Abstract. Seidell JC, Han TS, Feskens EJM, Lean MEJ (National Institute of Public Health and the Environment, Bilthoven, the Netherlands, and Royal Infirmary, University of Glasgow, Glasgow, UK). Narrow hips and broad waist circumferences independently contribute to increased risk of noninsulin-dependent diabetes mellitus. J Intern Med 1997; 242: 401-406.Objectives. Patients with non-insulin-dependent diabetes mellitus (NIDDM) have been shown to be more obese and have higher waist-to-hips circumference ratios compared to nondiabetics. In this study, we tried to dissociate obesity, waist and hip circumference from NIDDM. Design. A cross-sectional population-based casecontrol study. Setting. General population. Subjects. Controls: 5887 men and 7018 women, aged 20-59 years, without known diabetes or hyperglycaemia. Cases: 93 men and 66 women were diagnosed with NIDDM. Main outcome measures. We predicted waist and hips' circumference from the body mass index, weight/height 2 , (BMI) on the basis of linear regression. Differences between observed and expected values (residuals) of waist and hip circumference were categorized into tertiles. The relative odds of having NIDDM in tertiles of waist and hip residuals (middle tertile as reference) were calculated by multiple logistic regression analysis adjusted for each other and for age, smoking, physical activity, alcohol consumption and education. Results. NIDDM was most prevalent in men and women who had larger waists and in those who had smaller hip circumferences than expected from their BMI. Adjusted odds ratios (95% CI) for NIDDM were 2.9 (1.6-5.1) in men and 2.8 (1.5-5.1) in women who had a larger than expected waist, and were 3.7 (2.1-6.5) in men and 2.1 (1.1-3.8) in women who had smaller than expected hips. Conclusions. Our findings suggest that diabetics have larger waists and smaller hips compared to nondiabetics, irrespective of their degree of obesity, age and life-style factors. One possibility is that besides abdominal fatness, peripheral muscle atrophy is one of the factors associated with NIDDM.

[2-(14)C]quercetin-4'-glucoside (4 mg/kg body weight) was fed by gavage to rats housed in metabolic cages, and over an ensuing 72 h period, radiolabeled products in body tissues, plasma, feces, and urine were monitored by high-performance liquid chromatography with online radioactivity and MS2 detection. One and 6 h after ingestion, while in the small intestine, the flavonol glucoside was converted to glucuronide and methylated and sulfated derivatives of quercetin, but only trace amounts of these metabolites were excreted in urine. On entering the cecum and the colon, the flavonol metabolites declined as they were converted to phenolic acids, principally 3-hydroxyphenylacetic acid and 3,4-dihydroxyphenylacetic acid, by the colonic microflora. Feces contained mainly 3-hydroxyphenylacetic acid. Urine collected 0-12 and 0-24 h after ingestion contained radiolabeled hippuric acid and 3-hydroxyphenylacetic acid. 14C-Hippuric acid declined markedly in the 24-48 and 48-72 h urine samples, and there was a concomitant increase in labeled benzoic acid. There was minimal accumulation of radioactivity in plasma, despite a 69% recovery of label in urine over the 72 h period, and likewise, very little radioactivity was detected in body tissues out with the gastrointestinal tract. This is reflected in the fact that 72 h after ingestion 96% of the ingested radioactivity was recovered in feces, urine, and the cage washes, which comprise a mixture of urine and feces. The study reveals that as it passes through the gastrointestinal tract, almost all of the of [2-(14)C]quercetin-4'-glucoside is converted to phenolic acids, compounds not monitored in previous flavonol bioavailability studies with model animal systems, some of which have used exceedingly high doses of the aglycone quercetin (500 mg/kg body weight), which is not a normal dietary component.

OBJECTIVES: To assess the in¯uences of height and age on the differences in waist circumference between individuals of different stature. SUBJECTS: 3319 males and 4358 females from four studies in the UK and the Netherlands. MEASUREMENTS: Waist circumference, body weight, height, and age. RESULTS: Linear regression analysis of log 10 height as the independent variable on log 10 waist as the dependent variable was used to determine the optimal index powers (OIP) ( p ) to minimize the in¯uence of height in the relationships of waist/height p . Six out of eight samples of men and women had OIP of height not signi®cantly different from zero, with the remaining two groups had OIP between 0.15±0.58, indicating that height had very limited in¯uence on the differences in waist circumference measurement between individuals. Age adjustment increased the relationship between waist and height, with OIP of 0.19±0.89 in men and 0.02±0.58 in women. Without age adjustment, height explained 0.3±3.5% and 0.1±2.5% variance in waist in men and in women respectively, and the corresponding variances were 0.4±7.5% in men and 0.0±2.6% in women with age adjustment. A similar analysis of weight and height showed the OIP of height in weight/height p ratio ranged from 1.32±2.25 in men, and 0.87±1.74 in women without age adjustment, and from 1.47±2.24 in men and 1.25±1.96 in women with age adjustment. CONCLUSION: Height and age had limited in¯uences on the differences in waist between Caucasian subjects of different stature. Waist alone may be used to indicate adiposity or to re¯ect metabolic risk factors. In contrast, the in¯uence of height on body weight is important.