Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Body Fat Distribution”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,099 records · Page 61Linked to original sources
PubMedDate not supplied

Body composition analyses in normal weight obese women.

The purpose of this study was to identify new indexes of body composition that characterize the normal weight obese (NWO) women. We measured body composition by dual energy x-ray absorptiometry (DXA) and resting metabolic rate (RMR) by indirect calorimetry in a cohort of seventy-five healthy Italian women, subdivided into three groups (nonobese/controls, NWO, preobese-obese women). Despite a normal body mass index (BMI), the NWO women have a higher body fat mass percentage (FAT %) (38.99 +/- 6.03) associated to a significant (p = 0.02) lower amount of lean mass of legs (12.24 +/- 1.31) and lean mass of left leg (6.07 +/- 0.64) with respect to the control group. The NWO group showed a significant (p = 0.043) lower RMR (1201.25 +/- 349.02) in comparison with nonobese and preobese-obese women. To classify NWO individuals among general population, we identified three significant body composition indexes: abdominal index, leg index and trunk index. The NWO women showed significant increased value in the three indexes (p < 0.001). Our results suggest that, despite a normal BMI, the NWO women displayed a cluster of anthropometric characteristics (body fat mass percentage, leg indexes) not different to obese women ones. An appropriate diet-therapy and physical activity may be protecting NWO individuals from diabetes and cardiovascular diseases associated to preobese-obese women.

Abdominal Fat↗
PubMed1994

Contribution of body fat and fat pattern to blood pressure level in school children.

OBJECTIVE: The study aims are investigating the effect of body fat and fat localization on blood pressure. DESIGN: Case-control study. SETTING: The study was carried out in the school health primary care. SUBJECTS: The case-control study included 220 obese and 220 non-obese children aged 7-18 years from Al Ain city, United Arab Emirates between September 1992 to May 1993 inclusive. Each group consisted of 120 males and 100 females. Two schools were randomly selected from each of the three educational stages: primary, junior and secondary. The inclusion criterion for cases comprised children with body mass index (BMI; Quetelet index), > 90th percentile of age and sex-specific reference data of the French population. Non-obese healthy controls were randomly selected from the same classes from where obese children were identified in order to ascertain that cases and controls were matched by age and sex. MEASUREMENTS: Anthropometric measures (weight, height, waist and hip circumferences), systolic and diastolic blood pressure were measured. To minimize inter-observer error, blood pressure was measured by one physician. We also collected information about other confounding social variables (family history of obesity and mother's education) and behavioural variables (preferred diet and physical activity). RESULTS: There was significant difference of systolic and diastolic blood pressure means between obese and non-obese children (P < 0.001) in both males and females. Applying the multiple linear regression analysis to fix the confounding effect of age, sex, social and behavioural factors, the fatness index, BMI, was significantly related to systolic (P < 0.0004) and diastolic (P < 0.0001); while waist-to-hip circumference ratio (WHR) was not significant (P = 0.803 in systolic and P = 0.648 in diastolic blood pressure respectively). CONCLUSIONS: Systolic and diastolic blood pressure showed a positive relationship with the fatness index BMI, but not with WHR, in both boys and girls. This is an evidence that WHR may not be a reliable indicator of body fat distribution in children.

Adipose Tissue↗
PubMed2001

Endocrine, metabolic and nutritional factors in obesity and their relative significance as studied by factor analysis.

OBJECTIVE: To evaluate different aspects of obesity, such as body fat distribution, plasma hormone and lipid profiles, adipose tissue composition and dietary intake in an obese population in order to identify the most important factors that contribute to obesity. DESIGN: Eighty-five obese subjects, 30 men and 55 women (age, 30-70 y; body mass index (BMI), 27-35 kg/m2), were studied using anthropometric measurements, computed tomography, adipose tissue composition, serum hormone and lipid profiles and nutritional evaluations. To determine to what extent individual factors contributed to the general process of obesity, the data were subjected to a factor analysis. RESULTS: Three patterns of anthropometric and computed tomography data emerged that accounted for 69% of the variance. Factor 1 defined abdominal obesity and explained 30% of the total variance, factor 2 (gynoid obesity) accounted for 26%; and factor 3 (subcutaneous fat) explained 13% of the total variance. When other factors associated with obesity, such as lipid profile, hormonal profile and fat composition, were introduced, obesity itself, especially abdominal obesity, remained the principal factor, accounting for 23% of total variability. All factors were of secondary importance when dietary characteristics were introduced. In the overall factor analysis, more than 40% of the variability in obesity was related to dietary habits, particularly fat intake, followed by energy and saturated fatty acids intake. CONCLUSION: Even though obesity is a multifactorial phenomenon, the results suggest that dietary intake, especially fat intake, is the most important factor contributing to obesity. Secondary factors include endocrine and metabolic factors.

Adipose Tissue↗
PubMed2000

Hyperlipidemia and insulin resistance are induced by protease inhibitors independent of changes in body composition in patients with HIV infection.

Although protease inhibitor (PI) therapy has improved the clinical status of patients with HIV infection, concerns have arisen that such treatment may have deleterious effects on glucose control, lipid metabolism, and body fat distribution. To determine whether initiation of PI therapy uniquely affects glucose and lipid metabolism, we analyzed paired data in HIV-infected patients before and after beginning antiretroviral therapy that included a PI (PI; N = 20) or lamivudine (3TC) but no PI (3TC; N = 9); and a control group on stable regimens that included neither of these agents (CONT: N = 12). Measurements included fasting glucose; insulin; triglycerides; total, high-density lipoprotein (HDL) and low-density lipoprotein (LDL) cholesterol; HIV RNA; CD4+ lymphocytes; weight; and total and regional body composition. Neither weight nor total or regional fat content changed significantly in any group during the period of observation. Nonetheless, in patients beginning PI therapy, there were significant increases in glucose (+9+/-3 mg/dl; p = .0136), insulin (+12.2+/-4.9 U/ml; p = .023), triglycerides (+53+/-17 mg/dl; p = .0069), and total and LDL cholesterol (+32+/-11 and +18+/-5 mg/dl; p = .0082 and .0026, respectively). None of these parameters changed significantly in the 3TC or CONT groups. The PI and 3TC groups experienced comparable increases in CD4+ lymphocytes, suggesting that the observed metabolic effects may be associated with PIs uniquely, rather than improvement in clinical status. However, it is also possible that the metabolic effects of PIs are associated with more effective viral suppression, because a greater proportion of patients in the PI group achieved undetectable levels of virus. We conclude that changes in glucose and lipid metabolism are induced by PI therapy in the absence of significant changes in weight or fat distribution. Longer periods of follow-up will be required to determine the clinical significance of these changes. However, at the moment, the risks associated with these metabolic effects do not appear to outweigh improvements in survival seen with PI therapy.

Adult↗
PubMed1988

Body fat topography in women with androgen excess.

The relationship between body fat distribution and androgen excess was determined in 168 premenopausal women of varying degrees of androgen excess. In 84 women with the polycystic ovary syndrome or idiopathic hirsutism, the waist:hips girth ratio (WHR) was 0.81 +/- 0.01 (mean +/- s.e.m.) and was significantly greater (P less than 0.001) than in an age- and weight-matched group of 84 non-hirsute women (0.77 +/- 0.01), indicating upper body fat predominance in the hirsute subjects. Plasma testosterone was directly correlated with the WHR (r = 0.55, P less than 0.001) and this relationship was independent of obesity level, supporting the hypothesis that plasma androgens are an important determinant of body fat topography in premenopausal women.

Adipose Tissue↗
PubMed2001

Waist to hip circumference ratio as a significant predictor of preeclampsia, irrespective of overall adiposity.

OBJECTIVE: To investigate whether abnormal body fat distribution is a significant predictor of the development of preeclampsia, irrespective of overall adiposity. METHODS: Twenty-six women with preeclampsia and 198 control women were enrolled. Waist to hip circumference ratio (WHR), body weight (BW), and body mass index (BMI, wt/ht2) were measured early in pregnancy (< 9 weeks of gestational age). Age, height, parity, tobacco usage, education period, gestational duration, and weight gain during pregnancy were also recorded for each subject. RESULTS: WHR, BMI, and BW early in pregnancy were significantly higher in the preeclampsia group (p < 0.0001). WHR, BMI, and BW positively correlated with the development of preeclampsia on univariate regression analysis (Standardized regression coefficient = 0.410, 0.387, and 0.363, respectively, p < 0.0001). On stepwise multiple regression analysis, WHR still correlated with the development of preeclampsia irrespective of BMI and BW. When the WHR predictive of the development of preeclampsia was set at 0.9, the sensitivity was 46.2% (12/26), which was significantly better than that of 25 of BMI (19.2%, 5/26; p < 0.05). CONCLUSION: Higher WHR is a significant predictor of the development of preeclampsia. This relation is irrespective of overall adiposity.

Adipose Tissue↗
PubMed2006

Relationship of adiponectin with insulin sensitivity in humans, independent of lipid availability.

OBJECTIVE: To test in humans the hypothesis that part of the association of adiponectin with insulin sensitivity is independent of lipid availability. RESEARCH METHODS AND PROCEDURES: We studied relationships among plasma adiponectin, insulin sensitivity (by hyperinsulinemic-euglycemic clamp), total adiposity (by DXA), visceral adiposity (VAT; by magnetic resonance imaging), and indices of lipid available to muscle, including circulating and intramyocellular lipid (IMCL; by 1H-magnetic resonance spectroscopy). Our cohort included normal weight to obese men (n = 36). RESULTS: Plasma adiponectin was directly associated with insulin sensitivity and high-density lipoprotein-cholesterol and inversely with plasma triglycerides but not IMCL. These findings are consistent with adiponectin promoting lipid uptake and subsequent oxidation in muscle and inhibiting TG synthesis in the liver. In multiple regression models that also included visceral and total fat, free fatty acids, TGs, and IMCL, either alone or in combination, adiponectin independently predicted insulin sensitivity, consistent with some of its insulin-sensitizing effects being mediated through mechanisms other than modulation of lipid metabolism. Because VAT directly correlated with total fat and all three indices of local lipid availability, free fatty acids, and IMCL, an efficient regression model of insulin sensitivity (R2 = 0.69, p < 0.0001) contained only VAT (part R2 = 0.12, p < 0.002) and adiponectin (part R2 = 0.41, p < 0.0001) as independent variables. DISCUSSION: Given the broad range of total adiposity and body fat distribution in our cohort, we suggest that insulin sensitivity is robustly associated with adiponectin and VAT.

Adiponectin↗
PubMedDate not supplied

Obesity and diabetes.

Fat mass per se has little effect on the progression of obesity towards diabetes. Predominance of fat in the upper part of the body resulting in android obesity is at least the clinical reflection of factors which lead obesity to progress towards diabetes and atherosclerosis. Therefore this type of obesity may be termed diabetogenic and atherogenic obesity. Insulin and cortisol secretion in obesity are not correlated with body fat but with the predominance of fat in the upper part of the body. Diabetogenic obesity may evolve through 5 stages from initial obesity without diabetes to insulin-dependent diabetes in previously obese subjects. Aside from the characteristics of body fat distribution, the mechanisms which induce or interrupt this progression remain unknown.

Adipose Tissue↗
PubMed2005

Trends in waist circumferences in young British children: a comparative study.

OBJECTIVE: To compare waist circumference (a measure of central fatness) and body mass index (BMI, a measure of general fatness) between a contemporary cohort and similar aged British children measured in 1987. DESIGN: : Comparative study of two cohorts. The first was a cross-sectional survey and the second an ongoing longitudinal survey. SETTING: : The cross-sectional study was conducted throughout Great Britain during 1987. The ongoing longitudinal study was conducted in the Avon region between 1995 and 1998. PARTICIPANTS: A total of 1821 children were measured in the cross-sectional study, and around 1000 children were measured in the longitudinal study. Ages ranged between 2 and 5 y. MAIN OUTCOME MEASURES: Waist circumference, height, weight and BMI. Values were expressed as mean+/-s.d. Age-related measures were compared. RESULTS: At equivalent ages, mean waist circumference values were greater in the contemporary children, compared with children measured in the earlier cross-sectional study in 1987 (P<0.05). Although boys from the contemporary cohort had the larger absolute waist circumference measurements, the absolute and percentage differences in waist circumference between the cohorts were greater for the girls. Mean BMI values were slightly but significantly higher in the contemporary children compared with children measured in the earlier cross-sectional study (P<0.05). The proportional increase in waist circumference for each age-sex group generally exceeded the proportional increase in BMI. CONCLUSION: These results suggest that central fatness in young British children has increased over the period of 1987-97 to a greater extent than general fatness. This should be of concern since evidence associates greater central adiposity with adverse levels of cardiovascular risk factors in children. Furthermore, the study highlights important shortcomings of the BMI measurement, in that it provides no information on body fat distribution and can mask true obesity-related risk in children.

Adipose Tissue↗
PubMed2000

Modulation of blood pressure and obesity with the dopamine D2 receptor gene TaqI polymorphism.

Pharmacological data suggest that obesity and blood pressure (BP) may be modulated through the dopamine D2 receptor (DD2R), which may represent an underlying mechanism that links these conditions. A TAQ:I polymorphism near the DD2R gene has been associated with indices of obesity in white populations. We compared anthropometric and fasting plasma biochemical parameters between 209 nondiabetic hypertensive and 174 gender-matched normotensive Chinese subjects. The hypertensives had increased dyslipidemia, increased fasting plasma glucose concentrations, and a greater degree of obesity. The A1 and A2 alleles of the DD2R gene TAQ:I polymorphism were identified with a polymerase chain reaction-based restriction fragment length polymorphism protocol. The A1 allele frequency was decreased in the hypertensives (42.0%) compared with the control subjects (52.0%, P=0.006), and genotype frequencies were different (P=0.05) between the 2 groups. In the combined population (n=383), systolic, diastolic, and mean arterial BPs were 6, 5, and 6 mm Hg lower, respectively, in subjects with the A1A1 genotype relative to the A2A2 genotype (all P<0.05), whereas skinfold thickness was increased at the iliac (P<0.001) and triceps (P<0.03) sites but not at the biceps or subscapular sites. Furthermore, this DD2R gene polymorphism was shown to be a significant independent predictor of diastolic BP and iliac and triceps skinfold thicknesses (all P<0.03). These contrasting associations of the DD2R TAQ:I polymorphism A1 allele with lower BP but increased markers of "gynoidal" or peripheral subcutaneous obesity (iliac and triceps skinfold thicknesses) in our Chinese population may provide some insight into the underlying relationship between BP and body fat distribution, but the exact nature of this link remains to be determined.

Adult↗
PubMed1994

Factors associated with obesity in school children.

To investigate factors associated with the development of obesity in school children, the authors undertook a case-control study of 220 obese and 220 non-obese children aged 6 to 18 years from Al Ain, United Arab Emirates between September 1992 to May 1993 inclusive. The inclusion criterion for cases comprised children with body mass (Quetelet index) > 90th percentile of age-sex reference data of the French population. Questionnaires were used to obtain information about socioeconomic status (SES), family history of obesity, related diseases and behavioral factors (smoking, physical activity and preferred diet). Anthropometric measures (weight, height, waist and hip circumferences), systolic and diastolic blood pressure were measured. The waist/hip circumference ratio was used as an index for body fat distribution. The logistic regression analysis showed that family history of obesity, diet, physical activity and mother education were significant factors for development of obesity after adjusting for other confounding covariates. Neither smoking nor SES showed apparent relation to obesity. Both systolic and diastolic blood pressure showed stronger correlation with the fatness index, BMI, than with the fat distribution index, waist/hip ratio.

Adolescent↗
PubMed1995

The influence of body composition on left ventricular mass and other echocardiographic and Doppler measurements in 70-year-old males.

Heart morphology and function are often related to body size. The most common way to standardize for body composition is to divide these measures by body surface area (BSA). It has, however, been suggested that left ventricular mass (LV mass) ought to be indexed for height and that cardiac output is influenced by the body fat distribution. As part of a health screening programme of 70-year-old males in Uppsala, Sweden, 100 consecutive men were investigated with echocardiography and Doppler with regard to cardiac morphology and function. These measures were related to an evaluation of body composition using an X-ray method (dual energy X-ray absorptiometry). Lean body mass was strongly related to BSA and body weight (r = 0.90 and 0.89, respectively; P < 0.0001), and to a lesser degree to height (r = 0.70, P < 0.0001). In the healthy part of the population (n = 39), relationships between LV mass on the one hand and lean body mass, BSA and height on the other were seen (r = 0.51, 0.51 and 0.50, respectively; P < 0.01). Stroke volume calculated according to Teichholtz (but not when calculated with Doppler) showed a significant correlation to lean body mass (r = 0.38, P < 0.05). Also, left ventricular diameter and the thickness of the free wall were related to lean body mass (r = 0.40 and 0.39, respectively; P < 0.05). All of the evaluated relationships were generally weaker and in most cases no longer significant when evaluated in part of the population with diseases known to affect the heart. No significant correlations were seen between the size of the left atrium, septal thickness or indices of diastolic function (E/A-ratio, isovolumic relaxation time, deceleration time) and the indices of body composition. No measures of cardiac morphology and function were significantly related to the amount of body fat or body bone. LV mass, stroke volume and left ventricular diameter were found to be correlated with lean body mass, but not with the amount of fat, in healthy 70-year-old males. Theoretically, it is optimal to adjust for lean body mass, but in clinical practice, BSA is to be preferred.

Aged↗
PubMed2004

Adiposity and dietary intake in cardiovascular risk in an obese population from a Mediterranean area.

The aim of the study was to determine the particular relevance of android fat distribution and dietary intake in cardiovascular risk in an obese Mediterranean population with high intake of monounsaturated fatty acids (MUFA) and to compare the findings with those from normal-weight subjects. For the study, 193 subjects aged 25-60 were selected: 118 obese (BMI > or = 27 kg/m2), and 75 normal-weight (BMI < 25 kg/m2). Cardiovascular risk factors including hypertension, dyslipidaemia, glucose intolerance and insulin resistance were assessed. Nutrient intake and body fat distribution were determined. Results show that MUFA were highly consumed in the total population (21% of total energy). The obese population was normolipidemic and normoinsulinemic. However, cardiovascular risk factors (CVRF) were significantly higher than in normal-weight (P < 0.05). Obese subjects derived a greater percentage of their energy intake from total fat and lower from carbohydrates and saturated fats (P < 0.05). BMI and waist-hip ratio positively correlated with fat percentage of total energy intake and with MUFA (g/100 g fatty acids) in men, indicating that the excess of fat intake in obesity is due to a larger consumption of olive oil. CVRF were significantly and positively associated to waist circumference and WHR, both in obese and in normal-weight subjects. In conclusion, not only obesity but also android fat in normal-weight subjects are important factors in cardiovascular disease even in the Mediterranean population, with a high intake of MUFA, where these factors seem to be more relevant to cardiovascular risk than dietary composition.

Adipose Tissue↗
PubMed2006

CYP19A1 genetic polymorphisms may be associated with obesity-related phenotypes in Chinese women.

OBJECTIVE: To examine the relationship between genetic polymorphisms of the CYP19A1 gene and obesity-related phenotypes, body mass index (BMI) and waist-to-hip ratio (WHR). SUBJECTS: In total, 1241 Chinese women, who were recruited as community controls for a population-based case-control study of breast cancer. METHODS: Nineteen haplotype tagging single nucleotide polymorphisms (htSNPs) in four haplotype blocks were genotyped. RESULTS: Significant associations were observed for WHR at three SNPs that are located in haplotype block 1, including rs2445765, rs1004984 and rs1902584 (P=0.05, 0.04 and 0.01, respectively). Women, particularly premenopausal women, who carried the minor allele at any of these SNPs, had higher WHR than those without it. Of these three SNPs, the strongest association was observed at rs1902584, which is the closest to Promoter I.4, the major promoter for adipose tissue. Haplotype analyses indicated an association between the haplotype TCCAT in block 1 and WHR with a P-value of 0.02. CONCLUSION: These results suggested that CYP19A1 genetic polymorphisms may be associated with the risk of obesity among Chinese women, especially among premenopausal women. The CYP19A1 protein (aromatase) plays a critical role in estrogen biosynthesis and thus affects body fat distribution and regulation.

Adult↗
PubMed1996

Association of sex hormones and adiposity with plasma levels of fibrinogen and PAI-1 in postmenopausal women.

Blood levels of the clotting factor fibrinogen and tissue plasminogen activator inhibitor-1 (PAI-1), a primary inhibitor of fibrinolysis, have been positively linked to risk of coronary heart disease. The authors have reported previously that plasma fibrinogen appears to rise after menopause and to be reduced with use of postmenopausal hormonal therapy. There is also evidence to suggest that sex hormones may influence PAI-1. To examine whether plasma fibrinogen and PAI-1 antigen levels differ among older postmenopausal women according to use of hormone therapy and by blood level of estrogen and androgens, these variables were assessed among 277 healthy women aged 65-82 years, one half of whom were receiving therapy. The study population was drawn from the Study of Osteoporotic Fractures, Pittsburgh, Pennsylvania, during 1986-1988. Overall, results showed median PAI-1 levels to be lower on average with oral and transdermal use of hormone therapy (25.0 vs. 33.5 ng/ml, p < 0.01) and mean fibrinogen levels to be lower (279 vs. 295 mg/dl, p < 0.02) with use of oral estrogen (but not transdermal) therapy compared with women not receiving therapy. Among women not receiving therapy, PAI-1 and fibrinogen levels were not related to endogenous sex hormone levels, with the exception of a modest positive relation between PAI-1 and serum estrone concentrations (rs = 0.29). In addition, a markedly higher PAI-1 level was found for women with a preponderance of upper body fat, independent of obesity. In sum, results showed that older women receiving postmenopausal hormone therapy had more favorable plasma levels of the hemostatic factors PAI-1 and fibrinogen than did those not receiving therapy, which can be explained in large part by differences between the two groups in obesity and body fat distribution.

Adipose Tissue↗
PubMed1993

Differential health benefits of weight loss in upper-body and lower-body obese women.

Upper-body obesity (UB Ob) is more strongly associated with adverse health consequences; however, few obesity-treatment studies have examined outcome according to body-fat distribution. To examine whether diet and formal- or informal-exercise instruction causes differential changes in health and lipid profiles, ten LB Ob and nine UB Ob premenopausal women received dietary intervention (2.1 MJ-deficit/d for 16 wk) and were randomly assigned to either formal- or informal-exercise instruction. Weight loss was similar between groups (approximately 8 kg), and no change occurred in lean body mass or basal metabolic rate. Baseline cholesterol and triglycerides were greater (P < 0.01) in UB Ob than LB Ob women and decreased more (P < 0.01) in response to treatment in UB Ob women. Formal exercise instruction increased high-density-lipoprotein cholesterol (P < 0.05) especially in UB Ob women. Future studies on treatment of obesity should include consideration of regional fat distribution.

Adipose Tissue↗
PubMed1992

Determinants of total high density lipoprotein cholesterol and high density lipoprotein subfraction levels among Hispanic and non-Hispanic white persons with normal glucose tolerance: the San Luis Valley Diabetes Study.

Determinants of total high-density lipoprotein cholesterol (HDL-C) and HDL subfractions were assessed in Hispanic and non-Hispanic white persons (n = 932), aged 20-74 years, in the San Luis Valley, Colorado. Using multiple regression, BMI was negatively associated with HDL-C, HDL2-C, and HDL3-C in men and HDL-C and HDL3-C in women. Among females, current smokers had lower HDL-C and subfractions. Women on beta-blockers had lower HDL3-C levels. For both sexes, a positive association was observed between age and HDL-C and subfractions and physical activity with HDL-C and HDL3-C. Drinking alcohol (> or = 50 g/week) was associated with higher HDL-C and HDL3-C in both sexes and HDL2-C in women. The positive association of age and negative associations of the subscapular/triceps ratio and fasting insulin had consistent relationships with HDL-C, HDL2-C, and HDL3-C in men and women. Ethnicity was not significantly associated with HDL-C or subfractions after controlling for body fat distribution or fasting insulin.

Adrenergic beta-Antagonists↗
PubMed1999

The natural history of the metabolic syndrome in young women with the polycystic ovary syndrome and the effect of long-term oestrogen-progestagen treatment.

OBJECTIVE: Little is known about the natural history of polycystic ovary syndrome (PCOS), although preliminary data indicate that affected women are more susceptible than the general population to diabetes and cardiovascular diseases at post-menopausal ages. The aim of this study was to follow-up all main features of the metabolic syndrome in a group of young women with PCOS and to investigate the long-term effects on metabolism and body composition of oestrogen-progestagen (OP) compounds, which are frequently used in these women to treat hyperandrogenism and related clinical features. DESIGN: Long-term follow-up study. SUBJECTS AND METHODS: Thirty-seven women with PCOS were re-evaluated 10.3 +/- 0.8 years (range 6-18 years) after their first assessments (age: before 19.8 +/- 4.9 years; after 29.9 +/- 4.4 years). When first examined, women were instructed to follow a hypocaloric diet if they were obese plus OP, if they agreed to such treatment. Main anthropometric parameters, basal sex hormones and lipids, fasting and glucose-stimulated glucose and insulin levels and several clinical data were recorded before and after follow-up. RESULTS: In the whole group of women with PCOS we found no changes in body weight and fat mass, whereas both the waist-to-hip ratio and the waist-to-thigh ratio were significantly reduced. No significant changes occurred in mean fasting and glucose-stimulated glucose and insulin concentrations, whereas a significant increase in high-density lipoprotein-cholesterol was found. No significant changes occurred in testosterone levels. During the follow-up period 16 women took OP for an average of 97 +/- 18 months (range 12-180 months) (OP-users) whereas 21 women never took OP (non-OP-users). All OP-users were still taking OP when re-evaluated at the follow-up examination. With respect to baseline values, body mass index was higher in non-OP-users than in their counterparts. Waist circumference (P < 0.025), the waist-to-hip (P < 0.05) and the waist-to-thigh (P < 0.01) ratios decreased significantly only in the OP-users. In addition, percentage changes in waist circumference (P < 0.05) and waist-to-hip ratio (P < 0.05) during the follow-up period were significantly different between the groups. Glucose tolerance (as area under the curve (AUC)) improved (P < 0.05) in OP-users but not in non-OP-users. Moreover, compared to baseline values, basal insulin levels were significantly (P < 0.01) reduced in OP-users but not in non-OP-users. On the contrary, no significant change was found in insulinAUC in the former, whereas it significantly increased (P < 0.05) in the latter. Accordingly, fasting C-peptide decreased (P < 0.05) in OP-users, whereas both fasting (P < 0.01) and stimulated (P < 0.01) C-peptide significantly increased in non-OP-users. Changes in fasting or stimulated insulin and C-peptide in non-OP-users were not associated with parallel changes in testosterone levels. Total cholesterol and triglycerides did not change in either group, but HDL-cholesterol increased (P < 0.05) only in OP-users. Sex hormone-binding globulin concentrations increased significantly (P < 0.01) in OP-users, without any significant change in non-OP-users. Testosterone concentrations did not change significantly in either group, but the testosterone: SHBG ratio significantly decreased in OP-users (P < 0.05) but not in the non-OP-users. Among the clinical features, acanthosis nigricans significantly (P < 0.01) worsened in non-OP-users but not in the OP-users, without any significant change in the hirsutism and acne scores. Pregnancy rates during the follow-up were similar in both groups. CONCLUSIONS: These data indicate that hyperinsulinaemia and insulin resistance tended to worsen spontaneously in women with PCOS, without any worsening of the hyperandrogenism. Long-term oestrogen-progestagen treatment countered this tendency, probably because it improved the pattern of body fat distribution, by reducing abdominal fat depots.

Adolescent↗