Abstract
Purpose
Excess body fat is a commonly known risk factor for type 2 diabetes. However, whether lean body mass, or fat free mass, could have a protective effect against type 2 diabetes, remains unclear. The aim of this study was to explore the association between lean body mass, fat mass and type 2 diabetes.
Methods
This study used data from the Danish Diet, Cancer and Health cohort of 37,053 men and women, aged 50–64 years at baseline (1993–1997). The exposure was measurements of body composition using bioelectrical impedance analysis. Incident diabetes during follow-up was determined through linkage to the Danish National Diabetes Register. Cox proportional hazards regression analysis was used to estimate HR and 95%CI for the association between lean body mass and incident type 2 diabetes, with and without adjustment for fat mass. A sensitivity analysis was performed, excluding cases of incident type 2 diabetes within the first 2 years of follow-up.
Results
When adjusted for fat mass, the main analysis showed non-linear inverse association between lean body mass and risk of diabetes for men, but not for women. However, the sensitivity analysis found no association for either men or women.
Conclusions
Lean body mass was not associated with incident type 2 diabetes when excluding cases that may have been subclinical at baseline. The results imply that public health should focus on reduction of fat mass for diabetes prevention.
Keywords: Type 2 diabetes, Lean body mass, Fat mass, Body composition
Introduction
An estimated 422 billion people have diabetes worldwide and the prevalence is rising. Diabetes is a major cause of disability and premature death, and thus focus on prevention is important [1]. Commonly known risk factors for type 2 diabetes include overweight and obesity, as defined by excess body fat, and these are directly associated with incidence of type 2 diabetes [2, 3]. However, the role of lean body mass, or fat free mass, remains unclear, and only few prospective cohort studies have been performed. Some studies suggest an inverse association between lean body mass and development of type 2 diabetes [4–6], while others show no association [7, 8]. The existing literature on this topic is sparse and includes a broad range of methods and study populations, making comparison between studies difficult, in contrast to studies focusing on fat mass, which have consistently shown positive association between obesity and risk of type 2 diabetes across study populations [9, 10].
Skeletal muscle mass is a main component of lean body mass, and helps to regulate blood glucose levels by storing glucose, and converting it to energy [11]. Total lean body mass could therefore potentially reduce risk of type 2 diabetes by facilitating stable blood sugar levels, thus preventing hyperglycemia. On the other hand, the role of fat tissue in the development of type 2 diabetes includes the release of free fatty acids (FFA) which reduce insulin secretion from pancreatic ß-cells. This contributes to insulin resistance in muscle tissue and can lead to hyperglycemia [12]. Fat tissue and lean body mass thus affect blood glucose levels, in opposite directions, and the net contributions to the development of type 2 diabetes are unclear.
To investigate how lean body mass and fat tissue affect the development of type 2 diabetes, this study aimed to explore associations between lean body mass, fat mass and type 2 diabetes in a large, prospective cohort study.
Materials and methods
Study design
This cohort study used data from the Diet, Cancer and Health study, which recruited participants from December 1993 to May 1997, including a clinic visit in either Aarhus or Copenhagen where anthropometry was measured, and data from self-administered questionnaires was collected [13].
Subjects
People were invited to participate in the Diet, Cancer and Health cohort if they met the following inclusion criteria: 50–64 years of age at the time of invitation, born in Denmark, living in or near Aarhus or Copenhagen, the two largest cities in Denmark and not previously diagnosed with cancer. Eligible cohort participants were identified and located through the Danish Civil Registration system, which assigns an individual 10-digit number to every person living in Denmark [14].
For this study, participants were excluded if they reported diabetes at entry to the study, or had a diagnosis of diabetes prior to baseline in the National Diabetes Registry, if they were missing covariate information, or if their values for lean body mass(%) and fat mass(%) were implausible. While there is no general consensus on specific cut-off for ranges for plausible fat percentage [15], participants with a fat percentage lower than 5% for men and 12% for women were excluded from the study, as these values were considered implausible.
Exposures
Measurements of body composition were obtained by bioelectrical impedance analysis (BIA), performed by trained technicians at the study centers. Measurements were obtained using a 50 Hz single frequency device (BIA 101-F, Akern/RJL, Florence Italy), as described elsewhere [16]. Body composition was calculated using algorithms including sex, age, weight and height, developed previously based on data from validated BIA measurements of 139 Danish adults [17]. Absolute values for lean body mass and fat mass were calculated based on the relative measurements of lean body mass and fat mass, respectively, and the total weight of each participant. In this study, lean body mass(%) is defined as total body mass, minus fat mass(%).
Other variables
Measurements of height and weight were taken by trained lab technicians at the study centers during the clinical visit. Height was measured to the nearest 0.5 cm, with the participants standing without shoes. Weight was measured by digital scales with the participants wearing light clothing, and recorded to the nearest 0.1 kg. [18]. Information on other covariates, including physical activity (>/< 30mins/day), smoking (never, former or current smoker), alcohol intake (no alcohol intake, <10 g/d, <20 g/d, <30 g/d or > 30 g/d) and educational history (none, short, medium or long tertiary education) was obtained by completion of a self-reported questionnaire, filled in by the participants at the study centers [13].
Outcome
Incident cases of diabetes during follow-up were obtained by register linkage to the Danish National Diabetes Register. A person was registered in the National Diabetes Register as a case if he/she was either registered in the National Patient Register with a diagnosis of diabetes, was registered as a diabetic patient receiving chiropody in the National Health Service Register, had five blood-glucose measurements in a 1 year period in the National Health Service Register, had two blood-glucose measurements per year in 5 consecutive years in the National Health Service Register, or purchased anti-diabetic drugs, as recorded in the Danish National Prescription register. The date of inclusion in the register was therefore not the date of diabetes diagnosis, but the date on which one of the above inclusion criteria was met. Incidence data in the National Diabetes Register was complete from August 1997 onwards. The National Diabetes Register did not differentiate between types of diabetes [19].
Information regarding death and emigration was obtained by linking the records from the Diet, Cancer and Health cohort to those of the Civil Registration System, which contains complete information on death and emigration for all people living in Denmark.
All cohort participants were followed from entry to the study to censoring, which was the age of diagnosis in National Diabetes Register, death, emigration or end of follow-up on 31.12.2011, whichever came first.
Ethics
All cohort participants provided informed consent to participate and the study was approved by relevant Ethical Committees and the Danish Data Protection Agency [13]. The present study was approved by the Danish Data Protection Agency in December 2017.
Statistical analysis
Cox Proportional Hazards Regression models were used to estimate hazard ratios (HR) and 95% confidence intervals (CI) for the association between lean body mass and incident type 2 diabetes, with age as the underlying time-variable. Age at recruitment was defined as entry into the study, and exit was defined as age at censoring. The associations between lean body mass and type 2 diabetes were modeled as 4-knotted, restricted cubic splines, using the mean as reference. Analyses were performed separately for men and women, to account for sex differences in body composition.
Initially, the association between lean body mass and incident type 2 diabetes was investigated (Model 1a) and then adjusted for fat mass (Model 1b). Then a sensitivity analysis was performed with the aim to investigate any differences between short- and long-term risk. The sensitivity analysis excluded incident diabetes cases that occurred within the first 2 years of follow-up, with the aim to assess a potential difference in the association between lean body mass and all cases or later-onset cases. Finally, we assessed the association between fat mass (kg) and incident type 2 diabetes, with and without adjustment for lean body mass, in the full cohort (Model 2a and 2b).
All models included adjustment for age, physical activity, smoking, alcohol intake and education, and were stratified by quartiles of inclusion date and performed separately for each sex. Potential confounders were entered into all models as categorical variables, as shown in Table 1. The proportional hazards assumption could not be rejected for any of the analyses (p > 0,05). All statistical analyses were performed in Stata/IC version 15.1, with a statistical significance level of 5%.
Table 1.
Baseline characteristics of all study participants (total n = 54,295) and cases of incident type 2 diabetes
| Men | Women | |||
|---|---|---|---|---|
| All Participants (N = 25,770) |
Cases (N = 3973) |
All Participants (N = 28,525) |
Cases (N = 3178) |
|
| Age at entry (mean(sd)) | 56,1 (4,4) | 56,4 (4,4) | 56,2 (4,4) | 56,8 (4,4) |
| Lean body mass (kg) (mean(sd)) | 60,6 (6,1) | 62,3 (6,7) | 44,4 (4,5) | 46,0 (5,2) |
| Fat mass (kg) (mean(sd)) | 22,3 (7,4) | 26,8 (8,4) | 24,4 (8,7) | 30,3 (10,5) |
| Lean body mass (%) (mean(sd)) | 73,7 (5,4) | 70,5 (5,1) | 65,5 (6,6) | 61,3 (6,4) |
| Fat mass (%) (mean(sd)) | 26,3 (5,4) | 29,5 (5,1) | 34,5 (6,6) | 38,7 (6,4) |
| Physical activity (%(N)) | ||||
| <30 min/day | 62,0% (15,949) | 68,0% (2702) | 59,0% (16,821) | 64,0% (2036) |
| >30 min/day | 38,0% (9821) | 32,0% (1271) | 41,0% (11,704) | 36,0% (1142) |
| Smoker (%(N)) | ||||
| Never | 25,9% (6670) | 21,4% (854) | 43,6 (12,485) | 41,0% (1307) |
| Former | 34,7% (8925) | 36,2% (1436) | 23,6% (6709) | 23,3% (742) |
| Current | 39,5% (10,175) | 42,4% (1683) | 32,8% (9331) | 35,6% (1129) |
| Alcohol intake (%(N)) | ||||
| 0 g/day | 1,8% (454) | 1,9% (74) | 2,62% (752) | 3,8% (121) |
| <10 g/day | 22,2% (5721) | 24,1% (959) | 49,3% (14,067) | 53,9% (1709) |
| <20 g/day | 27,3% (7056) | 26,2% (1042) | 27,1% (7742) | 22,8% (727) |
| <30 g/day | 11,3% (2906) | 9,8% (386) | 4,52% (1293) | 4,4% (140) |
| ≥30 g/day | 37,44% (9641) | 38,1% (1513) | 16,4% (4704) | 15,1% (481) |
| Education (%(N)) | ||||
| No tertiary education | 9,8% (2531) | 12,8% (511) | 19,1% (5444) | 26,0% (827) |
| Short tertiary education | 13,7% (3517) | 14,9% (593) | 31,6% (8988) | 31,1% (986) |
| Medium tertiary education | 42,3% (10,893) | 43,2% (1714) | 38,0% (10,863) | 34,6% (1101) |
| Long tertiary education | 34,2% (8829) | 29,1% (1155) | 11,3% (3230) | 8,3% (264) |
| Body Mass Index (BMI) categories | ||||
| (%(N)) | ||||
| BMI <25 | 35,3% (9089) | 16,5% (655) | 52,2% (14,877) | 26,7% (850) |
| BMI 25–30 | 50,1% (12,911) | 50,6% (2008) | 34,3% (9786) | 39,8% (1266) |
| BMI >30 | 14,6% (3769) | 32,9% (1310) | 13,5% (3862) | 33,5% (1062) |
Results
Of the 160,725 people who met the inclusion criteria, and received an invitation, 57,053 people, including 27,190 men and 29,863 women, chose to participate [13]. Of these, 569 people were excluded due to cancer diagnosed prior to enrolment, 161 people were excluded due to self-reported type 2 diabetes at baseline, 1222 people were excluded due to a diagnosis of diabetes in the National Diabetes Register at baseline, 770 people were excluded due to missing information and 36 people were excluded due to implausible values of fat percentage. These exclusions resulted in a total of 54,295 eligible participants, including 25,770 men and 28,525 women, for this study. During a total of 765,682 person years of follow-up and an average follow-up time of 14,1 years, 7151 cases of incident diabetes were recorded, including 3973 men and 3178 women.
Participants who developed diabetes during follow-up had lower lean body mass(%) and a higher BMI at baseline than the full cohort (Table 1). Participants who developed diabetes during follow-up were less likely to report being physically active, were more likely to be current smokers, to be of lower educational attainment, and to report a history of hypertension or hypercholesterolemia at baseline, than the full cohort.
Baseline characteristics for the early incident cases, who were excluded in the secondary analysis, revealed poorer health compared to the later incident cases, with early cases being more likely to have a BMI > 30, a higher fat percentage, and be less physically active than the later cases.
Lean body mass was positively associated with type 2 diabetes in both sexes (model 1a). However, when adjusting for fat mass, a weak inverse association with risk of type 2 diabetes was seen in men, but not in women (Fig. 1).
Fig. 1.
The association between lean body mass and incident type 2 diabetes, modelled by 4-knotted restricted cubic splines for men and women (n = 54,295). Models 1a were adjusted for age, physical activity, smoking, alcohol consumption and education. Models 1b include additional adjustment for fat mass. The solid lines show HR while the dotted lines show 95% CI, with reference point at the mean for lean body mass.
The sensitivity analysis excluding cases occurring within the first 2 years of follow up (n = 465), showed no inverse association between lean body mass and risk of type 2 diabetes, either for men or for women, when adjusted for fat mass (Fig. 2).
Fig. 2.
The association between lean body mass and incident type 2 diabetes, modelled by 4-knotted restricted cubic splines for men and women, after exclusion of incident cases within the first 2 years of follow up were excluded (N = 465). Models 1a were adjusted for age, physical activity, smoking, alcohol consumption and education. Models 1b on the right include additional adjustment for fat mass. The solid lines show the association (HR) while the dotted lines show 95% CI, with reference point at the mean for lean mass (kg)
Fat mass was statistically significantly positively associated with incident type 2 diabetes in both sexes, and the association remained largely unchanged after adjustment for lean body mass (Fig. 3).
Fig. 3.
The association between fat mass and incident type 2 diabetes, modelled by 4-knotted restricted cubic splines for men and women. Models 2a were adjusted for age, physical activity, smoking, alcohol consumption and education. Models 2b on the right include additional adjustment for lean body mass. The solid lines show HR while the dotted lines show 95% CI, with reference point at the mean for fat mass
Discussion
This prospective cohort study aimed to investigate how lean body mass and fat tissue affect the development of incident type 2 diabetes. In the main analyses, we observed that a high level of lean body mass was associated with a slightly lower risk of type 2 diabetes in men after adjusting for fat mass, but not in women. In sensitivity analyses excluding cases occurring within the first 2 years of follow-up, we found no association between lean body mass and type 2 diabetes for either sex. High levels of fat (kg) were associated with a higher risk of type 2 diabetes, as previously identified in other studies [9, 10]. This association persisted when adjusting for lean body mass.
This study has strengths as well as limitations. Only 286 (0.53%) participants emigrated during follow-up, limiting the concern for selection bias. The use of BIA to determine body composition has previously been validated against DXA and has been found to be a valid tool, with very little difference in results of the measurements produces by these two methods [20–23], limiting concern for misclassification of the exposure. The National Diabetes Register does not differentiate between types of diabetes. However, given the age of the study population diagnoses of type 2 diabetes, rather than other types, are most likely in this study. Registration in the National Diabetes Register is independent of exposure measurement in the cohort, limiting concern for differential misclassification of the outcome. Nevertheless, because incident type 2 diabetes in this study was determined upon registration in the National Diabetes Register, which in many cases occurs when a patient needs treatment for their condition, incident cases registered early on in the follow-up period may actually have had type 2 diabetes prior to entry to the study. People with type 2 diabetes have been found to have an accelerated loss of muscle mass, compared to healthy individuals of a similar age [24, 25]. If the baseline measurements of lean body mass were lower for people who already had type 2 diabetes, unregistered in the National Diabetes Register, than for healthy participants, this could induce an association between lean body mass and type 2 diabetes, as a result of reverse causation. To assess if this was the case, we performed a sensitivity analysis, in which we excluded cases registered within the first 2 years of follow-up. The difference between the two analyses indicates that reverse causation may have caused the inverse association in the main analysis, as the association was attenuated in the sensitivity analysis. However, if the initial cases were those most sensitive to the development of diabetes as a result of their body composition, compared to later cases, this exclusion may introduce selection bias into the study. We adjusted for known confounders, which did not alter the strength of the observed associations. Nevertheless, residual confounding cannot be excluded. The study population was largely white, and caution should be taken when extrapolating results to populations of other ethnicity, for whom anthropometry may associate differently with risk of diabetes [26, 27].
Existing knowledge regarding the regulation of blood glucose supports the plausibility of an inverse association between lean mass and risk of type 2 diabetes [11]. lean body mass plays a role in reducing blood sugar, while fat tissue reduces insulin secretion, which can in turn lead to hyperglycemia [12]. Thus, a person with greater lean body mass could have a lower risk of developing type 2 diabetes than a person with less lean body mass. However, this study indicates that this is not the case.
Previous studies have defined lean body mass in different ways, using calculated measurements such as “Muscle Mass Index” [4] and “Relative Muscle Mass” [5]. These differences in definition of the exposure along with the variations between studies mentioned above, means comparison of existing results with those of this study must be done with some caution. Previous studies of the association between lean body mass and type 2 diabetes have unsurprisingly shown differing results, with some studies identifying an inverse association [4–6] and others not [7, 8]. The difference in results of the analysis of the association between lean body mass and type 2 diabetes may be due to differences in the analytical approach, where some studies include adjustment for fat mass, and others do not [5, 6]. One study analyzed tertiles of “Muscle Mass Index” and risk of type 2 diabetes, adjusting for fat percentage as a dichotomous variable [4]. However, the absolute values of fat are likely to vary between the tertiles, which would not be captured in this analytical approach. As a result, the observed inverse associations may actually be due to low fat mass, and not high lean body mass. Of the two existing studies that do not identify an association between lean body mass and type 2 diabetes, one included adjustment for total body fat [7]. The other [8] assessed differences between intervals of lean body mass, and similarly found no association, which is in line with the results of this study. However, several studies are in agreement that fat mass is an important risk factor for developing type 2 diabetes [9, 10, 12].
Conclusions
This study suggests that lean mass is not significantly associated with risk of incident type 2 diabetes for men or for women when early cases, that may have been subclinical at entry to the study, are excluded. The findings may be used in a public health context to inform health practitioners and the general public alike, that focus should be on reducing absolute fat mass, when the main aim is to reduce diabetes risk.
Compliance with ethical standards
Conflict of interest
The authors declared that they have no conflict of interest.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.World Health Organisation (WHO). Diabetes [Internet]. Available from: https://www.who.int/news-room/fact-sheets/detail/diabetes. Accessed 9 Sept 2019
- 2.WHO | Obesity and overweight [Internet]. WHO. [cited 2018 Feb 8]. Available from: http://www.who.int/mediacentre/factsheets/fs311/en/. Accessed 9 Sept 2019
- 3.Weisman A, Fazli GS, Johns A, Booth GL. Evolving trends in the epidemiology, risk factors, and prevention of Type 2 Diabetes: a review. Can J Cardiol. 2018;34(5):552–564. doi: 10.1016/j.cjca.2018.03.002. [DOI] [PubMed] [Google Scholar]
- 4.Son JW, Lee SS, Kim SR, Yoo SJ, Cha BY, Son HY, Cho NH. Low muscle mass and risk of type 2 diabetes in middle-aged and older adults: findings from the KoGES. Diabetologia. 2017;60(5):865–872. doi: 10.1007/s00125-016-4196-9. [DOI] [PubMed] [Google Scholar]
- 5.Hong Sungwoo, Chang Yoosoo, Jung Hyun-Suk, Yun Kyung Eun, Shin Hocheol, Ryu Seungho. Relative muscle mass and the risk of incident type 2 diabetes: A cohort study. PLOS ONE. 2017;12(11):e0188650. doi: 10.1371/journal.pone.0188650. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Tatsukawa Yoshimi, Misumi Munechika, Kim Young Min, Yamada Michiko, Ohishi Waka, Fujiwara Saeko, Nakanishi Shuhei, Yoneda Masayasu. Body composition and development of diabetes: a 15-year follow-up study in a Japanese population. European Journal of Clinical Nutrition. 2018;72(3):374–380. doi: 10.1038/s41430-017-0077-7. [DOI] [PubMed] [Google Scholar]
- 7.Larsen BA, Wassel CL, Kritchevsky SB, Strotmeyer ES, Criqui MH, Kanaya AM, Fried LF, Schwartz AV, Harris TB, Ix JH, Health ABC Study Association of Muscle Mass, area, and strength with incident diabetes in older adults: the health ABC study. J Clin Endocrinol Metab. 2016;101(4):1847–1855. doi: 10.1210/jc.2015-3643. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Li JJ, Wittert GA, Vincent A, Atlantis E, Shi Z, Appleton SL, Hill CL, Jenkins AJ, Januszewski AS, Adams RJ. Muscle grip strength predicts incident type 2 diabetes: population-based cohort study. Metabolism. 2016;65(6):883–892. doi: 10.1016/j.metabol.2016.03.011. [DOI] [PubMed] [Google Scholar]
- 9.Garg S. K., Maurer H., Reed K., Selagamsetty R. Diabetes and cancer: two diseases with obesity as a common risk factor. Diabetes, Obesity and Metabolism. 2013;16(2):97–110. doi: 10.1111/dom.12124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Bellou V, Belbasis L, Tzoulaki I, Evangelou E. Risk factors for type 2 diabetes mellitus: an exposure-wide umbrella review of meta-analyses. PLoS ONE. 2018;13(3):e0194127. doi: 10.1371/journal.pone.0194127. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kjøbsted R, Hingst JR, Fentz J, Foretz M, Sanz M-N, Pehmøller C, Shum M, Marette A, Mounier R, Treebak JT, Wojtaszewski JFP, Viollet B, Lantier L. AMPK in skeletal muscle function and metabolism. FASEB J. 2018;32(4):1741–1777. doi: 10.1096/fj.201700442R. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Roberts CK, Hevener AL, Barnard RJ. Metabolic syndrome and insulin resistance: underlying causes and modification by exercise training. Compr Physiol. 2013;3(1):1–58. doi: 10.1002/cphy.c110062. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Tjønneland A, Olsen A, Boll K, Stripp C, Christensen J, Engholm G, Overvad K. Study design, exposure variables, and socioeconomic determinants of participation in diet, cancer and health: a population-based prospective cohort study of 57,053 men and women in Denmark. Scand J Public Health. 2007;35(4):432–441. doi: 10.1080/14034940601047986. [DOI] [PubMed] [Google Scholar]
- 14.Pedersen CB. The Danish civil registration system. Scand J Public Health. 2011;39(7_suppl):22–25. doi: 10.1177/1403494810387965. [DOI] [PubMed] [Google Scholar]
- 15.Gallagher D, Heymsfield SB, Heo M, Jebb SA, Murgatroyd PR, Sakamoto Y. Healthy percentage body fat ranges: an approach for developing guidelines based on body mass index. Am J Clin Nutr. 2000;72(3):694–701. doi: 10.1093/ajcn/72.3.694. [DOI] [PubMed] [Google Scholar]
- 16.Stegger JG, Schmidt EB, Obel T, Berentzen TL, Tjønneland A, Sørensen TIA, Overvad K. Body composition and body fat distribution in relation to later risk of acute myocardial infarction: a Danish follow-up study. Int J Obes. 2011;35(11):1433–1441. doi: 10.1038/ijo.2010.278. [DOI] [PubMed] [Google Scholar]
- 17.Heitmann BL. Prediction of body water and fat in adult Danes from measurement of electrical impedance. A validation study. Int J Obes. 1990;14(9):789–802. [PubMed] [Google Scholar]
- 18.Bigaard J, Spanggaard I, Thomsen BL, Overvad K, Tjønneland A. Self-reported and technician-measured waist circumferences differ in middle-aged men and women. J Nutr. 2005;135(9):2263–2270. doi: 10.1093/jn/135.9.2263. [DOI] [PubMed] [Google Scholar]
- 19.Carstensen B, Kristensen JK, Marcussen MM, Borch-Johnsen K. The National Diabetes Register. Scand J Public Health. 2011;39(7_suppl):58–61. doi: 10.1177/1403494811404278. [DOI] [PubMed] [Google Scholar]
- 20.Malavolti M, Mussi C, Poli M, Fantuzzi AL, Salvioli G, Battistini N, Bedogni G. Cross-calibration of eight-polar bioelectrical impedance analysis versus dual-energy X-ray absorptiometry for the assessment of total and appendicular body composition in healthy subjects aged 21-82 years. Ann Hum Biol. 2003;30(4):380–391. doi: 10.1080/0301446031000095211. [DOI] [PubMed] [Google Scholar]
- 21.Achamrah Najate, Colange Guillaume, Delay Julie, Rimbert Agnès, Folope Vanessa, Petit André, Grigioni Sébastien, Déchelotte Pierre, Coëffier Moïse. Comparison of body composition assessment by DXA and BIA according to the body mass index: A retrospective study on 3655 measures. PLOS ONE. 2018;13(7):e0200465. doi: 10.1371/journal.pone.0200465. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Ling CHY, de Craen AJM, Slagboom PE, Gunn DA, Stokkel MPM, Westendorp RGJ, Maier AB. Accuracy of direct segmental multi-frequency bioimpedance analysis in the assessment of total body and segmental body composition in middle-aged adult population. Clin Nutr. 2011;30(5):610–615. doi: 10.1016/j.clnu.2011.04.001. [DOI] [PubMed] [Google Scholar]
- 23.Moon JR, Stout JR, Smith-Ryan AE, Kendall KL, Fukuda DH, Cramer JT, Moon SE. Tracking fat-free mass changes in elderly men and women using single-frequency bioimpedance and dual-energy X-ray absorptiometry: a four-compartment model comparison. Eur J Clin Nutr. 2013;67(S1):S40–S46. doi: 10.1038/ejcn.2012.163. [DOI] [PubMed] [Google Scholar]
- 24.Kalyani RR, Corriere M, Ferrucci L. Age-related and disease-related muscle loss: the effect of diabetes, obesity, and other diseases. Lancet Diabetes Endocrinol. 2014;2(10):819–829. doi: 10.1016/S2213-8587(14)70034-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Park SW, Goodpaster BH, Lee JS, Kuller LH, Boudreau R, de Rekeneire N, Harris TB, Kritchevsky S, Tylavsky FA, Nevitt M, Cho YW, Newman AB, for the Health, Aging, and Body Composition Study Excessive loss of skeletal muscle mass in older adults with type 2 diabetes. Diabetes Care. 2009;32(11):1993–1997. doi: 10.2337/dc09-0264. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Misra A, Khurana L. Obesity-related non-communicable diseases: south Asians vs white Caucasians. Int J Obes. 2011;35(2):167–187. doi: 10.1038/ijo.2010.135. [DOI] [PubMed] [Google Scholar]
- 27.Han SJ, Kim HJ, Kim DJ, Lee KW, Cho NH. Incidence and predictors of type 2 diabetes among Koreans: a 12-year follow up of the Korean genome and epidemiology study. Diabetes Res Clin Pract. 2017;123:173–180. doi: 10.1016/j.diabres.2016.10.004. [DOI] [PubMed] [Google Scholar]