 to 2018. Weighted to be nationally representative, values are expressed as n (weighted %). Adjusted by sex, race and ethnicity, and NLR.
Data Availability Statement
The data described in the manuscript, code book, and analytic code will be made available upon request pending application and approval.