Abstract
Objective
Body Adiposity Index (BAI), a new surrogate measure of body fat (hip circumference/[height 1.5–18]), has been proposed as a more accurate alternative to BMI. We compared BAI with BMI and their correlations with measures of body fat, waist circumference (WC), and indirect indices of fat pre- and post-Roux-en-Y gastric bypass (RYGB).
Methods
Sixteen clinically severe obese (CSO) non-diabetic women (age = 33.9± 7.9 SD; BMI = 46.5±9.5 kg/m2) were assessed pre-surgery, and at 2 (n=9) and 5 mo (n=8) post-surgery. Body fat percentage (% fat) was estimated with bioimpedance analysis (BIA), air displacement plethysmography (ADP), and dual-energy x-ray absorptiometry (DXA). WC, an indicator of central fat, and both plasma leptin (ng/ml) and insulin (mU/l) concentrations were measured as indirect body fat indices. Pre- and post-surgery values were analyzed with Pearson correlations and linear regressions.
Results
BAI and BMI correlated significantly with each other pre-surgery and at each time point post surgery. BAI and BMI also correlated significantly with % fat from BIA and ADP; however, only BMI correlated significantly with % fat from DXA pre- and post-RYGB. BMI was the single best predictor of WC and leptin at 2 and 5 mo post-surgery and had significant longitudinal changes correlating with % fat from BIA and DXA as well as with leptin.
Discussion
Both BAI and BMI were good surrogates of % fat as estimated from BIA and ADP, but only BMI was a good surrogate of % fat from DXA in CSO women. Thus, BAI may not be a better alternative to BMI.
Keywords: Body Adiposity Index, Body Mass Index, Roux-en-Y gastric bypass, leptin, insulin, waist circumference
Introduction
Bariatric surgery is the most effective intervention for clinically severe obesity (CSO), with Roux-en-Y gastric bypass (RYGB) the most commonly performed procedure in the US. Improvement in insulin sensitivity post-surgery; however, varies possibly related to individual differences in pre-surgery baseline and post-surgical changes in body composition (ie % fat and fat distribution) and adiposity related hormones (ie leptin and insulin) [1]. Relatively sophisticated techniques to estimate body fat (ie air displacement plethysmography [ADP] and dual energy x-ray absorptiometry [DXA]) are not suitable for clinical practice, and thus surrogate measures, such as body mass index (BMI), are widely used.
BMI, a ratio of body weight to height squared (kg)/m2), can be used to approximate body fat. However, a higher BMI does not necessarily reflect greater fat mass, as it can be unduly influenced by muscle mass, which can be related to genetics, age, sex, ethnicity, and exercise [2]. Body Adiposity Index (BAI), a newly defined clinical measure, is based on a ratio of hip circumference to height to the 1.5 power minus 18 (cm/[(m)1.5 − 18]), which has been shown to correlate well with body fat percentage (% fat) [3]. These non-invasive surrogate measures can be useful tools to help determine a patient’s risk for obesity-related diseases. Waist circumference (WC) by itself is a key marker of central adiposity and is commonly used to assess cardiometabolic risk associated with obesity [4–6].
This study examined how BAI compared with BMI, particularly in their correlations with sophisticated estimates of body fat (ie ADP, DXA), WC, as well as adiposity signals (ie leptin, insulin) in clinically severe obese (CSO) women pre-surgery and at both 2 and 5-mo post RYGB. This is the first study to examine the comparative usefulness of the surrogate measures, BAI and BMI before and after marked weight loss in relation to the expected improvement in fat-related cardiometabolic risk factors (WC, leptin and insulin).
Methods
Participants
Sixteen CSO non-diabetic women age, 33.9±7.9 (mean ± SD) and BMI, 46.5±9.5 kg/m2, were recruited prior to bariatric surgery (Table 1). Of the 16 candidates who underwent RYGB, 9 returned at 2 mo, and 8 returned at 5 mo post-surgery. Pre-surgery, each participant completed a physical examination, including medical history, electrocardiogram, and routine blood tests to exclude the presence of underlying disease. Participants underwent RYGB by the same surgeon. They signed consent forms, which were approved by St. Luke’s-Roosevelt Hospital’s Institutional Review Board (IRB).
Table 1.
Participant measures pre- and post- Roux-en-Y gastric bypass (means±SD).
| Measures | Pre-RYGB | 2 mo Post-RYGB | 5 mo Post-RYGB |
|---|---|---|---|
| Weight (kg) | 126.1±25.6 | 108.3±20.7*** | 101.6±27.4*** |
| Height (m) | 1.70±0.1 | 1.70±0.1 | 1.70±0.1 |
| Waist (cm) | 124.2±17.1 | 114.2±14.3** | 103.6±13.7*** |
| Hip (cm) | 139.7±17.4 | 132.2±18.6 | 124.9±16.8*** |
| WHR | 0.90±0.1 | 0.90±0.1 | 0.80±0.1 |
| BAI | 48.1±10.1 | 44.8±12.2** | 39.9±6.5*** |
| BMI (kg/m2) | 46.5±9.5 | 40.2±8.2*** | 36.5±7.5*** |
| BIA (% fat) | 49.8±4.5 | 46±5.9* | 42.7±4.6*** |
| ADP (% fat) | 53.6±6.0 | 49.6±6.0** | 45.4±7.8*** |
| DXA (% fat) | 58.3±9.4 | 52.2±13.6 | 48.4±18.4* |
| Plasma leptin (ng/ml) | 52.1±23.3 | 22.9±12.9** | 20.7±13.1* |
| Plasma insulin (mU/l) | 21.3±6.9 | 12.2±8.3** | 10.2±3.3** |
| Plasma glucose (mg/dl) | 91±12.8 | 85.5±9.8 | 90.4±8.9 |
| HOMA-IR | 4.9±01.9 | 2.7±1.9 | 2.3±1.0** |
| QUICKI-IS | 0.22±0.01 | 0.24±0.01* | 0.24±0.01* |
P <0.05,
P <0.01,
P <0.001 change from pre-surgery within the same individuals.
Of the 16 at pre-surgery, 8 were African American (AA), 6 Hispanic (H), and 2 Caucasian (C). At 2 mo post-RYGB, there were 5 AA, 2 H, and 2 C, and at 5 mo, 5 AA, 2H, and 1 C.
RYGB = Roux-en-Y gastric bypass; WHR = waist-to-hip ratio; BAI = body adiposity index; BMI = body mass index; BIA = bioimpedance analysis; ADP = air displacement plethysmography; DXA = dual-energy X-ray absorptiometry; HOMA-IR= homeostasis model of assessment- insulin resistance; QUICKI-IS = quantitative insulin sensitivity check index.
Study protocol
Following an overnight fast from 8 pm, procedures were performed, pre-surgery, 2 and 5 mo post-surgery and included body composition (9 – 11 am) and blood draws (11 am). The % fat was calculated with bioelectrical impedance analysis (BIA), air displacement plethysmography (ADP), and dual-energy x-ray absorptiometry (DXA). Following surgery, participants had weekly nutritional counseling sessions to help insure compliance with the recommended post-surgical diet, including mineral and vitamin supplements. All body composition procedures were performed and interpreted by trained technologists.
Anthropometric measurements
Body weight was measured to the nearest 0.1 kg, using a calibrated scale (Weight Tronix, New York, NY). Height was measured to the nearest 0.1 cm, using a stadiometer (SECA 217). WC was measured in a standing position, at the level located halfway between the lower rib margin and the iliac crest. Hip circumference (HC) was measured at the level of the greater trochanter of the femoral bone, which was palpated laterally.
Bioimpedance analysis (BIA)
Body fat was estimated by bioelectrical impedance analysis (BIA), using a Tanita instrument (TBF-300A).
Air displacement plethysmography (ADP)
ADP was performed using BodPod (Life Measurement Instruments, Concord CA, USA) to estimate body volume and density. Participants wore a tight-fitting bathing suit and acrylic bathing cap. They were weighed to the nearest 0.01 kg, using the ADP electronic scale (Tanita Corp., Tokyo, Japan) [7].
Whole body dual-energy X-ray absorptiometry (DXA)
Total body fat was estimated, using a 2-fan beam (GE Lunar Prodigy, Madison, WI) DXA system. Percent body fat was calculated by dividing total fat obtained by the scanner by total body weight. The intraclass correlation coefficient at the Core Lab for % fat estimation by DXA was 0.994.
Blood sampling
Blood was taken at the Clinical Research Resource (CRR) of the New York Obesity Nutrition Research Center (NYONRC). Blood samples were drawn through a catheter from an in-dwelling butterfly needle inserted into a forearm vein. The catheter was connected to a slow drip normal saline bag to keep the vein patent. Blood samples with EDTA added were cold centrifuged (3000 RPM) for 15 min, and stored in cryomicrotubes at −80°C until assayed. Plasma leptin (ng/ml) and insulin (mU/L) were assayed using RIA kits from Millipore (Linco) (leptin: intraclass CV = 4.7%, interclass CV = 5.9%; insulin: intraclass CV = 4.4%, interclass CV = 4.5%). Serum glucose was measured using the hexokinase method. Additional measures for insulin resistance and sensitivity (inverse of resistance) were calculated from insulin and glucose: Insulin resistance was obtained using the homeostasis model of assessment (HOMA-IR) [8] (fasting glucose [mgl/dl] × fasting insulin [mU/ml])/405, and insulin sensitivity was derived from the quantitative insulin sensitivity check index (QUICKI-IS) [9] 1/(log fasting glucose [mg/dl]+log fasting insulin [mU/ml].
Data analysis
Descriptive statistics are expressed as mean ± SD. Univariate and repeated measures ANOVA were performed for values at pre-surgery, 2 and 5 mo post-surgery. Pearson’s r correlations and stepwise linear regression analyses were conducted to determine the relationship between BMI and BAI and their relationship with % fat (ie BIA, ADP, DXA), WC (measure of central fat), and hormonal indices of fat (ie, plasma leptin and insulin). All statistical analyses were conducted with SPSS 18 (Chicago, IL) based on two-tailed (α = 0.05) tests of significance.
Results
The pre- and post-surgery participant measures, including anthropometric data and blood concentrations, are presented in Table 1.
Pre-RYGB (baseline)
BAI correlated significantly with BMI (r=0.89, P <0.001) and with % fat from both BIA (r=0.86, P <0.001) and ADP (r=0.71, P =0.002), but not with % fat from DXA, nor with WC, leptin, insulin, HOMA-IR, or QUICK-IS as shown in Table 2. On the other hand, BMI was significantly correlated with % fat from BIA (r=0.88, P <0.001), ADP (r=0.79 P <0.001) as well as DXA (r=0.62, P =0.01), but not with WC, leptin, insulin, HOMA-IR, or QUICKI-IS.
Table 2.
Correlations (Pearson’s r) between Body Adiposity Index (BAI) and Body Mass Index (BMI) and with fat-related measures pre and post- Roux-en-Y gastric bypass.
| Measures | Pre-RYGB (n=16) | 2 mo Post-RYGB (n=9) | 5 mo Post-RYGB (n=8) |
|---|---|---|---|
| BAI | |||
| BIA (% fat) | 0.86*** | 0.68* | 0.84** |
| ADP (% fat) | 0.71** | 0.71* | 0.72* |
| DXA (% fat) | 0.39 | 0.59 | 0.57 |
| WC (cm) | 0.43 | 0.76* | 0.72* |
| Plasma leptin (ng/ml) | 0.30 | 0.89 ** | 0.48 |
| Plasma insulin (mU/l) | 0.31 | −0.30 | −0.08 |
| HOMA-IR | 0.19 | −0.32 | −0.08 |
| QUICKI-IS | −0.23 | 0.37 | −0.05 |
|
| |||
| BMI
| |||
| BIA (% fat) | 0.88*** | 0.82** | 0.89** |
| ADP (% fat | 0.79*** | 0.78* | 0.86** |
| DXA (% fat) | 0.62* | 0.78* | 0.87** |
| WC (cm) | 0.45 | 0.89** | 0.78* |
| Plasma leptin (ng/ml) | 0.26 | 0.89** | 0.81* |
| Plasma insulin (mU/l) | 0.40 | 0.03 | −0.03 |
| HOMA-IR | 0.28 | −0.01 | −0.09 |
| QUICKI-IS | −0.29 | 0.10 | −0.03 |
P < 0.05,
P < 0.01,
P < 0.001
RYGB = Roux-en-Y gastric bypass; BIA = bioimpedance analysis; ADP = air displacement plethysmography; DXA = dual-energy X-ray absorptiometry; WC = waist circumference; HOMA-IR= homeostasis model of assessment- insulin resistance; QUICKI-IS = quantitative insulin sensitivity check index.
When % fat from BIA, ADP, and DXA were correlated with WC and the indirect fat indices (leptin and insulin), only the % fat derived from DXA correlated significantly with insulin levels (r=0.52, P =0.04), HOMA-IR (r=0.59, P =0.02,) and QUICKI-IS (r=−0.49, P =0.05).
2 mo Post-RYGB
As expected, body weight, BAI and BMI were all significantly lowered following the surgical intervention (Table 1). Weight loss (kg) at 2 mo did not correlate with a higher pre-surgery BAI (r =0.10, P =0.80), but as expected, correlated with a higher pre-surgery BMI (r=0.94, P <0.001). Postsurgical reductions in % fat were significant based on BIA (3.8±3.5 SD, P =0.01) and ADP (4.3±2.7, P =0.002) although not based on DXA (5.0±8.4, P =0.11). Both WC (8.1±6.5, P =0.006) and leptin (31.4±20.6, P =0.002) also decreased significantly. There was an apparent reduction in insulin resistance as reflected by a decrease in fasting insulin levels (9.2±7.6, P =0.007), a marginally significant decrease of HOMA-IR (1.8±2.4, P =0.06), and a significant increase in QUIC-KI-IS (0.01±0.01, P =0.01).
At 2 mo, BAI was correlated with BMI (r=0.94, P =0.001) and with % fat from BIA (r=0.68, P =0.04) and ADP (r=0.71, v=0.03), but not with % fat from DXA (r=0.59, P =0.09). On the other hand, BMI was significantly correlated with % fat from BIA (r=0.82, P =0.007), ADP (r=0.78, P =0.01) as well as DXA (r=0.78, P =0.01). Although BAI was positively correlated with both WC and leptin (Table 2), stepwise linear regression analyses showed that BMI was the single best predictor of WC (r=0.89, P =0.001) and leptin (r=0.89, P =0.001). There were no significant BAI or BMI correlations with insulin, HOMI-IR, and QUICKI-IS (Table 2),
Body composition changes from pre-surgery baseline to 2 mo post-RYGB
Weight loss (kg) from baseline to 2 mo did not correlate with a change in BAI (r =0.24, P =0.54), but as expected, it correlated with a change in BMI (r=0.92, P <0.001). No significant correlations were found between the change in BAI’s estimation of fat and change in any % fat derived from BIA, ADP, or DXA, or between the change in BAI and the change in WC, plasma leptin, and insulin, controlling for baseline BAI. From pre-surgery baseline to 2 mo, change in BMI was correlated with change in BIA (r=0.85, P =0.008) but not with change in ADP (r=0.15, P =0.72) nor with change in DXA (r=0.69, P =0.06), controlling for baseline BMI. In addition, change in BMI was also correlated with change in leptin (r=0.78, P =0.02) but not with change in WC (r=0.56, P =0.15), insulin (r=0.53, P =0.18), HOMA-IR (r=0.62, P =0.10), or QUICKI-IS (r=−0.56, P =0.15), controlling for baseline BMI.
5 mo Post-RYGB
Body weight (P <0.001) and % fat from BIA (P <0.001), ADP (P <0.001), and DXA (P =0.02) were significantly reduced at 5 mo post-RYGB as compared to baseline. WC, leptin, insulin, HOMA-IR and QUICKI-IS were also significantly reduced (Table 1). Weight loss at 5 mo did not correlate with a higher pre-surgery BAI (r =−0.39, P =0.34), and in contrast to 2 mo, not with a higher pre-surgery BMI (r = −0.06, P =0.90).
At 5 mo, BAI was correlated with BMI (r=0.85, P <0.008). BAI was significantly correlated with BIA (r=0.84, P =0.009) and ADP (r=0.72, P =0.045) but not with DXA (r=0.57, P =0.14) whereas BMI was significantly correlated with % fat from BIA (r=0.89, P = 0.003), ADP (r=0.86, P =0.007) as well as DXA (r=0.87, P =0.005). Although BAI was significantly correlated with WC and leptin (Table 2), stepwise linear regression showed that BMI was the single best predictor of WC (r=0.78, P =0.02) and leptin (r=0.81, P =0.01). Neither BAI nor BMI were correlated with any insulin measure (Table 2).
Body composition changes from baseline to 5 mo post-RYGB
From pre-surgery baseline to 5 mo, changes in any body composition measure, including weight, were not correlated with a change in BAI, controlling for baseline BAI. The only positive correlation found was between a change in BMI and a change in DXA (r=0.78, P =0.04), controlling for baseline BMI (Table 3).
Table 3.
Correlations (Pearson’s r) between Change (Δ) in Body Adiposity Index (BAI) and Δ in Body Mass Index (BMI) with Δ in Fat-Related Measures Pre and Post- Roux-en-Y Gastric Bypass.
| Change from baseline to 2 mo post-RYGB (n=9) | Change from baseline to 5 mo post-RYGB (n=8) | |
|---|---|---|
| ΔBAI
| ||
| BIA (% fat) | 0.05 | 0.70 |
| ADP (% fat) | 0.44 | 0.10 |
| DXA (% fat) | 0.59 | 0.67 |
| WC (cm) | 0.08 | 0.16 |
| Plasma leptin (ng/ml) | 0.45 | 0.49 |
| Plasma insulin (mU/l) | −0.003 | −0.37 |
| HOMA-IR | 0.04 | −0.40 |
| QUICKI-IS | −0.06 | 0.34 |
|
| ||
| ΔBMI
| ||
| BIA (% fat) | 0.85** | 0.30 |
| ADP (% fat) | 0.15 | 0.39 |
| DXA (% fat) | 0.69 | 0.78* |
| WC (cm) | 0.56 | 0.45 |
| Plasma leptin (ng/ml) | 0.78* | 0.44 |
| Plasma insulin (mU/l) | 0.53 | 0.29 |
| HOMA-IR | 0.62 | 0.42 |
| QUICKI-IS | −0.56 | −0.04 |
P < 0.05,
P < 0.01,
P < 0.001
RYGB = Roux-en-Y gastric bypass; BIA = bioimpedance analysis; ADP = air displacement plethysmography; DXA = dual-energy X-ray absorptiometry; WC = waist circumference; HOMA-IR= homeostasis model of assessment- insulin resistance; QUICKI-IS = quantitative insulin sensitivity check index.
Discussion
The primary utility of adiposity surrogates, such as BAI and BMI, is for rapid evaluation in a clinical setting to help determine a patient’s risk for obesity-related diseases. This study contributes to the growing literature on BAI, % fat, and fat-related measures [3, 17–19] and is the first to compare BAI and BMI in their estimations of body fat and fat-related measures before and after marked weight loss from RYGB. Other studies have compared standard body composition measures (ie BIA, DXA) in obese women undergoing bariatric surgery [10, 11].
The first study objective was to examine the relationship of the two surrogate measures with direct body fat indices. BAI and BMI correlated significantly with BIA and ADP measurements of % fat pre- and 2 and 5 mo post-RYGB. DXA gave somewhat higher body % fat than BIA and ADP, and only BMI, not BAI, was significantly correlated with % fat from DXA at each time point. Although BMI has been criticized for being confounded with muscle mass, this may not have been a limiting factor in CSO women.
The next objective was to compare BAI and BMI with indirect measures of % fat. At pre-surgery baseline, there were no significant correlations with WC, leptin, or insulin-related measures, likely due to the surrogate measures not being accurate enough to extend to indirect fat indices. BMI as compared to BAI, was better able to predict WC and leptin following RYGB, suggesting that body weight, on which BMI is based, contributes more than hip circumference, on which BAI is based, to these indirect fat-related surrogates in CSO women. Although total body fat per se may not optimally reflect health risk, larger hip circumferences in CSO women may actually be protective against cardiovascular disease [12–14], This may help explain the stronger associations between BMI as compared to BAI with fat-related cardiovascular risk parameters before and after surgically induced weight loss.
Pre-surgery, there was a significant positive correlation between insulin resistance (assessed by HOMA-IR and fasting insulin) and % fat from DXA only, which is consistent with DXA’s higher body % fat estimates. The rapid decline in fasting insulin following RYGB was associated with decreased insulin resistance and improvement in insulin sensitivity; however, as for pre-surgery neither BAI nor BMI correlated with insulin measures post-surgery most likely due to these surrogate measures not being precise indicators of body fat.
Finally, we assessed longitudinal changes in % fat and fat-related measures with changes in BAI and BMI estimations of % fat pre- and post-RYGB. No association was found with changes in % fat or fat-related measures with changes in BAI over time. However, change in BMI was significantly correlated with change in % fat from BIA and with change in leptin from pre-surgery baseline to 2 mo. The relationship between change in BMI and change in % fat from DXA was of marginal significance at 2 mo but reached statistical significance at 5 mo. Taken together, these findings suggest that BMI was better at approximating changes in direct and indirect fat indices using a within-person design following surgery compared to BAI. Further research with such a design may provide valuable information on clinical outcomes and aid in the development of effective nonsurgical weight loss interventions (e.g. leptin replacement therapy) [15, 16].
Limitations of the study include lack of a non-surgical group and a small sample size. However, our data have thus far been consistent with other research comparing BAI and BMI with % fat [17, 19]. For example, Schulze et al. [19] recently investigated associations between BAI and BMI estimations of fat with % fat derived from magnetic resonance tomography as well as incidence of type 2 diabetes mellitus (DM) in a heterogeneous study population (n=360). They found BMI had higher correlations with % fat and greater associations with DM risk than BAI.
In conclusion, BAI appears to be a valid surrogate for body fat but not an acceptable alternative to BMI in severely obese women.
Acknowledgments
We acknowledge the technical assistance of John Thornton and contribution of Drs. Dympna Gallagher and Edward Colt of the Body Composition Core Lab of the NIH-funded New York Obesity Nutrition Research Center (NYONRC).
Supported in part by Columbia University’s CTSA grant No. UL1RR024156 from NCATS-NCRR/NIH.
Supported in part by the National Institute of Health (R01DK080153, RO3 DK61519, and DK026687).
References
- 1.Trakhtenbroit MA, Leichman JG, Algahim MF, et al. Body weight, insulin resistance, and serum adipokine levels 2 years after 2 types of bariatric surgery. Am J Med. 2009;122:435–442. doi: 10.1016/j.amjmed.2008.10.035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Gallagher D, Visser M, Sepulveda D, Pierson RN, Harris T, Heymsfield SB. How useful is body mass index for comparison of body fatness across age, sex, and ethnic groups? Am J Epidemiol. 1996;143:228–239. doi: 10.1093/oxfordjournals.aje.a008733. [DOI] [PubMed] [Google Scholar]
- 3.Bergman RN, Stefanovski D, Buchanan TA, et al. A better index of body adiposity. Obesity (Silver Spring) 2011;19:1083–1089. doi: 10.1038/oby.2011.38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Nomura K, Eto M, Kojima T, et al. Visceral fat accumulation and metabolic risk factor clustering in older adults. J Am Geriatr Soc. 2010;58:1658–1663. doi: 10.1111/j.1532-5415.2010.03018.x. [DOI] [PubMed] [Google Scholar]
- 5.Faria AN, Ribeiro Filho FF, Gouveia Ferreira SR, Zanella MT. Impact of visceral fat on blood pressure and insulin sensitivity in hypertensive obese women. Obes Res. 2002;10:1203–1206. doi: 10.1038/oby.2002.164. [DOI] [PubMed] [Google Scholar]
- 6.Phillips GB, Jing T, Heymsfield SB. Relationships in men of sex hormones, insulin, adiposity, and risk factors for myocardial infarction. Metabolism. 2003;52:784–790. doi: 10.1016/s0026-0495(03)00072-6. [DOI] [PubMed] [Google Scholar]
- 7.Ginde SR, Geliebter A, et al. Air displacement plethysmography: validation in overweight and obese subjects. Obes Res. 2005;13(7):1232–1237. doi: 10.1038/oby.2005.146. [DOI] [PubMed] [Google Scholar]
- 8.Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28:412–419. doi: 10.1007/BF00280883. [DOI] [PubMed] [Google Scholar]
- 9.Katz A, Nambi SS, Mather K, et al. Quantitative insulin sensitivity check index: a simple, accurate method for assessing insulin sensitivity in humans. J Clin Endocrinol Metab. 2000;85:2402–2410. doi: 10.1210/jcem.85.7.6661. [DOI] [PubMed] [Google Scholar]
- 10.Savastano S, Belfiore A, Di Somma C, et al. Validity of bioelectrical impedance analysis to estimate body composition changes after bariatric surgery in pre-menopausal morbidly women. Obes Surg. 2010;20:332–339. doi: 10.1007/s11695-009-0006-5. [DOI] [PubMed] [Google Scholar]
- 11.Linares CL, Ciangura C, Bouillot JL, et al. Validity of leg-to-leg bioelectrical impedance analysis to estimate body fat in obesity. Obes Surg. 2011;21:917–923. doi: 10.1007/s11695-010-0296-7. [DOI] [PubMed] [Google Scholar]
- 12.Ferreira I, Snijder MB, Twisk JW, et al. Central fat mass versus peripheral fat and lean mass: opposite (adverse versus favorable) associations with arterial stiffness? The Amsterdam Growth and Health Longitudinal Study. J Clin Endocrinol Metab. 2004;89:2632–2639. doi: 10.1210/jc.2003-031619. [DOI] [PubMed] [Google Scholar]
- 13.Snijder MB, Dekker JM, Visser M, et al. Trunk fat and leg fat have independent and opposite associations with fasting and postload glucose levels: the Hoorn study. Diabetes Care. 2004;27:372–377. doi: 10.2337/diacare.27.2.372. [DOI] [PubMed] [Google Scholar]
- 14.Tanko LB, Bagger YZ, Alexandersen P, Larsen PJ, Christiansen C. Peripheral adiposity exhibits an independent dominant antiatherogenic effect in elderly women. Circulation. 2003;107:1626–1631. doi: 10.1161/01.CIR.0000057974.74060.68. [DOI] [PubMed] [Google Scholar]
- 15.Boozer CN, Leibel RL, Love RJ, Cha MC, Aronne LJ. Synergy of sibutramine and low-dose leptin in treatment of diet-induced obesity in rats. Metabolism. 2001;50:889–893. doi: 10.1053/meta.2001.24917. [DOI] [PubMed] [Google Scholar]
- 16.Rosenbaum M, Sy M, Pavlovich K, Leibel RL, Hirsch J. Leptin reverses weight loss-induced changes in regional neural activity responses to visual food stimuli. J Clin Invest. 2008;118:2583–2591. doi: 10.1172/JCI35055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Barreira TV, Harrington DM, Staiano AE, Heymsfield SB, Katzmarzyk PT. Body adiposity index, body mass index, and body fat in white and black adults. JAMA. 2011;306:828–830. doi: 10.1001/jama.2011.1189. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Schulze MB, Stefan N. The body adiposity index and the sexual dimorphism in body fat. Obesity (Silver Spring) 2011;19:1729. doi: 10.1038/oby.2011.153. [DOI] [PubMed] [Google Scholar]
- 19.Schulze MB, Thorand B, Fritsche A, et al. Body adiposity index, body fat content and incidence of type 2 diabetes. Diabetologia. 2012 doi: 10.1007/s00125-012-2499-z. [DOI] [PubMed] [Google Scholar]