Increased Intramyocellular Lipid Concentration Identifies Impaired Glucose Metabolism in Women With Previous Gestational Diabetes Alexandra Kautzky-Willer,1 Martin Krssak,1 Christine Winzer,1 Giovanni Pacini,2 Andrea Tura,2 Serdar Farhan,1 Oswald Wagner,3 Georg Brabant,4 Ru¨ diger Horn,4 Harald Stingl,1 Barbara Schneider,5 Werner Waldha¨usl,1 and Michael Roden1

Women with previous gestational diabetes (pGDM) are frequently insulin-resistant, which could relate to in- tramyocellular lipid content (IMCL). IMCL were mea- estational diabetes mellitus (GDM) is a fre- sured with 1H nuclear magnetic resonance spectroscopy quent metabolic complication during preg- in soleus (IMCL-S) and tibialis-anterior muscles nancy that does not completely normalize after -IMCL-T) of 39 pGDM (32 ؎ 2 years, waist-to-hip ratio Gdelivery (1–3). Women with a history of previ) -and 22 women with normal glucose toler- ous gestational diabetes (pGDM) are often insulin-resis (0.01 ؎ 0.81 ance (NGT; 31 ؎ 1 years, 0.76 ؎ 0.02) at 4–6 months tant and exhibit markedly increased risk for the later after delivery. Body fat mass (BFM) was assessed from development of type 2 diabetes (4,5). The most prominent bioimpedance analysis, insulin sensitivity index (SI), parameters that predict type 2 diabetes in later life are the and glucose effectiveness (SG) from insulin-modified need for insulin in addition to diet therapy to achieve frequently sampled glucose tolerance tests. pGDM ex- normoglycemia, early diagnosis of GDM during pregnancy, hibited 45% increased BFM, 35% reduced SI and SG (P < and maternal BMI and plasma glucose during the oral 0.05), and 40% (P < 0.05) and 55% (P < 0.005) higher glucose tolerance test (OGTT) at diagnosis as well as at IMCL-S and IMCL-T, respectively. IMCL related to body the first postpartum assessment (4,6,7). fat (BFM P < 0.005, leptin P < 0.03), but only IMCL-T Skeletal muscle insulin resistance is a key feature of the correlated (P < 0.03) with SI and glucose tolerance metabolic syndrome and predisposes to type 2 diabetes ؍ index independent of BMI. Insulin-resistant pGDM (n and premature cardiovascular complications (8). Although (had higher IMCL-S (؉66%) and IMCL-T (؉86% (17 lifestyle (9), obesity, and increased lipid supply play an than NGT and insulin-sensitive pGDM (؉28%). IMCL important role in this disease (8,10), the hierarchy of -in insulin (0.05 ؍ were also higher (P < 0.005, P sensitive pGDM requiring insulin treatment during events is still unclear. It was postulated that muscle fat content could contribute to insulin resistance and glucose pregnancy and inversely related to the gestational week 1 of GDM diagnosis. Thus, IMCL-T reflects insulin sensi- intolerance (11–19), but only the advent of H nuclear tivity, whereas IMCL-S relates to obesity. IMCL could magnetic resonance spectroscopy (NMRS) made it possi- serve as an additional parameter of increased diabetes ble to quantify and distinguish between extramyocellular risk because it identifies insulin-resistant pGDM and and intramyocellular lipid contents (IMCL) (11,14,15,17, those who were diagnosed earlier and/or required insu- 18,20–22). lin during pregnancy. Diabetes 52:244–251, 2003 We tested the hypotheses that intracellular fat content in different muscles diversely relates to insulin sensitivity and correlates with established risk markers for type 2 diabetes in pGDM, such as gestational week at diagnosis, insulin treatment during pregnancy, glucose levels during From the 1Department of Internal Medicine III, Division of Endocrinology and OGTT at diagnosis and postpartum, and the degree of Metabolism, University of Vienna, Vienna, Austria; 2Metabolic Unit, Institute 1 of Biomedical Engineering, National Research Council (ISIB-CNR), Padova, obesity. Thus, we applied H NMRS to measure rapidly and Italy; 3Institute for Medical Laboratory Diagnostics, University of Vienna, noninvasively IMCL in soleus (IMCL-S) and tibialis ante- Vienna, Austria; 4Division of Endocrinology, University of Hannover, Han- rior muscles (IMCL-T) in pGDM. IMCL were correlated nover, Germany; and 5Institute of Biostatistics, University of Vienna, Vienna, Austria. with parameters of glucose tolerance, insulin sensitivity, Address correspondence and reprint requests to Michael Roden, MD, cardiovascular risk, body fat content, and distribution. Division of Endocrinology and Metabolism, Department of Internal Medicine Furthermore, the study extends to potential links between III, University of Vienna, Wa¨hringer Gu¨ rtel 18-20, A-1090, Vienna, Austria. E-mail: [email protected]. IMCL and the leptin system, which participates in the Received for publication 25 June 2002 and accepted in revised form 16 regulation of body weight (BW) and energy metabolism October 2002. AIRg 3–10, acute insulin response 3–10 min after glucose ingestion; BFM, (19,23,25,26). body fat mass; BW, body weight; FFM, fat-free mass; GDM, gestational diabetes mellitus; IMCL, intramyocellular lipid content; IMCL-S, IMCL of soleus muscle; IMCL-T, IMCL of tibialis anterior; NGT, normal glucose RESEARCH DESIGN AND METHODS tolerance; NMRS, nuclear magnetic resonance spectroscopy; OGIS, insulin sensitivity parameter; OGTT, oral glucose tolerance test; pGDM, previous All women ingested an isocaloric diet containing 200 g of carbohydrates/day gestational diabetes mellitus; SI, insulin sensitivity index. and refrained from exercise for at least 3 days before the studies. Metabolic

244 DIABETES, VOL. 52, FEBRUARY 2003 A. KAUTZKY-WILLER AND ASSOCIATES

TABLE 1 Clinical characteristics of the total group of women with pGDM (n ϭ 39), the insulin-resistant subgroup (GDM-R, n ϭ 17), and the insulin-sensitive subgroup (GDM-S, n ϭ 22) compared with women with NGT during pregnancy (n ϭ 23) 4–6 months after delivery GDM GDM-R GDM-S NGT Age (years) 31.1 Ϯ 0.81 31.0 Ϯ 1.4 31.2 Ϯ 1.0 30.6 Ϯ 1.3 BMI (kg/m2) 26.4 Ϯ 1.1 29.8 Ϯ 1.8 24.9 Ϯ 0.8* 24.3 Ϯ 0.9† WHR 0.81 Ϯ 0.01 0.81 Ϯ 0.01 0.80 Ϯ 0.01 0.76 Ϯ 0.02 Waist (cm) 89.1 Ϯ 2.3§ 96.1 Ϯ 2.5 84.5 Ϯ 2.2* 75.1 Ϯ 2.3†‡ Triglycerides (mg/dl) 118.2 Ϯ 22.0 136.3 Ϯ 50.9 105.9 Ϯ 14.3 75.2 Ϯ 6.3 Cholesterol (mg/dl) 210.9 Ϯ 7.6 200.3 Ϯ 14.1 218.0 Ϯ 8.3 198.0 Ϯ 12.4 HDL cholesterol (mg/dl) 61.4 Ϯ 2.7 54.8 Ϯ 4.7 65.8 Ϯ 3.9* 62.1 Ϯ 2.9† LDL cholesterol (mg/dl) 126.2 Ϯ 6.5 118.9 Ϯ 10.9 131.2 Ϯ 7.9 120.8 Ϯ 11.0 Systolic blood pressure (mmHg) 116.7 Ϯ 2.2 123.7 Ϯ 3.9 112.3 Ϯ 2.3* 111.5 Ϯ 2.6† Diastolic blood pressure (mmHg) 80.0 Ϯ 1.5§ 83.0 Ϯ 2.5 78.1 Ϯ 1.9 74.1 Ϯ 1.9† Ϯ Ϯ Ϯ Ϯ HbA1c (%) 5.40 0.07§ 5.36 0.14 5.43 0.08 5.1 0.03† Basal metabolic rate (kJ/kg BW) 20.10 Ϯ 0.55§ 18.35 Ϯ 0.69 21.75 Ϯ 0.65* 22.03 Ϯ 0.51† Basal metabolic rate (kJ/kg FFM) 30.50 Ϯ 0.33 29.35 Ϯ 0.50 31.58 Ϯ 0.42* 30.57 Ϯ 0.39† *P Ͻ 0.05 GDM-R versus GDM-S; †P Ͻ 0.05 GDM-R versus NGT; ‡P Ͻ 0.05 GDM-S versus NGT; §P Ͻ 0.05 GDM versus NGT. tests were performed on different days during the first phase (days 5–8) of the sensitivity from OGTT (OGIS) was derived as glucose clearance (ml ⅐ minϪ1 ⅐ menstrual cycle after 10- to 12-h overnight fasting. mϪ2) (31). Study participants. Cross-sectional analysis was performed in 39 pGDM Localized 1H NMRS. IMCL was measured with localized 1H NMRS (17,20,21) women at 4–6 months after delivery. They were recruited from our division’s on a 3.0-T/80-cm NMR spectrometer (Medspec; Bruker, Ettlingen, FRG) outpatient service, where they had been seen previously during pregnancy. equipped with a whole-body gradient coil (40 mT/m; Fig. 1). A standard GDM had been diagnosed according to the criteria of the 4th Workshop birdcage 1H coil (inner diameter 25 cm) was used in the transmission/ Conference of Gestational Diabetes (27). During pregnancy, 26 women were reception mode. The STEAM sequence (echo time 20 ms; mixing time 30 ms; treated with diet plus insulin, because blood glucose exceeded 95 mg/dl at relaxation time 6 s; number of scans 32) was complemented by CHESS water fasting and/or 130 mg/dl at 60 min postprandially. A total of 22 age-matched suppression and applied on the 1.73-cm3 volume of interest, which was placed women without any risk for diabetes and with normal glucose tolerance in the soleus or tibialis anterior muscles of the subject’s right leg. Spectra were during pregnancy served as a control group (NGT). All subjects gave written line-broadened and -fitted using the MacNUTS-PPC software (Acorn NMR, informed consent for participation in the study, which was approved by the Livermore, CA). IMCL was quantified from processed spectra after T2- local ethics committee. relaxation correction as a ratio of the intensity of (CH2)n (1.25 ppm) group Patients with previous ketoacidosis and/or ␤-cell antibodies (GAD, ICA, resonance to the intensity of the water resonance from non–water-suppressed IA2) were excluded. The relationship among IMCL, insulin sensitivity, and spectra of the same volume of interest (Fig. 1). The T2 relaxation times were metabolic parameters was analyzed both in the total pGDM population and for of 82 Ϯ 3 ms for IMCL-S and 30 Ϯ 2 ms for water in soleus muscle and 90 Ϯ the insulin-sensitive (pGDM-S) and insulin-resistant subgroups (pGDM-R), which were separated by a cutoff value of 2.8 10Ϫ4 minϪ1/(␮U/ml) for the insulin sensitivity index (SI). This value was derived from analyzing SI of the NGT of the present study plus another 26 matched female control subjects from other studies. The lower 2.5% quantile of the distribution gave an SI of 2.79 10Ϫ4 minϪ1/(␮U/ml), which was defined as the cutoff point between normal and impaired (lower) SI. All pGDM had higher waist circumference, diastolic blood pressure, HbA1c, and basal metabolic rate corrected for BW than NGT (Table 1). After correction for BMI, differences in waist-to-hip-ratio disappeared, whereas waist circumference remained different (P Ͻ 0.01). Clinical characteristics were not different between NGT and pGDM-S except for HbA1c (Table 1). In contrast, pGDM-R had higher BMI and systolic blood pressure but lower HDL cholesterol. Basal metabolic rate adjusted for BW was lowest in pGDM-R (Ϫ16% versus NGT and pGDM-S) but not different between pGDM-S and NGT. After correction for lean body mass, i.e., fat-free mass (FFM), the basal metabolic rate was not different between pGDM (30.50 Ϯ 0.33 kJ/kg FFM) and NGT (30.57 Ϯ 0.39 kJ/kg FFM) but lower in pGDM-R (29.35 Ϯ 0.50 kJ/kg FFM) than pGDM-S (31.58 Ϯ 0.42 kJ/kg FFM, P Ͻ 0.003) and NGT (P Ͻ 0.03). Frequently sampled intravenous glucose tolerance test. Glucose (time 0–0.5 min: 300 mg/kg BW) and then normal insulin (time 20–25 min: 0.03 IU/kg, Humulin R; Eli Lilly, Indianapolis, IN) were infused intravenously, and venous blood samples were taken in timed intervals (28). Analysis of glucose and insulin concentrations provided indexes for glucose tolerance (KG), insulin sensitivity (SI), and glucose effectiveness (SG), which describe glucose disposal and the insulin effect on glucose disappearance (28). Insulin secre- ⌬ tion was assessed from incremental short-term insulin response ( AIRG) calculated by averaging insulin concentration above basal from times 3–10 ⌬ min. The disposition index was calculated as SI times AIRG and gives a measure of the combined effects of insulin secretion and sensitivity on glucose disposal (29). OGTT. Participants ingested 75 g of glucose solution, and venous blood samples were collected for glucose, insulin, and C-peptide measurements at FIG. 1. Left: Magnetic resonance cross-sectional image of human calf timed intervals (30). Modeling analysis yielded fasting prehepatic insulin muscle. The rectangles indicate the positioning of the volumes of secretion rate and the total amount of insulin per unit volume released during interest in soleus and tibialis anterior muscles. Right: 1H NMR spectra the OGTT in response to increments in glucose concentration (30). Insulin acquired from the volumes of interest in both muscles.

DIABETES, VOL. 52, FEBRUARY 2003 245 MUSCLE LIPIDS AND GESTATIONAL DIABETES

TABLE 2 Metabolic parameters of the total group of women with pGDM (n ϭ 39), the insulin-resistant subgroup (GDM-R), and the insulin-sensitive subgroup (GDM-S) compared with women with NGT 4–6 months after delivery GDM GDM-R GDM-S NGT FSIGT Insulin sensitivity index [10Ϫ4 minϪ1 (␮U/ml)Ϫ1] 4.00 Ϯ 0.35§ 1.92 Ϯ 0.10 5.39 Ϯ 0.36* 6.1 Ϯ 0.5† Glucose effectiveness (minϪ1) 0.022 Ϯ 0.007§ 0.021 Ϯ 0.001 0.023 Ϯ 0.006 0.027 Ϯ 0.001† Disposition index (10Ϫ2 minϪ1) 0.112 Ϯ 0.011§ 0.094 Ϯ 0.020 0.124 Ϯ 0.015 0.20 Ϯ 0.03† Ϫ1 Ϯ Ϯ Ϯ Ϯ Glucose tolerance index (% min )10–20min 1.85 0.13 1.90 0.25 1.79 0.15 2.54 0.29‡ Ϫ1 Ϯ Ϯ Ϯ Ϯ AIRg (pmol l )3–10min 38.1 4.7 53.5 9.6 27.4 3.4* 35.3 4.5 OGTT Fasting glucose (mg/dl) 91.1 Ϯ 1.4§ 93.1 Ϯ 1.9 90.0 Ϯ 1.7 77.1 Ϯ 1.2†‡ 1 h plasma glucose (mg/dl) 151.3 Ϯ 7.3§ 159.1 Ϯ 10.1 145.1 Ϯ 10.4 102.1 Ϯ 4.4†‡ 2 h plasma glucose (mg/dl) 119.1 Ϯ 5.8§ 121.5 Ϯ 9.0 117.3 Ϯ 7.9 90.5 Ϯ 3.0†‡ OGIS (ml ⅐ minϪ1 ⅐ m2) 423.5 Ϯ 11.5§ 385.4 Ϯ 16.5 454.8 Ϯ 11.4* 502.4 Ϯ 12.6†‡ Basal secretion rate (pmol ⅐ lϪ1 ⅐ minϪ1) 22.8 Ϯ 2.4§ 27.6 Ϯ 4.9 18.8 Ϯ 1.6 14.5 Ϯ 1.1† Dynamic insulin secretion (nmol/l) 51.7 Ϯ 10.3 73.4 Ϯ 20.0 31.83 Ϯ 2.51* 33.6 Ϯ 2.6† *P Ͻ 0.05 GDM-R versus GDM-S; †P Ͻ 0.05 GDM-R versus NGT; ‡P Ͻ 0.05 GDM-S versus NGT; §P Ͻ 0.05 GDM versus NGT.

6 ms for IMCL-T and 27 Ϯ 1 ms for water in tibialis anterior muscle. The reflecting reduced insulin release in response to a given 1 coefficients of variation for H NMRS of IMCL, as assessed from three glucose load. In contrast, pGDM-R featured 50% higher measurements of the lipid content in each muscle in six healthy young ⌬ subjects, were 0.3% for soleus muscle and 1.3% for tibialis anterior muscle, AIRG than in NGT, indicating compensation by the respectively. ␤-cells for their insulin resistance. Body fat mass and basal metabolic rate. Body fat mass (BFM) was Although fasting and postprandial plasma glucose con- assessed from bioimpedance measurements (Akern-RJL Systems, Florence, centrations during OGTT were higher in pGDM, no woman Italy) as was the basal metabolic rate. Measurement of resting energy expenditure with bioimpedance was validated in 10 subjects against indirect was diabetic according to American Diabetes Association calorimetry yielding comparable results (r ϭ 0.78, P Ͻ 0.01). Resting energy and World Health Organization criteria (Table 2). The expenditure is expressed per kg BW as well as per kg FFM. insulin sensitivity index (OGIS) was reduced by 10 and Metabolites and hormones. Plasma glucose was measured using an auto- 25% in pGDM-S and pGDM-R, respectively. The pGDM-R mated glucose analyzer (Beckman, Fullerton, CA). HbA1c (upper limit of normal range 5.8%) was quantified by online high-pressure liquid chromatog- compensated their marked insulin resistance by doubling raphy (C-R4A Chromatopac; Shimadzu, Kyoto, Japan) from capillary blood. basal and dynamic insulin secretion, which in contrast Cholesterol and triglyceride concentrations were measured by lipid gel were not different between pGDM-S and NGT. Correction electrophoresis. Insulin (Serono Diagnostics, Freiburg, FRG), C-peptide (CIS for BMI or BFM as a result of the higher degree of obesity Bio International, Cedex, France), proinsulin, and total leptin (both from in pGDM-R did not affect the differences in insulin sensi- Linco, St. Charles, MO) were quantified in duplicate by radioimmunoassays Ͻ Ͻ with interassay coefficients of variation of Ͻ5.5% for insulin, C-peptide, leptin tivity (P 0.001) and glucose effectiveness (P 0.05) and Ͻ8% for proinsulin. Plasma bound leptin and soluble leptin receptor between groups. Fasting plasma FFA were higher in both concentrations were measured by specific radioimmunoassay (25,26,32). the total group of pGDM (0.63 Ϯ 0.04 ␮mol/l, P Ͻ 0.005) Fasting plasma free fatty acids (FFA) were measured as described (33) and and the subgroups (P Ͻ 0.01) pGDM-R (0.61 Ϯ 0.09 available in most pGDM (n ϭ 30) and NGT (n ϭ 17). Ϯ Statistical analysis. Data are presented as means Ϯ SE. Relationships were mmol/l) and pGDM-S (0.60 0.05 mmol/l) than in NGT tested for statistical significance by linear regression analysis and Spearman (0.38 Ϯ 0.03 mmol/l) but did not differ (P ϭ 0.07) between correlation coefficient. Groups were compared with nonparametric ANOVA, subgroups. and multiple test statistics were used for subgroup analyses comparing NGT, IMCL. Mean IMCL-S and IMCL-T were 31 and 61% higher pGDM-R, and pGDM-S (SAS package, version 8.2, Tukey-Kramer). ANOVA in the total group of pGDM than in NGT (Table 3). In the was also performed including interaction between groups and BMI or BFM. In the absence of significant interaction, groups were also compared after subgroup pGDM-R, both IMCL were also higher (IMCL-S correction for BMI or BFM. Multiple regression analysis was computed to 66%; IMCL-T 86%) than in NGT and one third higher in identify independent regulators of IMCL. pGDM-R than in pGDM-S. Also, BFM was 78 and 56% One woman in the pGDM-R subgroup had an IMCL-T value within the range higher in GDM-R than in NGT and GDM-S, respectively of the other pGDM, whereas her IMCL-S value of 7.7% H2O resonance exceeded the interquartile range (five times above the 75% quartile of all (Table 3). After correction for either BMI or BFM, the pGDM-R, more than seven times above that of the total population) and was differences in IMCL-S disappeared (P ϭ 0.10 for BMI, P ϭ therefore excluded from analysis. Nevertheless, inclusion or exclusion of this 0.61 for BFM), whereas IMCL-T remained increased in outlier did not affect the statistically significant differences. pGDM-R (P ϭ 0.01 for BMI, P ϭ 0.04 for BFM), indicating that obesity strongly influences only IMCL-S. RESULTS Nearly all pGDM-R (n ϭ 15 of 17; 88%) but only 50% of ϭ Glucose metabolism. All pGDM featured lower SI GDM-S (n 11 of 22) had required insulin therapy during Ϫ Ϫ ( 35%), glucose effectiveness (SG 25%), and disposition pregnancy. The pGDM-S treated previously with insulin index, a marker of overall glucose homeostasis (Ϫ47%), presented with higher IMCL-S (P Ͻ 0.005) and IMCL-T ϭ than NGT (Table 2). In the subgroup, pGDM-R, SI was even (P 0.05) than insulin-naive patients who were treated 70 and 65% lower than in NGT and pGDM-S, respectively with diet despite no difference in metabolic or anthropo- (Table 2). The disposition index was lower in pGDM-R metric parameters (Fig. 2). than in NGT but not different between pGDM subgroups. Leptin system. Plasma total leptin was highest in Ͼ The glucose tolerance index (KG) was 30% lower in all pGDM-R (Table 3). The fraction of bound leptin was not ⌬ pGDM, whereas AIRG was 22% lower only in pGDM-S, different between pGDM-R and NGT but was 50% lower in

246 DIABETES, VOL. 52, FEBRUARY 2003 A. KAUTZKY-WILLER AND ASSOCIATES

TABLE 3 IMCL-S and IMCL-T: BFM and parameters of the leptin system in the total group of women with prior GDM, in the insulin-resistant subgroup (GDM-R) and the insulin-sensitive subgroup (GDM-S) compared with women with during pregnancy NGT 4–6 months after delivery GDM GDM-R GDM-S NGT IMCL-S (% water resonance) 1.70 Ϯ 0.18§ 2.00 Ϯ 0.38 1.51 Ϯ 0.13 1.21 Ϯ 0.08† IMCL-T (% water resonance) 0.66 Ϯ 0.06§ 0.78 Ϯ 0.09 0.58 Ϯ 0.05* 0.41 Ϯ 0.04† BFM (kg) 26.4 Ϯ 2.1§ 32.5 Ϯ 3.4 21.3 Ϯ 1.8* 18.1 Ϯ 1.4† Fasting leptin (nmol/l) 16.7 Ϯ 1.5§ 19.6 Ϯ 2.7 14.3 Ϯ 1.6 9.0 Ϯ 1.3† Bound leptin (nmol/l) 0.60 Ϯ 0.09§ 0.80 Ϯ 0.22 0.48 Ϯ 0.04 1.07 Ϯ 0.13‡ Soluble leptin receptor (nmol/l) 3.76 Ϯ 0.28§ 3.77 Ϯ 0.46 3.75 Ϯ 0.36 5.93 Ϯ 0.61†‡ Fasting leptin/BFM (nmol ⅐ lϪ1 ⅐ kgϪ1) 0.6 Ϯ 0.03§ 0.62 Ϯ 0.06 0.66 Ϯ 0.05 0.48 Ϯ 0.05‡ Bound leptin/BFM (nmol ⅐ lϪ1 ⅐ kgϪ1) 0.028 Ϯ 0.005§ 0.029 Ϯ 0.01 0.026 Ϯ 0.004 0.06 Ϯ 0.008†‡ Soluble leptin receptor/BFM (nmol ⅐ lϪ1 ⅐ kgϪ1) 0.17 Ϯ 0.02§ 0.14 Ϯ 0.02 0.20 Ϯ 0.03 0.36 Ϯ 0.05†‡ *P Ͻ 0.05 GDM-R versus GDM-S; †P Ͻ 0.05 GDM-R versus NGT; ‡P Ͻ 0.05 GDM-S versus NGT; §P Ͻ 0.05 GDM versus NGT. pGDM-S. However, when adjusted for BFM (Table 3) and GDM was diagnosed (IMCL-S r ϭϪ0.30, P Ͻ 0.05; IMCL-T corrected for BMI, bound leptin became markedly (P Ͻ r ϭϪ0.44, P Ͻ 0.02), but only IMCL-T also negatively 0.001) different, being lower in both GDM subgroups than related to systolic blood pressure and negatively corre- ϭϪ Ͻ in NGT. Plasma concentrations of soluble plasma leptin lated with SI,SG, and KG (r 0.31, P 0.03; Fig. 4). In receptor were decreased in pGDM independent of BMI, pGDM-R, IMCL-T negatively related to the disposition whereas total leptin was no more different between groups index derived from both intravenous (r ϭ 0.62, P Ͻ 0.02) after correction for BMI. and oral glucose challenge (r ϭϪ0.67, P Ͻ 0.014; Spear- Correlation analysis. In all women, IMCL correlated man, ANOVA). Multiple regression analysis including SI, positively with BMI and BFM and negatively with the basal BMI, and 2-h plasma glucose as explanatory variables metabolic rate after adjustment for BW (Fig. 3). However, showed that for IMCL-S only BMI and for IMCL-T only the when expressed per FFM, basal metabolic rate did not 2-h plasma glucose had an independent influence. With ϭ relate to IMCL-S or IMCL-T (P 0.09) but still to IMCL-T only BMI and SI in the model, SI remained as the only in pGDM-R (r ϭϪ0.55, P Ͻ 0.04). IMCL was also associ- significant independent parameter to explain IMCL-T. ated with 2-h plasma glucose during OGTT (Fig. 4), waist When BFM entered the multiple regression analysis in- circumference (r ϭ 0.4, P Ͻ 0.001), plasma total leptin stead of BMI, again only BFM influenced IMCL-S (P Ͻ (r ϭ 0.32, P Ͻ 0.03), and plasma FFA (IMCL-S: r ϭ 0.31, 0.02), and 2-h plasma glucose (P Ͻ 0.04) and BFM (P Ͻ Ͻ ϭ Ͻ P 0.03; IMCL-T: r 0.55, P 0.001). In all women, IMCL 0.05) influenced IMCL-T. With only BFM and SI in the also inversely related to the gestational week during which model, BFM still significantly affected IMCL-S (P Ͻ 0.01), whereas no parameter independently affected IMCL-T. Linear regression analysis for plasma leptin revealed its correlation to BMI (r ϭ 0.70, P Ͻ 0.0001), waist-to-hip ratio (r ϭ 0.40, P Ͻ 0.003), 2-h glucose during OGTT (r ϭ 0.45, P Ͻ 0.002), systolic and diastolic blood pressures (r ϭ 0.38, P Ͻ 0.009), triglycerides (r ϭ 0.30, P Ͻ 0.05), HDL cholesterol (r ϭϪ0.31, P Ͻ 0.03), insulin sensitivity (r ϭ Ϫ0.30 P Ͻ 0.04), and insulin secretion (r ϭ 0.38, P Ͻ 0.009). In contrast, bound leptin correlated inversely only with 2-h glucose (r ϭϪ0.30, P Ͻ 0.03) and positively with the soluble leptin receptor (r ϭ 0.53, P Ͻ 0.005) as well as parameters of glucose effectiveness (r ϭ 0.33, P Ͻ 0.01) and insulin sensitivity (r ϭ 0.43, P Ͻ 0.001). Soluble leptin receptor concentrations were inversely related to IMCL-T (Ϫ0.43, P Ͻ 0.05) and positively to HDL cholesterol (r ϭ 0.57, P Ͻ 0.02).

TABLE 4 Correlations of IMCL-T with modeling parameters (Spearman correlation coefficient, ANOVA) Spearman P values FIG. 2. Box plots for IMCL-S and IMCL-T in insulin-sensitive women GDM-R (11 ؍ or insulin (n (11 ؍ who had pGDM-S and were treated by diet (n during pregnancy. The median is represented by the horizontal line OGTT disposition index (nmol/m3) Ϫ0.67 0.014 inside the box; the top and bottom of the box represent the third Ϫ quartile (75th percentile) and the first quartile (25th percentile), FSIGT disposition index 0.62 0.028 respectively. Whiskers are drawn from the edge of the box to the GDM-S farthest observation within 1.5 times the interquartile range of the OGTT disposition index (nmol/m3) Ϫ0.007 0.98 edge of the box. Observations beyond the whiskers are individually FSIGT disposition index Ϫ0.14 0.56 identified (•).

DIABETES, VOL. 52, FEBRUARY 2003 247 MUSCLE LIPIDS AND GESTATIONAL DIABETES

FIG. 3. Relationship between BMI, fat mass (FAT), and basal metabolic rate adjusted for BW and IMCL-S and IMCL-T in all women by linear regression analysis.

DISCUSSION women with characteristic features of the metabolic syn- This study confirms that the disturbances in glucose drome also presented with highest IMCL. Moreover, even metabolism persist in women with pGDM. Of the present the 50% of pGDM-S on previous insulin therapy had higher population, 44% were severely insulin-resistant (pGDM-R) IMCL than the insulin-naive pGDM-S whose IMCL were with excessive insulin secretion and higher degree of almost identical to those of NGT. As increased IMCL was abdominal fat as indicated by their higher waist circum- the only parameter that differed in these pGDM-S sub- ference and waist-to-hip ratio than the other groups. groups, it might be an important marker to identify even However, 56% were insulin-sensitive (pGDM-S) but women who are lean and otherwise insulin-sensitive but at showed distinct abnormalities of early and dynamic insu- increased risk for type 2 diabetes. Furthermore, in all lin secretion during intravenous and oral glucose chal- pGDM, independent of their insulin sensitivity, the higher lenge. Both metabolic abnormalities resulted in similarly IMCL was related to earlier diagnosis of GDM, another higher plasma glucose in both pGDM subgroups than in widely known risk factor for deterioration of glucose NGT. tolerance at follow-up (6). Although parameters of long-term glucose control such IMCL was higher in GDM-R than in the other groups and as plasma glucose during OGTT and HbA1c were compa- correlated with body fat, insulin sensitivity, and cardiovas- rable at 4–6 months after delivery, almost all pGDM-R and cular risk parameters in all women. Thus, increased IMCL half of the lean pGDM-S had required insulin therapy may be added to the characteristics of the metabolic during pregnancy. Glucose control was similar throughout syndrome. Previous studies reported divergent results on pregnancy between pGDM-S and pGDM-R as reflected by the relationship between IMCL and anthropometric pa- HbA1c and self-recorded daily pre- and postprandial glu- rameters (34,35), which possibly results from differences cose profiles. As the requirement of insulin therapy indi- between individual skeletal muscle groups. cates a more severe derangement of glucose metabolism We found that IMCL-S more strongly relate to measures and is considered an important risk factor for the progres- of obesity, whereas IMCL-T are more tightly associated sion to overt diabetes, it seems that pGDM-R are at higher with insulin resistance per se. This is in line with one risk of type 2 diabetes than pGDM-S. It is of note that these report (11) but partly in contrast to another study report-

248 DIABETES, VOL. 52, FEBRUARY 2003 A. KAUTZKY-WILLER AND ASSOCIATES

FIG. 4. Relationship between 2-h plasma glucose concentra- tions during the OGTT (OGTT2 h), the insulin sensitivity

index (SI) from frequently sampled insulin modified intrave- nous glucose tolerance tests (FSIGT) and the systolic blood pressure and intramyocellular lipid content in soleus muscle (IMCL-S) and in tibialis anterior (IMCL-T) in all women by linear regression analysis. ing a better correlation of insulin resistance with IMCL-S be channeled preferentially into triglycerides (8, 10, 19, 42, and increased IMCL-T only in women (20). Although the 43). Increased FFA uptake or lipolysis of IMCL would correlation of IMCL-S with the degree of obesity might increase cytosolic long-chain acyl-CoA (LCA-CoA), which suggest that increased intracellular muscle fat simply correlate well with insulin resistance (43) and can inhibit reflects whole-body adiposity, the correlation of IMCL-T insulin action via decreased insulin-receptor substrate-1 with parameters of the insulin resistance syndrome was phosphorylation (42,43). In line with this hypothesis, found to be independent of BMI. In untrained humans, plasma FFA elevation induces insulin resistance and gives skeletal muscle represents a mixed fiber type (36) contain- rise to IMCL under high (21,44) but not fasting insulin ing fiber types of different insulin sensitivity (37). Soleus conditions (33). Likewise, improvement of insulin resis- muscle is prevalently composed of slow-twitch oxidative tance by the thiazolidinedione pioglitazone reduces mus- type I fibers, whereas tibialis anterior muscle contains cle LCA-CoA and lipid accumulation in high-fat–fed rats more fast-twitch glycolytic type IIb fibers. Type I fibers (45) so that the beneficial effect of troglitazone to improve generally have a higher lipid content yet also higher insulin action and reduce progression to type 2 diabetes in oxidative enzyme capacity and greater insulin sensitivity high-risk Latino women with pGDM (46) could at least than type IIb fibers (38). In obesity and type 2 diabetes, the partly result from altered muscle lipid supply. proportion of type IIb fibers with reduced oxidative en- Alternatively, the correlation between IMCL-T and 2-h zyme activity may increase (39,40) so that a dynamic plasma glucose during OGTT hints at an interaction be- interaction between fiber type and metabolic capacity with tween postprandial hyperglycemia and IMCL. Kelley et al. more lipid stored in relation to oxidative capacity can be (47) proposed the concept that provision of glucose inhib- postulated for metabolic disorders. Furthermore, reduc- its lipid oxidation, which could contribute to pathogenesis tion of type I fibers with decreased expression of insulin of lipid accumulation in obesity. In contrast to the present sensitive glucose transporters (GLUT4) was detected in study, Phillips et al. (15) detected no correlation between type 2 diabetes (41). 2-h plasma glucose and IMCL as measured from muscle At present, a cause–effect relationship for IMCL and biopsies of women at a mean age of 52 years. Cross- insulin sensitivity is not clear and the mechanisms of sectional studies in other populations (11,12,20,22) found interaction are not yet fully understood. Plasma FFA no significant correlation for fasting plasma glucose and resulting from dietary fat supply and/or increased lipolysis IMCL. We have previously reported that short-term hyper- in fat tissue may directly induce insulin resistance or could glycemia for2hinthepresence of fasting (peripheral)

DIABETES, VOL. 52, FEBRUARY 2003 249 MUSCLE LIPIDS AND GESTATIONAL DIABETES insulinemia increases glucose disposal without changes in mellitus and gestational diabetes mellitus: same disease, another name? IMCL-S (33). Nevertheless, it cannot be excluded that Diabetes Rev 3:566–583, 1996 6. Buchanan TA, Xiang A, Kjos S, Lee WP, Trigo E, Nader I, Bergner A, Palmer chronic postprandial increases in plasma glucose may JP, Peters RK: Gestational diabetes: antepartum characteristics that pre- contribute to IMCL accumulation in pGDM. dict postpartum glucose intolerance and type 2 diabetes in Latino women. The close association between IMCL and elevated Diabetes 47:1302–1310, 1998 plasma total leptin concentrations correlating with insulin 7. Metzger BE, Cho NH, Roston SM, Radvany R: Prepregnancy weight and secretion, insulin resistance, and BFM in pGDM could antepartum insulin secretion predict glucose tolerance five years after point at increased weight retention postpartum (48,49) gestational diabetes mellitus. Diabetes Care 16:1598–1605, 1993 8. Shulman GI: Cellular mechanisms of insulin resistance. J Clin Invest and risk for later type 2 diabetes (50,51). In the present 106:171–176, 2000 study, increased total leptin was mostly explained by 9. Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, increased free leptin, because bound leptin and soluble Walker EA, Nathan DM: Reduction in the incidence of type 2 diabetes with leptin receptor were even lower in pGDM. Of note, the lifestyle intervention or metformin. N Engl J Med 346:393–403, 2002 bound form was positively associated with glucose dis- 10. Roden M: Non-invasive studies of glycogen metabolism in human skeletal muscle using nuclear magnetic resonance spectroscopy. Curr Opin Clin posal and tolerance, whereas the soluble leptin receptor Nutr Metab Care 4:261–266, 2001 was inversely related to IMCL-T. The soluble leptin recep- 11. Perseghin G, Scifo P, DeCobelli F, Pagliato E, Battezzati A, Arcelloni C, tor represents the major leptin binding protein in human Vanzulli A, Testolin G, Pozza G, DelMaschio A, Luzi L: Intramyocellular blood and may therefore determine the circulating amount triglyceride content is a determinant of in vivo insulin resistance in 1 13 of total leptin (26,52). It was postulated that the bound humans. A H- C nuclear magnetic resonance spectroscopy assessment in offspring of type 2 diabetic parents. Diabetes 48:1600–1606, 1999 form is involved in the regulation of energy expenditure, 12. Jacob S, Machann J, Rett K, Brechtel K, Volk A, Renn W, Maerker E, whereas the free form simply reflects the degree of adi- Matthaei S, Schick F, Claussen CD, Ha¨ring HU: Association of increased posity (25,32). The present study supports this hypothesis, intramyocellular lipid content with insulin resistance in lean nondiabetic because the degree of adiposity explains variations in offspring of type 2 diabetic subjects. Diabetes 48:1113–1119, 1999 plasma total leptin, whereas the protein-bound form and 13. Storlien LF, Jenks AB, Chisholm DJ, Pascoe WS, Khouri S, Kraegen EW: Influence of dietary fat composition on development of insulin resistance the soluble leptin receptor were similarly reduced in all in rats: relationship to muscle triglyceride and w-3 fatty acids in muscle pGDM after correction for BFM. phospholipid. Diabetes 40:280–289, 1991 1 In conclusion, IMCL-T as measured with HNMRS 14. Pan DA, Lillioja S, Kriketos AD, Milner MR, Baur LA, Bogardus C, Jenkins reflects insulin sensitivity and glucose homeostasis, AB, Storlien LH: Skeletal muscle triglyceride levels are inversely related to whereas IMCL-S predominantly relates to the degree of insulin action. Diabetes 46:983–988, 1997 15. Simoneau JA, Colberg SR, Thaete FL, Kelley DE: Skeletal muscle glycolytic obesity in women with pGDM. Increased IMCL particu- and oxidative enzyme capacities are determinants of insulin sensitivity and larly identifies those women who are markedly insulin- muscle composition in obese women. FASEB J 9:273–278, 1995 resistant and/or require insulin during pregnancy and who 16. Phillips DIW, Caddy S, Ilic V, Fielding BA, Frayn KN, Borthwick AC, Taylor receive a diagnosis earlier in the course of pregnancy. R: Intramuscular triglyceride and muscle insulin sensitivity: evidence for a Thus, higher IMCL relate to classical risk factors for type relationship in non-diabetic subjects. Metabolism 45:947–950, 1996 17. Oakes ND, Cooney GJ, Camilleri S, Chisholm DJ, Kraegen EW: Mecha- 2 diabetes in this cohort of young women and could be nisms of liver and muscle insulin resistance induced by chronic high-fat added to features of the metabolic syndrome and serve as feeding. Diabetes 46:1768–1774, 1997 an additional marker of risk for later type 2 diabetes in 18. Szczepaniak LS, Babcock EE, Schick F, Dobbins RL, Garg A, Burns DK, Mc women with pGDM. Garry JD, Stein DT: Measurement of intracellular triglyceride stores by 1H spectroscopy: validation in vivo. Am J Physiol 276:E977–E989, 1999 19. Boesch C, Slotboom J, Hoppeler H, Kreis R: In vivo determination of 1 ACKNOWLEDGMENTS intramyocellular lipids in human muscle by means of localized H-NMR spectroscopy. Magn Reson Med 37:484–493, 1997 These studies were supported by the Austrian Science 20. Kelley DE, Mandarino LJ: Fuel selection in human skeletal muscle in Fund to A.K.-W. (P14515-MED), M.R. (P13722-MED, insulin resistance. A reexamination. Diabetes 49:677–683, 2000 P13213-MOB), and the Austrian National Bank to H.S. and 21. Boden G, Lebed B, Schatz M, Homko C, Lemieux S: Effects of acute M.R. (O¨ NB 9127). G.P. and A.T. participated in this study changes of plasma free fatty acids on intramyocellular fat content and insulin resistance in healthy subjects. Diabetes 50:1612–1617, 2001 within a cooperative project between ISIB-CNR (formerly 22. Krssak M, Petersen K, Dresner A, DiPietro L, Vogel SM, Rothman DL, LADSEB) and the Department of Internal Medicine III, Roden M, Shulman GI: Intramyocellular lipid concentrations are correlated University of Vienna. with insulin sensitivity in humans: a 1H NMR spectroscopy study. Diabe- We are indebted to A. Hofer, O.H. Lentner, P. Nowotny, tologia 42:113–116, 1999 23. Fru¨ hbeck G, Salvador J: Relation between leptin and the regulation of and the laboratory staff of the Division of Endocrinology glucose metabolism. Diabetologia 43:3–13, 2000 and Metabolism. 24. Lewandowski K, O’Callaghan CJ, Dunlop D, Medley GF, O’Hare P, Brabant G: Free leptin, bound leptin, and soluble leptin receptor in normal and diabetic pregnancies. J Clin Endocrinol Metab 84:300–306, 1999 REFERENCES 25. Brabant G, Horn R, von zur Mu¨ hlen A, Mayr B, Wurster U, Heidenreich F, 1. Kjos S, Buchanan T: Gestational diabetes mellitus. N Engl J Med 341:1749– Schnabel D, Gru¨ ters-Kieslich A, Zimmermann-Belsing T, Feldt-Rasmussen 1756, 1999 U: Free and protein bound leptin are distinct and independently controlled 2. Kautzky-Willer A, Prager R, Waldha¨usl W, Pacini G, Thomaseth K, Wagner factors in energy regulation. Diabetologia 43:438–442, 2000 OF, Ulm M, Streli C, Ludvik B: Pronounced insulin resistance and 26. Lammert A, Kiess W, Bottner A, Glasow A, Kratzsch J: Soluble leptin inadequate B-cell secretion in lean gestational diabetes mellitus during and receptor represents the main leptin binding activity in human blood. after pregnancy. Diabetes Care 20:1717–1723, 1997 Biochem Biophys Res Commun 283:982–988, 2001 3. Ward WK, Johnston CLW, Beard JC, Benedetti TJ, Halter JB, Porte D: 27. American Diabetes Association: Clinical practice recommendations 2001. Insulin resistance and impaired insulin secretion in subjects with histories Diabetes Care 25 (Suppl. 1):S94–S96, 2002 of gestational diabetes mellitus. Diabetes 34:861–869, 1985 28. Pacini G, Tonolo G, Sambataro M, Maiolo M, Ciccarese M, Brocco E, 4. Damm P, Kuhl C, Bertelsen A, Molsted-Pedersen L: Predictive factors for Avagoro A, Nosadini R: Insulin sensitivity and glucose effectiveness: the development of diabetes in women with previous gestational diabetes minimal model analysis of regular and insulin modified FSIGT. Am J mellitus. Am J Obstet Gynecol 167:607–616, 1992 Physiol Endocrinol Metab 37:E592–E599, 1998 5. Pendergrass M, Fazioni E, De Fronzo R: Non-insulin-dependent diabetes 29. Kahn SE, Prigeon RL, McCulloch DSK, Boyko EJ, Bergman RN, Schwartz

250 DIABETES, VOL. 52, FEBRUARY 2003 A. KAUTZKY-WILLER AND ASSOCIATES

MW, Neifing JL, Ward WK, Beard JC, Palmer JP, et al: Quantification of the resistance in type 2 diabetes a slow, type 1 fiber disease? Diabetes relationship between insulin sensitivity and beta-cell function in human 50:1324–1329, 2001 subjects. Evidence for a hyperbolic function. Diabetes 42:1663–1672, 1993 42. Waldha¨usl W, Roden M: The effects of free fatty acids on glucose transport 30. Thomaseth K, Kautzky-Willer A, Ludvik B, Prager R, Pacini G: Integrated and phosphorylation in human skeletal muscle. Curr Opin Endocrinol mathematical model to assess B-cell activity during the oral glucose Diabetes 7:211–216, 2000 tolerance test. Am J Physiol 270:E522–E531, 1996 43. Oakes ND, Bell KS, Furler SM, Camilleri S, Saha AK, Ruderman NB, 31. Mari A, Pacini G, Murphy E, Ludvik B, Nolan JJ: A model-based method for Chisholm DJ, Kraegen EW: Diet-induced muscle insulin resistance in rats assessing insulin sensitivity from the oral glucose tolerance test. Diabetes is ameliorated by acute dietary lipid withdrawal or a single bout of Care 24:539–548, 2001 exercise: parallel relationship between insulin stimulation of glucose 32. Ockenga J, Biscoff SC, Tillmann HL, Rifai K, Widjaja A, Boker KWH, Manns uptake and suppression of long-chain fatty acyl-CoA. Diabetes 46:2022– MP, Brabant G: Elevated bound leptin correlates with energy expenditure 2028, 1997 in cirrhotics. Gastroenterology 119:1656–1662, 2000 44. Brechtel K, Dahl DB, Machann J, Bachmann OP, Wenzel I, Maier T, 33. Krebs M, Krssak M, Nowotny P, Weghuber D, Gruber S, Mlynarik V, Claussen CD, Haring HU, Jacob S, Schick F: Fast elevation of the Bischof M, Stingl H, Fu¨ rnsinn C, Waldha¨usl W, Roden M: Free fatty acids intramyocellular lipid content in the presence of circulating free fatty acids inhibit the glucose-stimulated increase of intramuscular glucose-6-phos- and hyperinsulinemia: a dynamic 1H-MRS study. Magn Reson Med 45:179– J Clin Endocrinol Metab phate concentration in humans. 86:2153–2160, 183, 2001 2001 45. Ye JM, Doyle PJ, Iglesias MA, Watson DG, Cooney GJ, Kraegen EW: 34. Malenfant P, Tremblay A, Doucet E, Imbeault P, Simoneau JA, Joanisse Peroxisome proliferator-activated receptor (PPAR)-alpha activation low- DR: Elevated intramyocellular lipid concentration in obese subjects is not ers muscle lipids and improves insulin sensitivity in high-fat-fed rats. reduced after diet and exercise. Am J Physiol Endocrinol Metab 280:E632– Comparison with PPAR-gamma activation. Diabetes 50:411–417, 2001 E639, 2001 46. Azen SP, Peters RK, Berkowitz K, Kjos S, Xian A, Buchanan TA.TRIPOD 35. Goodpaster BH, Thaete FL, Simoneau JA, Kelley DE: Subcutaneous (troglitazone: in the prevention of diabetes): a randomized, placebo- abdominal fat and thigh muscle composition predict insulin sensitivity independently of visceral fat. Diabetes 46:1579–1585, 1997 controlled trial of troglitazone in women with prior gestational diabetes 36. Johnson MA, Polgar J, Weightman D, Appleton D: Data of the distribution mellitus. Control Clin Trials 19:217–231, 1998 of fiber types in 36 human muscles. An autopsy study. J Neurol Sci 47. Kelley DE, Goodpasture BH: Skeletal muscle triglyceride. An aspect of 18:111–129, 1973 regional adiposity and insulin resistance. Diabetes Care 24:933–941, 2001 37. Hickey MS, Weidner MD, Gavigan KE, Zheng D, Tyndall GL, Houmard JA: 48. Kautzky-Willer A, Pacini G, Tura A, Bieglmeyer C, Schneider B, Ludvik B, The insulin action-fiber type relationship in humans is muscle groups Prager R, Waldha¨usl W: Increased plasma leptin in gestational diabetes. specific. Am J Physiol 269:E150–E154, 1995 Diabetologia 44:164–172, 2001 38. Dyck DJ, Peters SJ, Glatz J, Gorski J, Keizer H, Kiens B, Liu S, Richter EA, 49. Stock P, Scholl Th, Schluter M, Schroeder Ch: Plasma leptin influences Spriett LL, van der Vusse GJ, Bonen GA: Functional differences in lipid gestational weight gain and postpartum weight retention. Am J Clin Nutr metabolism in resting skeletal muscle of various fiber types. Am J Physiol 8:1236–1240, 1998 271:E340–E351, 1997 50. McNeely MJ, Boyko EJ, Weigle DS, Shofer JB, Chessler SD, Leonnetti DL, 39. Hickey MS, Carey JO, Azevedo JL, Houmard JA, Pories WJ, Israel RG, Fujimoto WY: Association between baseline plasma leptin levels and Dohm GL: Skeletal muscle fiber type is related to adiposity and in vitro subsequent development of diabetes in Japanese Americans. Diabetes glucose transport rate in humans. Am J Physiol 268:E453–E457, 1995 Care 22:65–70, 1999 40. Nyholm B, Qu Z, Kaal A, Pedersen SB, Gravholt CA, Andersen JL, Saltin B, 51. Seufert J, Kiefer TJ, Leech CA, Holz GG, Moritz W, Ricordi C, Habener JF: Schmitz O: Evidence of an increased number of type IIb muscle fibers in Leptin suppression of insulin secretion and gene expression in human insulin-resistant first-degree-relatives of patients with NIDDM. Diabetes pancreatic islets: implications for the development of adipogenic diabetes 46:1822–1828, 1997 mellitus. J Clin Endocrinol Metab 84:670–676, 1999 41. Gaster M, Staehr P, Beck-Nielsen H, Schroder HD, Handberg A: GLUT4 is 52. Huang L, Wang Z, Li C: Modulation of circulating leptin levels by its soluble reduced in slow muscle fibers of type 2 diabetic patients. Is insulin- receptor. J Biol Chem 276:6343–6349, 2001

DIABETES, VOL. 52, FEBRUARY 2003 251