Abstract
Aim. To study the association between anthropometric measures: body mass index (BMI), percent body fat, waist circumference (WC), waist-to-hip ratio (WHR), waist-to-height ratio (WHtR), waist-to-hip-to-height ratio (WHHR), and A Body Shape Index (ABSI); to see if individuals in the lowest 5 percentiles for these measures have an increased risk of mortality.
Methods. A population-based prospective cohort study (10,304 men and 16,549 women), the Malmö Diet and Cancer study (MDC), aged 45–73 years.
Results. During a mean follow-up of 14 ± 3 years, 2,224 men and 1,983 women died. There was a significant increased mortality risk after adjustments for potential confounders in the group with the 5% lowest BMI (referent 25%–75%); hazard ratios (HR) with 95% confidence intervals were 1.33 (1.10–1.61) for women and 1.27 (1.07–1.52) for men. A similar significant increased mortality risk was seen with the 5% lowest percent body fat, HR 1.31 (1.07–1.60) for women and 1.25 (1.04–1.50) for men. Women with an ABSI in the lowest 5 percentiles had a lower mortality risk HR 0.64 (0.48–0.85).
Conclusion. These results imply that BMI or percent body fat could be used to identify lean individuals at increased mortality risk.
All measures identified increased risks in individuals with values above the 75th percentile.
BMI as well as percent body fat showed had a U-shaped relationship with mortality.
There was no increased mortality risk associated with having a WC, WHR, WHtR or WHHR in the lowest 5 percentiles when using individuals with the 25th–75th percentiles of each measure as referent.
Introduction
For centuries, and in many parts of the world today, overweight has been considered high status and a sign of health, which may be explained by a higher survival in infectious diseases such as tuberculosis (Citation1). Today, the trend in the United States is alarming, with more than one-third of the US population being classified as obese (Citation2). A similar trend with increased central obesity has been observed in Europe (Citation3).
Health research has mainly focused on the increased risks associated with overweight and obesity (Citation4–9). Traditionally, body mass index (BMI) has been the most common anthropometric measurement for deciding and grading under- and overweight (Citation4). There are, however, a number of anthropometric measurements, namely: percent body fat, waist circumference (WC), waist-to-hip ratio (WHR), sagittal abdominal diameter (SAD), waist-to-height ratio (WHtR), waist-to-hip-to-height ratio (WHHR), and A Body Shape Index (ABSI) (Citation5,Citation6,Citation10–12). A low BMI has been associated with increased risk of mortality (Citation8). Less is known about other anthropometrical measures, and whether a clinically relevant proportion of the population, such as 5% or more, has an increased mortality risk. Novel anthropometrical measures, especially WHHR and ABSI (Citation5,Citation6,Citation12), have recently been shown to be superior to BMI in predicting increased risks of cardiovascular disease and mortality. Furthermore, WHHR can identify the increased risk of cardiovascular disease among men classified as overweight and obese by their BMI, and in women with normal BMI (Citation5).
We hypothesize that other anthropometric measures could be superior to BMI, also when it comes to identifying an increased risk of mortality in the leanest individuals in the population. Herein, we primarily aimed to investigate whether anthropometric measures can identify increased risk of mortality among those in the lowest 5 percentiles of each measure in a cohort study of 26,853 participants in Sweden during a mean follow-up of 14 ± 3 years. As a secondary aim, we wanted to study the risk associated with being above the 75th percentile of each measure to verify if they identify the risk associated with overweight and obesity.
Materials/subjects and methods
Study population
We used the Malmö Diet and Cancer (MDC) cohort, from the city of Malmö in southern Sweden, for the present study. Detailed information on MDC regarding sample characteristics and data collection has been given previously (Citation13,Citation14). During March 1991 and September 1996, 28,449 individuals (men: n = 11,246, born 1923–1945; and women: n = 17,203, born 1923–1950) attended a baseline examination, underwent measurement of blood pressure and anthropometric measures, and filled out a self-administered questionnaire. A total of 1,596 subjects were excluded due to missing values of anthropometric measurements and other biological, life-style, and socioeconomic variables. Thus, the final study population in the analysis consisted of 26,853 (10,304 men and 16,549 women) subjects, aged 45–73 years.
The ethics committee at Lund University approved the study (LU 51/90), and all participants provided informed consent.
Baseline examinations
The examinations were performed by trained nurses at the screening center. Standing height was measured with a fixed stadiometer calibrated in centimeters. Weight was measured to the nearest 0.1 kg using balance-beam scales, with subjects wearing light clothing and no shoes. BMI was calculated as weight (kg) divided by the square of the height (m2). WC was measured as the circumference (cm) between the lowest rib margin and iliac crest, and hip circumference (cm) as the largest circumference between waist and thighs. WHR was defined as the ratio of circumference of waist to hip. WHHR was calculated as the WHR divided by height (m-1). WHtR was defined as the ratio of circumference of waist to height. Bioelectrical impedance analyzers (BIA) were used for estimating body composition, and percent body fat was calculated using an algorithm according to procedures provided by the manufacturer (BIA 103, single-frequency analyzer, JRL Systems, Detroit, IL, USA) (Citation15). A Body Shape Index (ABSI) (m11/6 kg-2/3), WC/(BMI(2/3)height(1/2)) (Citation16–18). BMI, WC, WHtR, WHR, WHHR, percent body fat, and ABSI were categorized into the following percentiles: lowest 5%, 5%–10%, 10%–25%, 25%–75%, and highest 25% in men and women, respectively.
Information on smoking habits, alcohol consumption, leisure-time physical activity, education level, and civil and immigrant status was obtained from a self-administered questionnaire. Comorbidities at baseline (history of cancer, diabetes, coronary events, and stroke) were retrieved from the local and national registers (Citation19,Citation20). Low level of leisure-time physical activity was defined as the lowest category of a score revealed through 18 questions covering a range of activities in the four seasons (Citation21). Subjects were divided into current smokers (i.e. those who smoked regularly or occasionally) or non-smokers (i.e. former smokers or never smokers). High alcohol consumption was defined as > 40 g alcohol per day for men and > 30 g per day for women. Education was categorized into three groups: school year < 9, 9–12, and > 12, respectively. Marital status was categorized into married or not (Citation22). Immigrant status was divided into Swedish-born or foreign-born.
All subjects were followed from the baseline examination until death, emigration from Sweden, or 30 June 2009, whichever came first. Vital status and causes of death were retrieved through record linkage of the personal identification number and the Swedish Cause of Death Register (SCDR). Cardiovascular mortality was defined on the basis of ICD-9 codes 390–459 or ICD-10 codes I00–I99 in the SCDR. Cancer mortality was based on ICD-9 codes 140–239 or ICD-10 codes D00–D48.
Statistical analysis
Cox proportional hazards regression was used to examine the association between anthropometric measures and mortality in men and women, using the 25%–75% group as referent. Proportional hazards were confirmed with Schoenfeld's test. The follow-up time until death, emigration, or end of follow-up was used. Hazard ratios (HR), with 95% confidence interval (CI) were calculated. Age was included as covariate in the basic model. Multivariable models were adjusted for current smoking, high alcohol consumption, low leisure physical activity, low education, marital and immigrant status, and comorbidities at baseline. Potential multiplicative interactions with age (below/above 50 at baseline) were tested. All analyses were performed using IBM SPSS Statistics (version 22; www.spss.com).
Results
Baseline characteristics for the participants are shown in , in the following BMI strata: lowest 5%, > 5%–10%, > 10%–25%, > 25%–75%, > 75%.
Table I. Baseline characteristics.
Anthropometric measures and total mortality
During a mean follow-up of 14 ± 3 years, 4,207 participants (2,224 men and 1,983 women) died.
The age-adjusted total mortality risk associated with low and high anthropometric values are shown in . The participants in the 25th to the 75th percentile of each measure were used as referents. All anthropometric values in the 75th percentile had higher risk in mortality. There was an increased risk of mortality for women in the lowest 5 percentiles of BMI (P < 0.001) and percent body fat (P < 0.001). Similar and significant findings were seen among the men: the lowest 5 percentiles of BMI (P < 0.001), percent body fat (P < 0.01), and WC (P < 0.05) had higher mortality risk. The mortality risk estimate was higher for both women and men with low BMI (lowest 5%) than for those with a high BMI (above the 75th percentile). The mortality risk estimate was also higher in women with the lowest 5% percent body fat. Women and men in the lowest 5 percentiles of ABSI had lower mortality risks than in the 25th–75th percentiles.
Table II. The age-adjusted total mortality risk associated with anthropometric measures.
For the women and men with values above the 75th percentile of each measure, the other anthropometric measures resembled the results for BMI, i.e. significantly higher risks were seen (P < 0.001).
The results of the multivariable models are shown in (adjusted for age, education level, marital status, immigrant status, smoking status, alcohol intake, physical activity level, and baseline comorbidities). The results regarding BMI and percent body fat were attenuated but still significant for women and men. The lower risk for women in the lowest 5th percentiles of ABSI than in those with an ABSI in the 25th–75th percentiles remained significant. There were no significant multiplicative interactions with age (above or below 50 years) for any of the measures.
Table III. Multivariablea models of the association between anthropometric measures and total mortality.
We performed a sensitivity analysis where all individuals who reported to be current smokers at the baseline investigation were excluded (data not shown in tables). The increased risk for mortality remained significant for men, HR 1.33 (1.01–1.76, P = 0.049), but not for women, HR 1.32 (0.98–1.81, P = 0.069), in the lowest 5% of BMI group. For the lowest 5 percentiles of percent body fat, the risk was significantly increased for women, HR 1.36 (1.01–1.83, P = 0.044), but non-significantly in men, HR 1.11 (0.86–1.44, P = 0.421).
Anthropometric measures and cause-specific mortality
The association between BMI and percent body fat, and cardiovascular, cancer, and other fatalities are shown separately in and , in men and women, respectively. For men, the lowest 5 percentiles for BMI and percent body fat had an increased risk of cancer in the age-adjusted model. Only percent body fat remained significant after further adjustments. There was an increased risk of cancer mortality in women in the lowest 5 percentiles of percent body fat, which was attenuated and no longer significant after further adjustments. The lowest 5 percentiles in percent body fat had an increased risk of other mortalities (non-cancer and non-cardiovascular disease (CVD)) that remained after adjustments in women and men.
Table IV. Anthropometric measures BMI, body fat% and cancer, CVD, and other mortality in women.
Table V. Anthropometric measures BMI, body fat% and cancer, CVD, and other mortality in men.
Additionally, we analyzed some specific causes of death (non-cancer and non-CVD mortality). In the multivariable model, the lowest 5 percentiles of BMI were associated with a high risk of infectious disease mortality (n = 21, HR 6.48 (2.06–20.40), P < 0.001) as well as respiratory disease mortality (n = 109, HR 3.58 (1.94–6.62), P < 0.001) among men. Women with a BMI and percent body fat in the lowest 5 percentiles had a higher risk of respiratory disease mortality: HR 4.18 (2.49–7.00), P < 0.001, and HR 3.81 (2.06–7.04), P < 0.001 (n = 104), respectively.
Discussion
In accordance with previous studies, the associations between different anthropometric measures and an increased risk of mortality differ in shape (Citation12). BMI as well as percent body fat had a U-shaped relationship. In contrast, a lower risk was seen for the women in the lowest 5 percentiles of ABSI, i.e. a positive linear relationship with mortality. There was no increased mortality risk associated with having a WC, WHR, WHtR, or WHHR in the lowest 5 percentiles when using individuals with the 25th– 75th percentiles of each measure as referent. All measures identified increased risks in individuals with values above the 75th percentile.
There have been several studies and also meta-analyses proving the association between very low BMI and increased risk of mortality, though using BMI as the only measurement (Citation23). The large number of participants and the long follow-up of the present study enabled us to study the lowest 5 percentiles of each anthropometric measure. Similar studies have been done, but only with data obtained for the lowest 10th percentile (Citation12). Having data for the lowest 5 percentiles allowed us to get an even more tuned relationship and to show that the lowest 6%–10% was not as saliently associated with mortality. Even though our study did not identify an increased mortality in the lowest 5 percentiles of most anthropometric measures, we found an association between low BMI and mortality among women as well as among men, confirming previous findings (Citation8,Citation23–27). Our study supported the findings of Chen et al. (Citation24) and Heitmann et al. (Citation27) showing a U-shaped association between BMI and mortality risk. Likewise, this study confirmed previous findings concerning low percent body fat and increased mortality risk (Citation25).
The present study further supports previous findings of a positive linear relationship between ABSI and risk of mortality (Citation12,Citation16–18): both women and men in the lowest 5 percentiles of ABSI had lower mortality risks than in the 25th–75th percentiles. One potential explanation for this could be the fact that ABSI is the only anthropometric measure taking into account three different parameters (weight, height, and WC), and a high ABSI implies that the WC is higher than expected for a given height and weight.
In accordance with previous studies (Citation6–8,Citation28), all anthropometric measures identified an increased mortality risk in overweight and obese individuals. The present study supports the use of anthropometric measures in clinical practice in order to assess obesity-related mortality in older adults (Citation12,Citation29).
There are several potential explanations to a higher risk in lean individuals. Individuals with low BMI are more often smokers, and had a higher risk of respiratory disease mortality in the present study. The nicotine in cigarettes has a negative effect on appetite and other feelings of hunger (Citation30), thus permitting a regular smoker to go without food for longer periods of time than non-smokers. According to this relationship the smoking would assist the person to lose weight and to stay underweight. Being a smoker imposes a higher risk for mortality, which has been shown repeatedly (Citation31–33). However, when the findings in this study were adjusted for smoking, the results remained significant, indicating that smoking cannot explain the increased risk with a low BMI and percent body fat.
Underweight individuals might not receive the same kind of attention once they come in contact with medical care. Since the hazardous consequences of overweight and obesity have been known for a long time (Citation6–8,Citation28), there would be less doubt about what to do with an overweight patient—besides taking care of the acute symptoms for which the patient had come into contact with medical care, weight loss would very often be recommended in order to prevent new symptoms. For the underweight patient the link between the symptoms and the weight may perhaps not be as obvious, since less is known about the connection between underweight and an increased risk of mortality (Citation34,Citation35).
Another possible explanation could be that underweight people are more frail than normal-weight or overweight people. This frailty would make these people more vulnerable to serious consequences once affected by a disease (Citation36); the limited amount of fat in the body would imply that there would be less of stored reserve energy to be consumed in case of a serious illness when extra energy is needed in order to combat the disease, and also to compensate for loss in appetite. The higher mortality may also be a sign of malnutrition, which may result in a suppressed immune system (Citation37) besides also smaller amounts of adipose tissue and thus fewer adipocytes. Adipocytes are associated with the dendritic cells in the lymph nodes, which are activated during inflammation or infection (Citation38). These hypotheses are strengthened by the increase in mortality in infectious diseases as well as respiratory diseases in the lowest 5 percentiles of BMI in the present study.
In contrast to our initial hypothesis, this study identified the well-established as well as novel anthropometric measures to be lacking in their potential to predict an increased risk of mortality for underweight people. New additional methods would be desirable. BMI is a fairly reliable measurement, though it is still not optimal in determining the actual fat content of the body (Citation39). The method used to quantify the percent body fat in the present study could be criticized for being too vague; if the percent body fat was to be measured in more detail, a more accurate method would have to be found. In previous studies it has been shown that detailed fat analysis of different organs such as the liver, pancreas, and some muscles can be obtained through MRI scans (Citation40). Such a precise fat analysis would be valuable in order to decide whether the overweight/underweight is benign or malign. MRI is a recommended method in order to obtain a thorough and organ-specific fat analysis that is objective and detailed as well as without any harmful risks (Citation41). Maybe MRI scans could be a more reliable method in order to identify the risk groups when it comes to underweight as well as overweight, than the studied anthropometric measures. Further research highlighting the use of MRI in identifying underweight and overweight is warranted.
Limitations and strengths
The present study has several strengths. The cohort was fairly big with a high rate of participation. Moreover, the participants in this study were thoroughly examined, characterized by their BMI, percent body fat, WC, WHR, WHtR, WHHR, and ABSI, with enough participants (n = 26,853) and a mean follow-up of 14 years, which allowed analysis of the participants in the lowest 5 percentiles of each measure. This is to be compared to similar studies where participants with the lowest BMI have been identified by the 10% lowest BMI (Citation12).
The cohort was a Swedish one with all participants being residents of Sweden; it is unclear to what extent the results can be extended to other populations. Different ethnicities respond differently to obesity (Citation42), in fact so differently that it has been proposed that Asians ought to have lower BMI cut-offs than Caucasians and Afro-Americans (Citation43). Yet, approximately 15% of the present cohort was made up of participants born in non- Scandinavian countries (Citation44). Another limitation of this study is the fact that the cohort did not include participants younger than the age of 45 at the baseline investigation, limiting the extrapolation to other age groups. There could also be residual confounding, i.e. factors that we did not have the possibility to adjust for, such as anxiety and depression. Finally, although we assessed several anthropometric measures, there are some important anthropometric measures that were not available in the present cohort, e.g. SAD, which has been shown to predict coronary heart disease (CHD) risk in other Swedish populations (Citation6,Citation45).
Conclusion
This present study highlights the shortcomings of many anthropometric measures including WC, WHR, WHtR, WHHR, and ABSI when it comes to identifying the increased risk of mortality among underweight individuals. BMI or percent body fat should be used to identify lean individuals at increased mortality risk. Our results, however, support earlier findings with increased mortality risk in overweight individuals as identified by general and abdominal anthropometric measures.
Funding: The Malmö Diet and Cancer study was supported by grants from the Swedish Medical Research Council, the Swedish Heart and Lung Foundation, and by funds from the Region Skåne, Skåne University Hospital, Malmö, and Lundströms Foundation.
Declaration of interest: The study was investigator-initiated and -driven. The authors report no conflicts of interests in connection with this study. I.W. is employed by Executive Health AB.
References
- Flegal KM, Graubard BI, Williamson DF, Gail MH. Cause-specific excess deaths associated with underweight, overweight, and obesity. JAMA. 2007;298:2028–37.
- Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of childhood and adult obesity in the United States, 2011-2012. JAMA. 2014;311: 806–14.
- Rosengren A, Eriksson H, Hansson PO, Svardsudd K, Wilhelmsen L, Johansson S, et al. Obesity and trends in cardiovascular risk factors over 40 years in Swedish men aged 50. J Intern Med. 2009;266:268–76.
- Cabal-Hierro L, Lazo PS. Signal transduction by tumor necrosis factor receptors. Cell Signal. 2012;24:1297–305.
- Carlsson AC, Riserus U, Arnlov J, Borne Y, Leander K, Gigante B, et al. Prediction of cardiovascular disease by abdominal obesity measures is dependent on body weight and sex. Results from two community based cohort studies. Nutr Metab Cardiovasc Dis. 2014;24:891–9.
- Carlsson AC, Riserus U, Engstrom G, Arnlov J, Melander O, Leander K, et al. Novel and established anthropometric measures and the prediction of incident cardiovascular disease: a cohort study. Int J Obes (Lond). 2013;37:1579–85.
- Pischon T, Boeing H, Hoffmann K, Bergmann M, Schulze MB, Overvad K, et al. General and abdominal adiposity and risk of death in Europe. N Engl J Med. 2008;359:2105–20.
- Wandell PE, Carlsson AC, Theobald H. The association between BMI value and long-term mortality. Int J Obes (Lond). 2009;33: 577–82.
- Li C, Engstrom G, Hedblad B, Calling S, Berglund G, Janzon L. Sex differences in the relationships between BMI, WHR and incidence of cardiovascular disease: a population-based cohort study. Int J Obes (Lond). 2006;30:1775–81.
- Sierra-Johnson J, Unden AL, Linestrand M, Rosell M, Sjogren P, Kolak M, et al. Eating meals irregularly: a novel environmental risk factor for the metabolic syndrome. Obesity (Silver Spring). 2008;16:1302–7.
- Song X, Jousilahti P, Stehouwer CD, Soderberg S, Onat A, Laatikainen T, et al. Comparison of various surrogate obesity indicators as predictors of cardiovascular mortality in four European populations. Eur J Clin Nutr. 2013;67:1298–302.
- Song X, Jousilahti P, Stehouwer CD, Soderberg S, Onat A, Laatikainen T, et al. Cardiovascular and all-cause mortality in relation to various anthropometric measures of obesity in Europeans. Nutr Metab Cardiovasc Dis. 2015;25:295–304
- Berglund G, Elmstahl S, Janzon L, Larsson SA. The Malmo Diet and Cancer Study. Design and feasibility. J Intern Med. 1993;233:45–51.
- Melander O, Newton-Cheh C, Almgren P, Hedblad B, Berglund G, Engstrom G, et al. Novel and conventional biomarkers for prediction of incident cardiovascular events in the community. JAMA. 2009;302: 49–57.
- Calling S, Hedblad B, Engstrom G, Berglund G, Janzon L. Effects of body fatness and physical activity on cardiovascular risk: risk prediction using the bioelectrical impedance method. Scand J Public Health. 2006;34:568–75.
- Kahan T, Wandell P. Risk factors in established coronary heart disease: evaluation of a secondary prevention programme. J Cardiovasc Risk. 2001;8:73–80.
- Krakauer NY, Krakauer JC. A new body shape index predicts mortality hazard independently of body mass index. PLoS One. 2012;7:e39504.
- Krakauer NY, Krakauer JC. Dynamic association of mortality hazard with body shape. PLoS One. 2014;9:e88793.
- Ludvigsson JF, Andersson E, Ekbom A, Feychting M, Kim JL, Reuterwall C, et al. External review and validation of the Swedish national inpatient register. BMC Public Health. 2011;11:450.
- Zia E, Hedblad B, Pessah-Rasmussen H, Berglund G, Janzon L, Engstrom G. Blood pressure in relation to the incidence of cerebral infarction and intracerebral hemorrhage. Hypertensive hemorrhage: debated nomenclature is still relevant. Stroke. 2007;38:2681–5.
- Li C, Aronsson CA, Hedblad B, Gullberg B, Wirfalt E, Berglund G. Ability of physical activity measurements to assess health-related risks. Eur J Clin Nutr. 2009;63:1448–51.
- Rosvall M, Ostergren PO, Hedblad B, Isacsson SO, Janzon L, Berglund G. Occupational status, educational level, and the prevalence of carotid atherosclerosis in a general population sample of middle-aged Swedish men and women: results from the Malmo Diet and Cancer Study. Am J Epidemiol. 2000;152:334–46.
- Winter JE, MacInnis RJ, Wattanapenpaiboon N, Nowson CA. BMI and all-cause mortality in older adults: a meta-analysis. Am J Clin Nutr. 2014;99:875–90.
- Chen Y, Copeland WK, Vedanthan R, Grant E, Lee JE, Gu D, et al. Association between body mass index and cardiovascular disease mortality in east Asians and south Asians: pooled analysis of prospective data from the Asia Cohort Consortium. BMJ. 2013;347:f5446.
- Rolland Y, Gallini A, Cristini C, Schott AM, Blain H, Beauchet O, et al. Body-composition predictors of mortality in women aged >/ = 75 y: data from a large population-based cohort study with a 17-y follow-up. Am J Clin Nutr. 2014;100:1352–60.
- Thorogood M, Appleby PN, Key TJ, Mann J. Relation between body mass index and mortality in an unusually slim cohort. J Epidemiol Community Health. 2003;57:130–3.
- Heitmann BL, Erikson H, Ellsinger BM, Mikkelsen KL, Larsson B. Mortality associated with body fat, fat-free mass and body mass index among 60-year-old swedish men-a 22-year follow-up. The study of men born in 1913.Int J Obes Relat Metab Disord. 2000;24:33–7.
- 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.
- Carmienke S, Freitag MH, Pischon T, Schlattmann P, Fankhaenel T, Goebel H, et al. General and abdominal obesity parameters and their combination in relation to mortality: a systematic review and meta-regression analysis. Eur J Clin Nutr. 2013;67:573–85.
- Jo YH, Talmage DA, Role LW. Nicotinic receptor-mediated effects on appetite and food intake. J Neurobiol. 2002;53:618–32.
- Ezzati M, Lopez AD. Estimates of global mortality attributable to smoking in 2000. Lancet. 2003;362:847–52.
- Jacobs DR Jr, Adachi H, Mulder I, Kromhout D, Menotti A, Nissinen A, et al. Cigarette smoking and mortality risk: twenty-five-year follow-up of the Seven Countries Study. Arch Intern Med. 1999;159:733–40.
- Mokdad AH, Marks JS, Stroup DF, Gerberding JL. Actual causes of death in the United States, 2000. JAMA. 2004;291:1238–45.
- Berrington de Gonzalez A, Hartge P, Cerhan JR, Flint AJ, Hannan L, MacInnis RJ, et al. Body-mass index and mortality among 1.46 million white adults. N Engl J Med. 2010;363:2211–19.
- Troiano RP, Frongillo EA Jr, Sobal J, Levitsky DA. The relationship between body weight and mortality: a quantitative analysis of combined information from existing studies. Int J Obes Relat Metab Disord. 1996;20:63–75.
- Wandell PE, Theobald H. The association between low fasting blood glucose value and mortality. Curr Diabetes Rev. 2007;3:274–9.
- Rytter MJ, Kolte L, Briend A, Friis H, Christensen VB. The immune system in children with malnutrition—a systematic review. PLoS One. 2014;9:e105017.
- Pond CM. Adipose tissue and the immune system. Prostaglandins Leukot Essent Fatty Acids. 2005;73:17–30.
- Kragelund C, Omland T. A farewell to body-mass index? Lancet. 2005;366:1589–91.
- Thomas EL, Frost G, Taylor-Robinson SD, Bell JD. Excess body fat in obese and normal-weight subjects. Nutr Res Rev. 2012;25:150–61.
- Brennan DD, Whelan PF, Robinson K, Ghita O, O’Brien JM, Sadleir R, et al. Rapid automated measurement of body fat distribution from whole-body MRI. AJR Am J Roentgenol. 2005;185:418–23.
- Shai I, Jiang R, Manson JE, Stampfer MJ, Willett WC, Colditz GA, et al. Ethnicity, obesity, and risk of type 2 diabetes in women: a 20-year follow-up study. Diabetes Care. 2006;29:1585–90.
- Pan WH, Flegal KM, Chang HY, Yeh WT, Yeh CJ, Lee WC. Body mass index and obesity-related metabolic disorders in Taiwanese and US whites and blacks: implications for definitions of overweight and obesity for Asians. Am J Clin Nutr. 2004;79:31–9.
- Carlsson AC, Wandell P, Riserus U, Arnlov J, Borne Y, Engstrom G, et al. Differences in anthropometric measures in immigrants and Swedish-born individuals: results from two community-based cohort studies. Prev Med. 2014;69C:151–6.
- Riserus U, de Faire U, Berglund L, Hellenius ML. Sagittal abdominal diameter as a screening tool in clinical research: cutoffs for cardiometabolic risk. J Obes. 2010;2010.