Introduction

Type 2 diabetes is characterized by insulin resistance and impaired insulin secretion. Obesity in general, and abdominal (truncal) obesity in particular, is well-known to be associated with insulin resistance and cardiometabolic diseases. In addition to visceral adiposity, a low amount of leg (gluteofemoral) fat mass is reported to be associated with insulin resistance and cardiometabolic abnormalities1,2. We3 previously reported that after adjustment for trunk fat, leg fat mass showed an inverse association with triglyceride and a positive association with HDL cholesterol. Leg fat mass also showed a positive association with serum adiponectin and fasting lipoprotein lipase (LPL) even in young female Japanese university students whose BMI and waist circumference averaged 20.9 kg/m2 and 72 cm, respectively3. We also previously reported a low amount of leg fat mass in young female Japanese university students with reduced birth weight4 and a family history of diabetes5. Lotta et al.6 suggest that lower gluteofemoral and higher abdominal fat may be associated with higher blood pressure and triglyceride levels and a higher risk of type 2 diabetes and coronary disease.

Lipoprotein lipase (LPL) promotes postprandial triglyceride breakdown in the circulation and lipid storage in adipose tissue. Previous Mendelian randomization analyses suggested potential causal effects of LPL on insulin levels and the development of type 2 diabetes7. Moreover, studies showed that triglyceride-lowering alleles in the LPL pathway were associated with a lower risk of type 2 diabetes8,9. Recent proteomic analyses in Chinese adults confirmed this association10. The mechanisms linking increased LPL activity to decreased insulin resistance and diabetes may include triglyceride storage in gluteofemoral rather than visceral fat9,11.

A small quantity of LPL protein, which is catalytically inactive, exists in serum12. Serum LPL mass concentrations were correlated with adipose tissue release of LPL protein13. This suggests that the rate of adipose tissue release of LPL protein is a major determinant of systemic LPL mass concentrations13. Rip et al.14 reported that reduced levels of serum LPL are associated with an increased risk for future coronary artery disease. As discussed in detail by Rip et al.14, serum LPL concentrations may be a marker for the amount of systemically available (catalytically) active LPL.

Prediabetes, an intermediate hyperglycemia between normal glucose tolerance and diabetes, is defined by fasting glucose or 2 h glucose concentrations during the oral glucose tolerance test (OGTT)15. However, there is ample evidence that 1 h plasma glucose concentrations (1hPG) predict the future development of type 2 diabetes, with equal or even superior performance to 2 h glucose [see Table 2 of Ref. 16]. Therefore, the International Diabetes Federation advocated that 1hPG ≥ 155 mg/dL may potentially identify individuals with normal glucose tolerance at high risk of type 2 diabetes at an earlier stage16. Elevated 1hPG was associated with many cardiovascular risk factors, target organ damage, and adverse outcomes, including obesity and visceral adiposity [see Table 6 of Ref. 16]. However, we found no study that investigates the associations of elevated 1hPG, gluteofemoral fat mass, and serum LPL. Type 2 diabetes in Asians was characterized by diabetes onset at a much lower BMI17, and body composition is related to socioeconomic status18. We have recently suggested that impaired pancreatic β-cell function found in nonobese Japanese with low leg fat may explain why type 2 diabetes develops at a much lower BMI and predominates insulin secretion failure in Japanese people19. Therefore, the present study tested whether elevated 1hPG was associated with decreased leg fat and serum LPL concentrations in Japanese female university students and their biological parents, a population where body composition is less related to socioeconomic status20.

Methods

We previously reported results of OGTT in 168 daughters and 124 parents19,20,21,22. In the present study, we cross-sectionally investigated 164 Japanese female university students and 94 middle-aged parents (54 mothers and 40 fathers). They all had normal glucose tolerance, defined as fasting glucose levels < 110 mg/dL and 2 h glucose concentrations <140 mg/dL23. Subjects who reported having clinically diagnosed acute or chronic inflammatory diseases, endocrine, cardiovascular, hepatic, and renal diseases were excluded from the study. Those on a diet to lose weight and hormonal contraception were also excluded. This research followed the tenets of the Declaration of Helsinki. All participants gave written informed consent after the experimental procedure had been explained. The study was approved by the Ethics Committees of the Mukogawa Women’s University (No. 07-28 on 19/02/2008).

More than 90% of first-year students were 18 years old19. This may decrease the interference of age and environmental factors such as alcohol intake, smoking, educational and socioeconomic status. Further, almost all school expenses were covered by parents, suggesting that socioeconomic status may be less heterogeneous among parents19.

Blood samples were obtained in the morning after a 12 h overnight fast. Plasma glucose, serum insulin, lipids, lipoproteins, and free fatty acid (FFA) were measured as previously reported19,20,21,22. Serum LPL concentrations were measured using ELISA24 using a commercial kit from Daiichi Pure Chemicals (Tokyo, Japan, with an interassay CV = 2.8%). HbA1c values were determined by high-performance liquid chromatography (HLC723-G7, Tosoh, Tokyo, Japan). During a 75-g OGTT, blood samples were taken at min 0 (fasting), 30, 60, and 120 min for glucose and insulin analysis. Homeostasis model assessment insulin resistance (HOMA-IR) was determined using fasting plasma glucose and insulin levels25, and insulin sensitivity by the Matsuda index using glucose and insulin levels during OGTT26. Early-phase glucose-stimulated insulin secretion was estimated using the insulinogenic index (IGI)27. This was calculated as the incremental insulin (μU/mL) divided by the incremental glucose (mg/dL) during the first 30 min of OGTT. The oral disposition index (ODI) was calculated as the product of the IGI and the Matsuda index28. The area under the glucose and insulin concentration curves during OGTT (AUCg and AUCi, respectively) were calculated using the trapezoidal method. Adipose tissue insulin resistance index (AT-IR) was calculated as a product of fasting insulin (μU/mL) and FFA (mEq/L)29.

Body weight, height, and waist circumference were measured after an overnight fast with light clothing and shoes off, and BMI was calculated. Whole-body dual-energy X-ray absorptiometry (DXA) (Hologic QDR-2000, software version 7.20 D, Waltham, MA) was used to measure lean tissue mass, fat mass, and bone mineral mass for arms, legs (gluteofemoral regions), the trunk, and the total body20. The ratio of leg fat to body fat (LF/BF ratio) and the ratio of trunk fat to body fat (TF/BF ratio) were calculated in each participant. Fat mass index and leg fat mass index were calculated as body fat and leg fat in kg divided by squared height in meters, respectively.

Data were presented as mean ± SD unless otherwise stated. Due to deviation from normal distribution, fasting insulin, triglyceride, HOMA-IR, AT-IR, IGI, and ODI were logarithmically transformed for analyses. Comparisons between those with and without elevated 1hPG were made with a two-sample t-test. Multivariate logistic regression analyses were performed to identify the most significant variables contributing to elevated 1hPG. Variables that showed significant differences between the two groups were included as independent variables. A two-tailed p < 0.05 was considered statistically significant. All calculations were performed with SPSS system 23 (SPSS Inc., Chicago, IL).

Results

Female students and their parents were on average nonobese (Table 1), and 95.1% of daughters and 84.0% of parents had a BMI < 25 kg/m2. Elevated 1hPG was observed in 6% (10/164) of students, 14.8% (8/54) of mothers, 32.5% (13/40) of fathers, and 22.3% (21/94) of parents. Parents had higher BMI, fat mass index, waist circumference, and TF/BF ratio, and lower LF/BF ratio. Fasting and 2 h insulin, HOMA-IR, AT-IR, and Matsuda index, as indicated in bold, did not differ. However, serum TG and systolic blood pressure were higher, and HDL cholesterol and LPL were lower in parents. They had lower IGI and ODI, reflecting β-cell function adjusted for insulin sensitivity, and higher fasting and 2 h glucose and HbA1c. Serum adipokines did not differ.

Table 1 Features of female students and parents with normal glucose tolerance.
Full size table

Daughters with elevated 1hPG compared to those without had lower birth weight (Fig. 1 and Table 2). There was no difference in anthropometric and fat distribution measures and serum leptin (Table 2 and Supplementary Table S1). However, daughters with elevated 1hPG had lower serum adiponectin (Fig. 1). They had lower log IGI, although HOMA-IR, AT-IR, and Matsuda index did not differ (Table 2). There was also no difference in metabolic variables, including serum triglyceride and HDL cholesterol (Table 2). However, serum LPL was lower in daughters with elevated 1hPG (Fig. 1). 

Fig. 1
figure 1

Jitter plots of birth weight, serum adiponectin, log insulinogenic index (IGI), and serum lipoprotein lipase (LPL) in young female Japanese students with (n = 10 in red) and without (n = 154 in green) elevated 1 h post-load plasma glucose. They all had normal glucose tolerance. See Table 2 for means, SD, and p values.

Full size image
Table 2 Insulin sensitivity/resistance, insulin secretion, and cardiometabolic risk factors in students with normal glucose tolerance and elevated 1 h plasma glucose.
Full size table

Parents with elevated 1hPG had a lower LF/BF ratio and higher BMI, fat mass index, waist circumference, and TF/BF ratio (Table 3). However, the ratio of arm fat to body fat, leg fat mass index, ASMI, and %ASM did not differ.

Table 3 Body composition of parents with normal glucose tolerance and elevated 1 h plasma glucose.
Full size table

Parents with elevated 1-hPG had lower log IGI and Matsuda index and higher fasting and 2 h insulin, HOMA-IR, and AT-IR (Table 4). They also had higher log triglyceride and systolic blood pressure, and lower HDL cholesterol and LPL. Serum adipokines did not differ.

Table 4 Insulin sensitivity/resistance, insulin secretion, and cardiometabolic risk factors in parents with normal glucose tolerance and elevated 1 h plasma glucose.
Full size table

Multivariate logistic regression analyses in daughters revealed that elevated 1hPG was associated with log IGI and serum LPL independently of birth weight and serum adiponectin (Table 5). Independent associations with log IGI (Odds ratio; 0.051, 95% confidence interval: 0.008–0.341, p = 0.002) and serum LPL (Odds ratio; 0.916, 95% confidence interval: 0.866–0.969, p = 0.002) were confirmed in a model that included log IGI and serum LPL as independent variables. In parents, LF/BF ratio, log IGI, and Matsuda index were related to elevated 1hPG independently of fat mass index, TF/BF ratio, serum LPL, and HOMA-IR.

Table 5 Multivariate logistic regression analysis for elevated 1 h plasma glucose.
Full size table

Propensity score analyses confirmed the results described above, except for birth weight in daughters and log IGI in parents (Table 6).

Table 6 Propensity score analyses for elevated 1 h plasma glucose in daughters (A) and parents (B).
Full size table

Discussion

The present study demonstrated that young Japanese women with elevated 1hPG, whose BMI averaged 21 kg/m2, had decreased early-phase insulin secretion (estimated by IGI), decreased birth weight, serum adiponectin, and LPL. The associations with decreased IGI and serum LPL were confirmed in middle-aged Japanese parents. In addition, elevated 1hPG was associated with decreased IGI and serum LPL independently of serum adiponectin in young Japanese. In middle-aged Japanese parents, decreased IGI and leg fat (relative to total body fat) were related to elevated 1hPG independently of serum LPL.

A reduced amount of leg (gluteofemoral) fat mass may be associated with insulin resistance and cardiometabolic abnormalities1,2. LPL promotes lipid storage in subcutaneous adipose tissue, mainly located in the gluteofemoral region in non-obese individuals. Serum LPL concentrations may be a possible marker for the amount of systemically available (catalytically) active LPL14. We have previously reported associations of serum LPL concentrations between mothers and daughters, though borderline significant (p = 0.06)30. As mothers and fathers with elevated 1hPG had decreased serum LPL concentrations, we hypothesized that a prolonged deficiency of LPL, a rate-limiting enzyme for lipid storage in subcutaneous leg fat, may explain the decreased amount of leg fat in middle-aged Japanese with elevated 1hPG. Studies reported that higher subcutaneous adipose tissue LPL activity was associated with greater triglyceride storage in subcutaneous adipose tissue, thereby reducing visceral adipose tissue accumulation and metabolic dysfunction31,32.

Early β-cell dysfunction appeared to be a predominant feature in Asian populations who progressed to type 2 diabetes17. In the present study, impaired early-phase insulin secretion was evident in middle-aged parents and in young Japanese women with elevated 1hPG, whose BMI averaged 21 kg/m2 and waist circumference 73 cm. There was no difference in four measures of insulin resistance/ sensitivity (fasting insulin, HOMA-IR, AT-IR, and Matsuda index) between young women with and without elevated 1hPG. Recent studies from China and Japan reported impaired early-phase insulin secretion in people with elevated 1hPG and normal glucose tolerance33,34. Taken together, in nonobese East Asians, impaired early insulin secretion might be an underlying mechanism of an elevated 1hPG during an OGTT.

Studies from North America and Europe reported that elevated 1hPG was associated with obesity and visceral adiposity16. However, few studies investigated associations with leg fat mass. In middle-aged Japanese parents, elevated 1hPG was related to decreased leg fat (relative to total body fat), not to elevated trunk fat (relative to total body fat), despite higher BMI and waist circumference. This finding may be associated with an observation that a decreased amount of leg (gluteofemoral) fat mass was associated with insulin secretion failure in normal-weight but not overweight and obese middle-aged Caucasians1.

The proportion of participants with elevated 1hPG and normal glucose tolerance was 12% in obese Italian youth35 and 35% in obese Latino youth with a family history of type 2 diabetes36. It was 6% in lean young Japanese women whose BMI and waist circumference averaged 21 kg/m2 and 73 cm, respectively. The current study may be the first to report the proportion of participants with elevated 1hPG and normal glucose tolerance in young Asian populations.

Parents compared to their daughters with normal glucose tolerance had higher fasting glucose and triglyceride, blood pressure, and lower HDL cholesterol (Table 1). However, there was no difference in four measures of insulin resistance/sensitivity (fasting insulin, HOMA-IR, AT-IR, and Matsuda index) between the two groups. Body fat percentage and %ASM also did not differ, although fat mass index, BMI, and waist circumference were higher in parents with normal glucose tolerance. ASMI was higher in parents. A likely explanation is lower leg fat (relative to total body fat), which is associated with unfavorable cardiometabolic health1,19,37,38.

The strength of this study includes a homogeneous study population with scarce confounding factors19,20,21,22. Several limitations of this study include the cross-sectional design, relatively small sample size, and a single measurement of biochemical variables. We used many surrogates of insulin secretion and insulin resistance/sensitivity in the present study, which may be less accurate. Finally, as we studied Japanese only, results may not be generalized to other races or ethnicities.

In conclusion, a decreased amount of leg fat in middle-aged Japanese with elevated 1hPG may be explained hypothetically by a prolonged deficiency of LPL, a rate-limiting enzyme for lipid storage in subcutaneous leg fat. Further studies in other populations and other genres will be needed.