Increased intake of fruits and vegetables in overweight subjects: effects on body weight, body composition, metabolic risk factors and dietary intake.

A diet rich in fruits and vegetables has been associated with several health benefits. However, the effects on body weight (BW) and metabolic markers are not fully known. The present study investigated the effects of increased intake of fruits and vegetables in overweight and obese men and women on dietary habits, anthropometry and metabolic control. In a 16-week controlled intervention, thirty-four men and thirty-four women aged 35-65 years (BMI>27 kg/m2) were randomised to an intervention (IN) or a reference (RG) group. All participants received general dietary advice, and subjects in the IN group received fruits and vegetables for free, of which ≥500 g had to be eaten daily. BW, waist circumference (WC), sagittal abdominal diameter (SAD), plasma insulin, blood glucose, glycated Hb (HbA1c), serum lipids, blood pressure, plasminogen activator inhibitor-1 activity, urinary isoprostane (iso-8-PGF 2α) and serum carotenoids were measured. Diet was assessed using 3-d weighed food records. In all, thirty subjects in the IN group and thirty-two in the RG group completed the intervention. Intake of fruits and vegetables doubled in the IN group, whereas intake of fruits increased in the RG group. Serum α- and ß-carotene concentrations and intakes of folate and vitamin C increased significantly in the IN group. Energy intake, BW, WC and SAD decreased significantly in both groups. Supine systolic blood pressure decreased significantly in the IN group, with no between-group differences. No significant changes were observed for other metabolic markers. Provision of fruits and vegetables led to substantially increased intakes, with subsequent favourable changes in anthropometry and insulin levels, which tended to be more pronounced in the IN group. The observed improvements may, in combination with improved nutritional markers, have health benefits in the long term.

A gene variation (rs12691) in the CCAT/enhancer binding protein α modulates glucose metabolism in metabolic syndrome.

BACKGROUND AND AIMS: CCAAT/enhancer-binding protein alpha (CEBPA) is a transcription factor involved in adipogenesis and energy homeostasis. Caloric restriction reduces CEBPA protein expression in patients with metabolic syndrome (MetS). A previous report linked rs12691 SNP in CEBPA to altered concentration of fasting triglycerides. Our objective was to assess the effects of rs12691 in glucose metabolism in Metabolic Syndrome (MetS) patients. METHODS AND RESULTS: Glucose metabolism was assessed by static (glucose, insulin, adiponectin, leptin and resistin plasma concentrations) and dynamic (disposition index, insulin sensitivity index, HOMA-IR and acute insulin response to glucose) indices, performed at baseline and after 12 weeks of 4 dietary interventions (high saturated fatty acid (SFA), high monounsaturated fatty acid (MUFA), low-fat and low-fat-high-n3 polyunsaturated fatty acid (PUFA)) in 486 subjects with MetS. Carriers of the minor A allele of rs12691 had altered disposition index (p = 0.0003), lower acute insulin response (p = 0.005) and a lower insulin sensitivity index (p = 0.025) indicating a lower insulin sensitivity and a lower insulin secretion, at baseline and at the end of the diets. Furthermore, A allele carriers displayed lower HDL concentration. CONCLUSION: The presence of the A allele of rs12691 influences glucose metabolism of MetS patients.

Conservative treatment of obesity in an academic obesity unit. Long-term outcome and drop-out.

OBJECTIVE: The aim of this study was to explore the outcome and the problems of drop-out in the treatment of obese outpatients at an academic obesity unit. DESIGN: A two-year clinical treatment evaluation. SUBJECTS: A total of 117 obese subjects, 83 women and 34 men, mean aged 50 (23-70) years, with an average body mass index (BMI) of 39.0 kg/m2 (28.8- 64.7). INTERVENTION: All treatment was based on group therapy and included behaviour modification and nutrition counselling. A team of nurses, dieticians, a physiotherapist, a psychotherapist and a physician supervised the treatment. Two programmes were used. Group 1 initially received a low-calorie diet (LCD) for seven weeks combined with the behaviour treatment programme. Group 2 was treated with the behaviour treatment programme only. All subjects were offered complementary treatment according to their medical needs. RESULTS: There was a continuous drop-out of subjects during the two-year treatment period with an overall drop-out rate of 53%. Anthropometric characteristics, medical history or reasons for drop-out had no impact on the drop-out rate. In completers the weight reduction after two years was 9.2 [+/-10.8 standard deviation (S.D.) kg. In non-completers the weight reduction of the last observed weight measurement was 4.7 (+/-7.9 S.D.) kg. After year two, the weight reduction in Group 1 was 8.8 (+/-12.2 S.D.) kg, and in Group 2 was 9.7 (+/-8.0 S.D.) kg. CONCLUSION: This study has showed the difficulties of long-term clinical treatment of obese outpatients, even in a specialised obesity clinic. The findings demonstrate that educated and experienced staff together with an extended package of treatment options is not enough to keep patients in treatment for two years. However though the drop-out rate was high, two thirds of the included subjects reduced their weight, which is a satisfactory result in a clinical setting. The drop-out rate and the reasons for dropping out could give a clue in which direction the diagnostics and analysis of the subject's individual needs in health care should be directed.

Glucose and insulin responses in healthy women after intake of composite meals containing cod-, milk-, and soy protein.

OBJECTIVE: To evaluate the metabolic effect of three different kinds of dietary proteins as part of composite meals with similar macronutrient composition in healthy subjects. DESIGN: A randomised meal study. SETTING: Metabolic ward. SUBJECTS AND METHODS: In total, 17 healthy women, 30-65 years old, consumed three meals in randomised order. The meals consisted of foodstuffs with similar nutrient composition but different types of protein (cod, cottage cheese, or soy protein isolate). The distribution of energy from protein, fat and carbohydrates was 33, 26, and 41 energy percent, respectively. Total amount of energy was 2300 kJ. Blood samples were drawn for assay of B-glucose, S-insulin, S-free fatty acids, S-triglycerides, and C-peptide in the fasting state and at seven times (20, 40, 60, 90, 120, 180, and 240 min) after starting to eat the test meal. RESULTS: The blood glucose response after the cod protein meal differed from that of the soy protein meal, with a larger area under the curve (AUC) calculated up to 120 min. The serum insulin response after the milk protein meal differed from that of the cod protein meal with a larger AUC calculated up to 240 min. The insulin/C-peptide and the insulin/glucose ratios differed between the meals; the insulin/C-peptide ratio was higher after the milk protein meal compared to the cod, and soy protein meal at 120 min. The insulin/glucose ratio was lower after the cod protein meal compared to the milk, and soy protein meals at 120 min. The results showed that the metabolic responses differed after meals with similar macronutrient composition containing cod-, milk-, or soy protein.

Blood flow is an important determinant of forearm glucose uptake following a mixed meal.

Insulin-mediated vasodilation has been suggested to be of importance for glucose uptake during normoglycemic hyperinsulinemia. If this also is valid after an ordinary mixed meal remains to be evaluated. Forearm blood flow (FBF) and forearm glucose uptake change (evaluated by venous occlusion plethysmography) and glucose arteriovenous differences were evaluated over 120 minutes in 10 healthy volunteers following an ordinary mixed meal (700-900 kcal, 34% of energy from fat). Fasting arterial glucose level was 4.9+/-0.9 mmol/l, and the maximum glucose level was reached 30 minutes after the start of ingestion (6.6+/-0.8 mmol/l, p<0.0001). Plasma insulin levels were increased four-fold. FBF increased rapidly within 20 minutes after the start of ingestion and reached its maximum after 50 minutes (94% higher than baseline level, p<0.01). After 2 hours FBF was still substantially elevated (75% above baseline level, p<0.01). Forearm glucose uptake increased fivefold already after 20 minutes ( p<0.01). During the 2 hours, the increase in FBF contributed to 41% of the forearm glucose uptake ( p<0.05). The present study showed that the increase in FBF seen after an ordinary mixed meal is important for the change in forearm glucose uptake. These results support the view that modulation of limb blood flow is a determinant of glucose uptake.

A programme of behaviour modification and nutrition counselling in the treatment of obesity: a randomised 2-y clinical trial.

OBJECTIVE: To compare two group treatment programmes for obese outpatients. Both programmes included behaviour modification, nutrition counselling, very-low-calorie diet (VLCD) and a continuous measuring of metabolic and anthropometrical status, but they differed regarding the treatment intensiveness. The main aim was to study whether intensive treatment gives a larger weight reduction compared with less intensive treatment and what level of input from health care personnel is needed to reach adequate treatment results. DESIGN: A 2-y randomised clinical trial. SUBJECTS: A total of 43 obese subjects aged 24-60 y, BMI 35 kg/m(2) (29-48). INTERVENTION: Two programmes were used. Both were based on group therapy and were supervised by a dietitian and a psychologist. Group 1 received a continuous intensive treatment with planned group meetings every fortnight during the first year and six group meetings the second year. Group 2 had planned group meetings every third month. Anthropometrical and metabolic data were measured every third month in both groups. The VLCD periods were the same. RESULTS: There was no evidence that a more intensive treatment promotes a larger weight reduction. Weight reduction after 1 y: group 1, -7.6 (+/-0.97) kg, BMI -2.6 (+/-0.3) kg/m(2); group 2, -6.4 (+/-1.16) kg, BMI -2.2 (+/-0.4) kg/m(2). Weight reduction after 2 y: group 1, -6.8 (+/-1.4) kg, BMI -2.4 (+/-0.3) kg/m(2); group 2, -8.6 (+/-1.6) kg, BMI -3.0 (+/-0.3) kg/m(2). The dropout rate was 26%. CONCLUSION: There were no significant differences in weight reduction, compliance or dropout rate between the groups and there was no evidence that a more intensive treatment promotes a larger weight reduction. This observation is of value when setting up treatment programmes. To measure the metabolic and anthropometrical status during the treatment and to give continuous feedback to the subjects seem to be important factors for compliance. Both treatment programmes gave highly significant weight reductions in the range of 5-10%, which has been referred to as a realistic goal for the treatment of obese patients.

Changes in food habits in healthy Swedish adolescents during the transition from adolescence to adulthood.

OBJECTIVE: To investigate the change in food habits in Swedish adolescents between 15 and 21 y of age with reference to age, sex, region and socioeconomic background. DESIGN: A longitudinal study from 1993 to 1999. SETTING: Two different regions in Sweden, the university city of Uppsala and the industrial town of Trollhättan. SUBJECTS: On three different occasions, 1993, 1995 and 1999, 208 adolescents, 96 males and 112 females, were studied. METHODS: A food frequency questionnaire containing 29 different food groups was used. The questionnaire also contained questions about food habits and amounts of some food items and socioeconomic conditions of the participants and their families. RESULTS: At 17 and 21 y of age, the adolescents consumed significantly more often pasta, vegetables, coffee and tea compared to age 15, while the frequency consumption of fat spread, milk, bread, potatoes, carrots and buns and biscuits decreased. The changes between 15 and 17 were smaller than between age 17 and 21. At age 21, the males decreased their intake of fruit, while the females decreased their intake of meat. No-meat consumers among females increased from 2 to 13%. Higher educational level of the mothers of the adolescents was associated with more frequent consumption of vegetables and pasta between ages 17 and 21. Milk consumption decreased significantly in both sexes. Breakfast habits did not change: 90% had breakfast five times/week or more. CONCLUSIONS: Food habits change significantly during adolescence along with lifestyle changes. Therefore, health promotion during adolescence ought to be more supported by the society.

An ordinary mixed meal transiently impairs endothelium-dependent vasodilation in healthy subjects.

This study was designed to evaluate the effects of an ordinary mixed meal on endothelium-dependent vasodilation. Ten young healthy volunteers were given a mixed meal (minced meat sauce with rice, 900 kcal, 34% of the energy content was fat). In the fasting state, at 60 and 120 min after the start of the meal, endothelium-dependent vasodilation and endothelium-independent vasodilation were evaluated by local infusion of metacholine (4 microg min (-1)) and sodium nitroprusside (10 microg min (-1)) in the brachial artery. Blood flow in the forearm was measured using venous occlusion plethysmography. Endothelium-dependent vasodilation decreased from 15.4 +/- 3.3 (mean +/- SD) at fasting to 13.7 +/- 3.5 mL min (-1) (100 mL tissue)-1 (P < 0.01) 60 min after feeding, but had returned to the fasting level at 120 min. At 60 min, but not in the fasting state, the serum level of free fatty acids was inversely related to endothelium-dependent vasodilation (r=-0.74, P < 0.05), although no significant net changes in FFA levels were seen. Endothelium-independent vasodilation was not affected by the mixed meal. No similar attenuations in endothelium-dependent vasodilation were seen during control meals. In conclusion, an ordinary mixed meal transiently attenuated endothelium-dependent vasodilation. Free fatty acids may be involved in this effect on endothelial function.

Diet, nutrition and diabetes mellitus.

Nutritional management of diabetes mellitus, and the importance of diet in the development of insulin resistance, have for many years been important areas of research and education at the Unit for Clinical Nutrition Research at the Department of Public Health and Caring Sciences (formerly Department of Geriatrics) at Uppsala University. The research has more recently focussed on effects of dietary fat quality in the development of insulin resistance and in treatment of diabetes, on interaction between dietary fat and physical activity in relation to insulin sensitivity and on the importance of carbohydrate rich foods with low glycaemic index in the diabetic diet. Much work has also been directed towards development of educational material about nutrition recommendations and dietary treatment in diabetes mellitus. The ultimate goals for all our efforts are to visualize, and promote, the possibilities and fundamental importance of lifestyle changes. This includes an improved diet and increased physical activity, in the prevention and treatment of diabetes mellitus.

Reliability of anthropometric measurements in overweight and lean subjects: consequences for correlations between anthropometric and other variables.

OBJECTIVE: To estimate the reliability of anthropometric measurements in overweight and lean subjects, and to examine the influence of this reliability on correlations to other variables, since low reliability leads to underestimation of correlations. DESIGN: Replicate measurements by two observers in 26 overweight and 25 lean subjects measured at two occasions. MEASUREMENTS: Sagittal abdominal diameter (SAD), waist circumference (waist), waist-to-hip ratio (W/H) and skinfold measurements. RESULTS: Intra-class correlation coefficients (ICCs) for SAD and waist were higher than for W/H (0.98 vs. 0.90, P<0.001, and 0.97 vs. 0.90, P = 0.001, respectively). For waist, the ICC was lower for overweight than for lean subjects (0.85 vs 0.95, P=0.030), but the ICC values were comparable for SAD and W/H (0.92 vs. 0.95 and 0.78 vs. 0.83, respectively). Intra-observer variations (IOV) for SAD and waist were lower than for W/H (coefficients of variation; 1.6%, 1.4% and 2.3%, respectively), as were intra-subject variations (ISV) (2.7%, 3.0% and 3.4%, respectively). ICC values ranged from 0.84 to 0.93 and were lower for overweight than for lean subjects for biceps, subscapular and umbilical skinfolds (P=0.031, P<0.001 and P=0.048, respectively). Coefficients of variations for skinfold measurements ranged between 7.3% and 16.0% for IOV and between 14.9% and 20.8% for ISV. CONCLUSIONS: The low ICC values imply that correlations can be underestimated in overweight groups. We propose that, because of their higher reliability, SAD and waist have a higher predictive capacity for cardiovascular risk than W/H. SAD is the only measurement with high reliability in both weight groups and its use is recommended.

Improved glycemic control and lipid profile and normalized fibrinolytic activity on a low-glycemic index diet in type 2 diabetic patients.

OBJECTIVE: To evaluate the effects of varying the glycemic index (GI) of carbohydrate-rich foods on metabolic control in type 2 diabetic patients. RESEARCH DESIGN AND METHODS: In a randomized crossover study, 20 patients, 5 women and 15 men, were given preweighed diets with different GIs during two consecutive 24-day periods. Both diets were composed in accordance with dietary recommendations for people with diabetes. The macronutrient composition and type and amount of dietary fiber were identical. Differences in GI were achieved mainly by altering the structure of the starchy foods. RESULTS: Peripheral insulin sensitivity increased significantly and fasting plasma glucose decreased during both treatment periods. There was a significant difference in the changes of serum fructosamine concentrations between the diets (P < 0.05). The incremental area under the curve for both blood glucose and plasma insulin was approximately 30% lower after the low- than after the high-GI diet. LDL cholesterol was significantly lowered on both diets, with a significantly more pronounced reduction on the low-GI diet. Plasminogen activator inhibitor-1 activity was normalized on the low-GI diet, (-54%, P < 0.001), but remained unchanged on the high-GI diet. CONCLUSIONS: A diet characterized by low-GI starchy foods lowers the glucose and insulin responses throughout the day and improves the lipid profile and capacity for fibrinolysis, suggesting a therapeutic potential in diabetes.

The influence of food structure on postprandial metabolism in patients with non-insulin-dependent diabetes mellitus.

Two natural-food mixed meals composed in accordance with the present dietary recommendations were given to 10 diabetic patients on two occasions. The meals were planned to achieve large differences in glycemic index (GI) by altering the food structure but maintaining an identical nutrient composition. In the first study, two meals were compared that contained either pasta or bread made from durum wheat flour. In the second experiment a meal with parboiled rice, red kidney beans, and bread made from whole-wheat grains was compared with a meal of sticky rice, ground red kidney beans, and bread made from ground wheat. All of the other ingredients in these meals were identical. The area under the curve for blood glucose (BG) and plasma insulin (PI) was significantly lower after the meal with pasta (BG: -35%, P < 0.05; PI: -39%, P < 0.05) and the meal with parboiled rice (BG: -42%, P < 0.001; PI: -39%, P < 0.01), respectively, compared with the corresponding high-GI meals. The results shows the importance of preserved structure in common foods and support the applicability of the GI concept to mixed meals.

Polyunsaturated fatty acids may impair blood glucose control in type 2 diabetic patients.

Fifteen patients with Type 2 diabetes were given two diets rich in either saturated fat or polyunsaturated fat in alternate order over two consecutive 3-week periods on a metabolic ward. Both diets contained the same amount of fat, protein, carbohydrates, dietary fibre, and cholesterol. The proportions of saturated, monounsaturated and polyunsaturated fatty acids in the saturated fat diet were 16, 10, and 5%-energy and in the polyunsaturated fat diet (PUFA) 9, 10, and 12%-energy. The PUFA diet contained a high proportion of n-3 fatty acids. Metabolic control improved significantly in both dietary periods, due to both qualitative dietary changes and a negative energy balance. The serum lipoprotein concentrations decreased on both diets but the serum lipids were significantly lower after the PUFA diet (serum triglycerides -20%, p = 0.001; serum cholesterol -5%, p = 0.03; VLDL-triglycerides -29%, p less than 0.001; and VLDL-cholesterol -31%, p = 0.001) than after the saturated fat diet. Average blood glucose concentrations during the third week were significantly higher fasting (+15%, p less than 0.01), and during the day at 1100 h (+18%, p less than 0.001) and 1500 h (+17%, p = 0.002) on PUFA than on the saturated fat diet. Significantly higher blood glucose levels were also recorded with a standard breakfast, while the sum of the insulin values was lower (-19%, p = 0.01). HbA1c did not differ significantly between the two dietary periods.(ABSTRACT TRUNCATED AT 250 WORDS)

[Diet treatment in clinical practice. Advice on diet--more than a knowledge of dietetics]. / Kostbehandling i praktiken. Råd om mat--mer än kunskap i näringslära.

Eating habits are deeply rooted and founded early in life. The need to change one's eating habits in order to treat a certain disease or a metabolic disorder may seem to impose a well nigh impossible task. It is therefore extremely important that the dietitian adjust the recommendations to each individual's needs, wishes and circumstances. The article translates nutritional objectives into practical advice. The desired dietary composition is essentially the same for the foods which are recommended for the treatment of diabetes, hyperlipoproteinemia and hypertension.

Dietary habits in renal stone patients compared with healthy subjects.

The dietary habits of renal calcium stone patients were investigated both by dietary history and by 4-day food records and compared with the dietary habits of control subjects, matched on the basis of age, sex, social and professional status. The method using 4-day records seemed to be more precise, judged by the correlation to the urinary output of nutrients. There was no difference in the daily intake of major nutrients between stone formers and controls, but a higher ingestion of vitamin C in controls and a larger consumption of alcohol among stone formers. In contrast to some epidemiological evidence, there were no significant differences in consumption of animal protein when stone formers were matched for social class. Despite a similar total intake of calcium, the stone formers excreted more calcium in the urine, probably reflecting a higher intestinal absorption of calcium. There seem to be only marginally different dietary habits between stone formers and carefully matched control subjects. Differences in the urinary output of minerals and electrolytes are mainly due to variations in gastrointestinal uptake.

Dietary habits and effects of dietary advice in patients with type 2 diabetes. Results from a one-year intervention study.

The diets of 22 patients with type 2 diabetes were evaluated by a dietary-history interview. Sixteen of the patients were also interviewed after a 1-year intervention period, to find out whether continuous contact with a dietitian at a health centre could influence the food habits and thereby improve the metabolic control. In most participants the diabetes was fairly well controlled on entry into the study. Another 24 patients took part in the same intervention programme but did not agree to participate in the dietary-history interview. These subjects had, in general, a poorer metabolic control on admission. According to the study design, the patients were to make four visits to the dietitian during the first 6 months, and one visit at the end of the year. The patients were advised to eat a fat-modified, fibre-enriched diet. Among the patients who completed both the intervention programme and the dietary-history interviews there were significant reductions in mean body weight, haemoglobin A1c and serum triglycerides during the first half of the study. After 12 months, however, only the mean body weight and serum triglycerides remained lower than before the intervention period. A study of the fatty acid composition of the plasma cholesterol esters showed a significant reduction of the content of oleic acid (18:1 n-9). The dietary interviews indicated only minor changes (not significant) in the intake of energy and different nutrients, including dietary fibre, after 1 year. However, the effects of the intervention were better in the patients who did not participate in the dietary interview with significant reductions in the mean body weight, fasting blood glucose, haemoglobin A1c and serum triglycerides also after 1 year.

Effects of four meals with different kinds of dietary fibre on glucose metabolism in healthy subjects and non-insulin-dependent diabetic patients.

The effects of four mixed meals on glucose metabolism were studied during a 4-h period in 12 healthy women and in 13 women with non-insulin-dependent diabetes. Three test meals containing different types of dietary fibre in realistic amounts (cereal, leguminous and mixed-fibre), and one control meal were prepared. Each meal was calculated to contain 2.1 MJ (500 kcal) and the energy distributions of protein, fat and carbohydrate (by difference) were 15, 29-31 and 56-58 per cent, respectively. Results of analyses of the mixed meals, performed after the clinical study, showed unexpectedly large deviations from the calculated values regarding digestible carbohydrates in the leguminous meal. In the healthy group the blood glucose concentrations after the four meals did not differ significantly. In the diabetic group the area under the curve of blood glucose values was significantly smaller after both the leguminous meal and that containing mixed dietary fibre than after the control meal (P less than 0.001). There were also significant differences in blood glucose between the cereal meal and the leguminous and mixed-fibre meals (P less than 0.001), and between the leguminous meal and the mixed-fibre meal (P less than 0.05). In the healthy group the insulin response was significantly lower after the leguminous meal than after the control meal (P less than 0.05) whilst the diabetic group showed lower insulin responses after all the high-fibre test meals. Serum triglycerides, cholesterol and HDL cholesterol were similar after the different meals. The lower content of digestible carbohydrates, especially in the leguminous meal, complicates the interpretation of the results.(ABSTRACT TRUNCATED AT 250 WORDS)

Effects of leguminous seeds in a mixed diet in non-insulin-dependent diabetic patients.

The effects of moderate addition of leguminous seeds to a mixed diabetic diet were evaluated in 15 inadequately controlled patients with non-insulin-dependent diabetes mellitus. A control diet and a diet with an increased content of peas and beans (leguminous diet) were given consecutively for a 3-week period each on a metabolic ward. Seven of the patients started with the control diet and 8 with the leguminous diet. On an energy basis the diets comprised 20% protein and 33-34% fat, the ratio between polyunsaturated and saturated fatty acids was 1.4. The analysed dietary fibre contents of the control and leguminous diets were 24 and 37 g/day per 6.7 MJ (1,600 kcal) respectively. The content of digestible carbohydrates was similar in the 2 diets. The mean body weight did not differ after the 2 dietary treatment periods. All patients showed improved glucose control during treatment irrespective of the type of diet given. The fasting blood glucose concentration on admission was 14.3 +/- 0.9 mmol/l (mean +/- SEM). This value had decreased to 9.9 +/- 0.8 during the last week of the control period and to 9.7 +/- 0.6 during the last week of the leguminous diet period. The mean postprandial glucose concentration at 3 pm was significantly lower during the leguminous diet period. The mean urine glucose excretion was significantly lower during the leguminous diet period. The fasting serum lipid concentrations decreased during treatment with each diet, but there were no significant differences between these values at the end of the 2 diet treatment periods.(ABSTRACT TRUNCATED AT 250 WORDS)