Bone metabolism markers in adolescent girls with eating disorders and weight loss: effects of growth, weight trend, developmental and menstrual status.

UNLABELLED: Serum concentrations of osteocalcin (OC) decrease and those of C-terminal telopeptide of type 1 collagen (CTX) increase during weight loss in adolescent girls with eating disorders (ED). The impact of weight loss on bone metabolism markers is greatest in premenarcheal girls. INTRODUCTION: Adolescents with ED stand a risk of not reaching optimal peak bone mass and develop osteoporosis. Previous investigations are contradictory as to how markers of bone formation and resorption change during weight loss and nutritional rehabilitation. METHODS: Serum OC and CTX were measured at assessment of 461 adolescent girls with ED and during treatment of 55 girls with anorexia nervosa. Bone metabolism was related to weight, weight change and growth rate. RESULTS: At assessment, OC concentrations were positively correlated with growth rate and inversely with age and (rate of) weight loss. Growth rate was the only predictor of CTX concentrations in premenarcheal girls. In postmenarcheal girls, CTX concentrations were inversely correlated with age and rate of weight loss. During weight gain, there was an increase of OC concentrations. CTX concentrations decreased at the onset of weight gain and increased when near normal weight was reached. CONCLUSIONS: Bone formation markers decrease and resorption markers increase during weight loss. The effects are independent of menstrual status but the impact on bone formation markers is greater in young, premenarcheal girls. Markers are normalised during weight gain but it is conceivable that repeated and/or prolonged weight loss in adolescents reduces peak bone mass.

Essential fatty acid status in teenage girls with eating disorders and weight loss.

AIM: To explore the relationship between essential fatty acids (FA) and weight changes in adolescent girls with eating disorders (ED). METHODS: Blood samples were obtained from 220 girls with ED and 39 healthy controls. The girls with ED were 15.3 ± 1.5 years of age and weighed 49.8 ± 8.7 kg (BMI 18.3 ± 2.8 kg/m(2)) after a weight loss of 6.8 ± 6.4 kg. FA were analysed in plasma phospholipids (PPL) and erythrocyte membranes (ERY). RESULTS: The proportions of saturated and monounsaturated FA were increased during weight loss, while linoleic acid (18:2ω6) was decreased. The proportions of eicosapentanoic acid (EPA) (20:5ω3) and docosahexanoic acid (DHA) (22:6ω3) in PPL and ERY did not differ from controls. The activity of stearoyl-CoA-desaturase was increased as evidenced by an increased product/precursor ratio and correlated with the rate of weight loss. The activities of delta-6-desaturase and delta-5-desaturase did not differ from controls. The rate of weight loss was inversely correlated with delta-6-desaturase and directly correlated with delta-5-desaturase. CONCLUSION: The FA profile indicates low-fat intake, fat mobilization from stores and an increased conversion of essential FA at the delta-5-desaturase step during weight loss in adolescent girls with ED. Normal levels of EPA and DHA were maintained.

Experimental intrauterine growth retardation in the rat causes a reduction of pancreatic B-cell mass, which persists into adulthood.

BACKGROUND AND OBJECTIVES: The aim of this study was to investigate the possibility that intrauterine growth retardation (IUGR) causes alterations of glucose tolerance, insulin secretory response to glucose, and pancreatic B-cell growth, and if such changes may persist into adulthood. METHODS: Pregnant rats were operated on day 16 of pregnancy ad modum Wigglesworth to induce IUGR. Operated rats gave birth to viable offspring but litter size was reduced. The mothers nursed their pups, which were subsequently weaned and reared to an age of 3 months in apparent good health. RESULTS: At 1 day of age, IUGR pups were 10% lighter than control newborns whose mothers had been subjected to a sham operation. Pancreatic B-cell mass and insulin content were reduced by 35-40% in newborn IUGR offspring. Postnatal growth did not differ between IUGR and control animals of either sex and the difference in body weight at birth was not apparent from 1 week of age and onwards. Tests performed at 3 months of age could not demonstrate differences in glucose tolerance between IUGR and control animals. In females, but not in males, the peak insulin secretory response to glucose was lower in IUGR animals compared to controls. In the 3-month-old rats, B-cell mass was reduced by 40% in male and by 45% in female IUGR rats compared to controls, a reduction corresponding to a similar decrease in pancreatic insulin content (male reduction 48%, female reduction 45%). CONCLUSIONS: In the rat, IUGR causes a diminution of pancreatic B-cell mass which persists into adulthood. Normal glucose tolerance could be maintained but it is conceivable that increasing demands on insulin secretion may not be met by the reduced B-cell mass and that impaired glucose tolerance and even diabetes would hence develop.

Endogenous glucose production and lipolysis in anorexia nervosa--a study using stable isotope-labelled compounds.

UNLABELLED: Glucose production and lipolysis were investigated in teenage girls with anorexia nervosa using glucose and glycerol labelled with stable isotopes. The production of energy substrates were also maintained in the very underweight patients who showed higher relative rates of substrate mobilization compared to those who had gained some weight. CONCLUSION: The investigation shows that glucose production, necessary to provide fuel for the central nervous system, is also maintained in starving patients with anorexia nervosa.

Weight requirements for return of menstruations in teenage girls with eating disorders, weight loss and secondary amenorrhoea.

AIM: To investigate the weight requirements for return of menstruation in teenage girls with eating disorders (ED), weight loss and secondary amenorrhoea. METHODS: Growth charts from the school health services and measurements of weight and stature at assessment and during follow-up were obtained for 127 girls with ED, secondary amenorrhoea and subsequent return of menstruation. Measurements were used to estimate weight and body mass index (BMI) before puberty, at menarche, at the highest weight prior to the onset of the ED, at the last menstruation preceding amenorrhoea, at the lowest weight during treatment, and at return of menstruation. RESULTS: Before onset of the ED, the girls were taller, heavier and less lean than the population average as evidenced by standard deviation scores (SDS) for weight, height and BMI above zero. Weight loss started from an average weight of 58.9 +/- 9.8 kg (mean +/- SD), a last menstruation occurred at 51.5 +/- 6.9 kg, the lowest weight during treatment was 45.6 +/- 7.0 kg and menstruation returned at 52.9 +/- 6.0 kg. Return of menstruation occurred within a wide weight range. However, if weight at return of menstruation was expressed in SDS, it could be predicted by a linear regression on weight SDS at loss of menstruation (r2 = 0.76; p < 0.001). CONCLUSIONS: The weight level required for return of menstruation is highly individual but can be predicted by the weight at which menstruations cease. In the treatment of ED, there is a need for such individual weight targets--a target based on the population weight for height and/or age may be too generalized and too low.

The significance of routine laboratory analyses in the assessment of teenage girls with eating disorders and weight loss.

Routine laboratory investigations that had been performed at disease assessment on 327 teenage girls with eating disorders and weight loss were analyzed. The laboratory investigations included erythrocyte sedimentation rate (ESR), blood haemoglobin concentration (Hb), white blood cell count (WBC), platelet count, serum alkaline phosphatase (ALP) activity, serum aspartate aminotransferase (ASAT) activity, serum alanine aminotransferase (ALAT) activity, serum albumin concentration, glycated haemoglobin (HbA1c) and serum concentrations of sodium, potassium, magnesium, calcium (corrected for albumin), inorganic phosphate, creatinine and urea. The results were for ESR, Hb, WBC, platelet count, ALP, ASAT, ALAT, inorganic phosphate, creatinine, urea and HBA1C related to weight and (ongoing) weight loss. The variations of the biochemical measurements were, however, largely within reference ranges, weight and weight changes predicted the biochemical measurements only to a small degree and in individual patients the results of the analyses often suggested normality. These analyses may therefore not be suited to assess the degree of weight loss and starvation in eating disorders. They may, however, be useful for the exclusion of other diseases which could show weight loss and biochemical abnormalities.

Clinical onset and diagnosis of eating disorders in premenarcheal girls is preceded by inadequate weight gain and growth retardation.

AIM: To study growth and weight changes before the presentation of an eating disorder (ED) with premenarcheal onset. METHODS: Growth charts from the school health services were obtained for 45 girls assessed during the period 1990-2000 at Uppsala University Children's Hospital. Measurements of weight and height from the charts and at presentation were recalculated into standard deviation scores (SDS). RESULTS: At their maximal weight the girls were 12.5 +/- 1.7 (mean +/- SE) y old. They were then lighter, shorter and leaner than the general population mean, as evidenced by SDS below zero for weight (-0.43 +/- 1.08; mean +/- SD), height (-0.45 +/- 1.01) and body mass index (BMI) (-0.35 +/- 1.07). At presentation approximately 1 y later they had lost 5.8 +/- 4.3 kg and were considerably underweight (weight SDS -2.27 +/- 1.33) and further stunted (height SDS -0.76 +/- 0.97). The point on the growth curves with the highest SDS for weight was observed at 8.5 +/- 1.4 y of age. The girls were then heavier (weight SDS 0.35 +/- 0.93) and less lean (BMI SDS 0.42 +/- 0.97) than the population average. A tendency to track down through weight and height curves before the onset of weight loss was thus observed. Total weight deficit was as much as 31 +/- 10% of expected body weight. Analyses of weight and height deficits indicated that two-thirds of the weight deficit and 60% of the height deficit was generated before the onset of weight loss. CONCLUSION: Girls with eating disorders presenting before menarche may have a long history of poor weight gain and growth retardation before the onset of weight loss. This is in contrast to older girls, who commonly start weight loss at an above-average weight without prior poor weight gain. Since the psychopathology of ED in young girls may be different and less evident compared with older teenagers it is important to be aware that poor weight gain and growth retardation may be associated with early-onset ED.

Changes in body weight and body mass index (BMI) in teenage girls prior to the onset and diagnosis of an eating disorder.

UNLABELLED: Growth charts from the school health services were used to study weight changes prior to the onset and diagnosis of eating disorders (ED) with weight loss in 122 teenage girls. In these girls menarche started at 12.6 +/- 1.0 y (mean +/- SD), which is the same for the general population. At their maximal premorbid weight, the girls were 14.6 +/- 1.2 y of age and had a higher weight and body mass index (BMI) than the general population, as evidenced by standard deviation scores (SDS) above zero for weight (0.65 +/- 0.95, p < 0.001) and BMI (0.61 +/- 0.94, p < 0.001). Weight gain prior to the onset of weight loss followed two different patterns. Eighty-three (68%) girls tracked upwards through the weight curves to reach their maximal weight, thus putting on weight at a higher rate than expected. Thirty-nine (32%) girls tracked downwards through the weight curves to reach their maximal weight, but did so starting from a weight and BMI at well above average. CONCLUSIONS: Weight gain in teenage girls prior to the onset of an ED deviates from that of the general population. Being or becoming heavier and less lean than their peers could influence body image and decisions on dieting, which, in turn, may be the start of the development of an ED in the otherwise predisposed. The deviation from normal tracking patterns prior to the onset of disease also causes problems in diagnostic procedures and in setting treatment targets for weight gain, since "normal" or expected weight becomes more difficult to establish.

Heart risk associated with weight loss in anorexia nervosa and eating disorders: electrocardiographic changes during the early phase of refeeding.

Refeeding syndromes with electrolyte aberrations, heart failure and arrhythmias may complicate the nutritional rehabilitation of emaciated patients with eating disorders. Therefore, electrocardiographic (ECG) changes and changes in serum electrolyte concentrations following refeeding were studied in 37 admissions of 32 teenage girls with eating disorders. On admission they were all on a weight-losing course and weighed 37.0+/-8.0 kg (mean +/- SD) following a weight loss of 14.2+/-7.2 kg. On ECG recordings there was a prolongation of the QT interval and an increased QT dispersion. Serum concentrations of sodium, potassium and magnesium were with few exceptions normal. Serum concentrations of creatinine were high in relation to the low body weight, indicating protein catabolism. The first 2 wk of refeeding resulted in a weight gain of 1.7+/-0.2 kg without signs of refeeding syndromes or electrolyte aberrations. QT prolongation and dispersion normalized within the 3 d of refeeding. It is concluded that oral refeeding of patients with eating disorders and weight loss can be performed efficiently and without causing refeeding syndromes. QT pathology, a consequence of acute starvation and a risk factor for cardiac arrhythmias, normalizes within days. In view of the need to balance adequate refeeding and reduction of QT pathology against the risks of refeeding syndromes the start of refeeding of severely emaciated patients is best performed in a hospital setting where monitoring of ECG and serum electrolytes is possible.

Heart risk associated with weight loss in anorexia nervosa and eating disorders: risk factors for QTc interval prolongation and dispersion.

Risk factors for QTc interval prolongation and dispersion, indicators of an increased risk for cardiac arrhythmia and sudden death, have been investigated in patients with eating disorders (ED) and ongoing weight loss. Patients were characterized with regard to weight, body mass index (BMI; weight/length2), duration of weight loss, rate of weight loss and rate of weight loss immediately preceding examination. At examination, a 12-lead electrocardiographic (ECG) registration and blood samples for analysis of serum electrolytes were obtained. In total, 92 examinations in 58 female patients aged 15.5+/-1.7 (mean +/- SD) y were analysed. Control ECG recordings were obtained from 38 normal-weight teenage girls with no known heart disease. Patients with ED weighed 40.7+/-7.8 kg, corresponding to BMI 15.2+/-2.4 kg/m2 following a weight loss of 11.8+/-6.5 kg. In ED patients, the ECG showed bradycardia, a shift to the right of the QRS axis, diminished amplitudes of the QRS complex and T wave, and prolongation and increased dispersion of the QTc interval. In multiple regression analyses low weight, low BMI and rapid weight loss immediately preceding the examination were the most important independent predictors of QTc interval prolongation and dispersion. It is concluded that an ECG examination is an important part of the assessment of patients with ED and ongoing weight loss, even in the absence of electrolyte disturbances, and especially if the patient is severely underweight or weight loss is rapid.

Cyanide detoxification in rats exposed to acetonitrile and fed a low protein diet.

Different neurological syndromes have been associated with exposure to cyanide. Dietary cyanide exposure from cassava roots combined with a low intake of the sulfur amino acids necessary for cyanide detoxification has been implicated in the causation of konzo, an upper motoneuron disease identified in Africa. We have investigated the effect of a low protein diet on the capacity for cyanide detoxification. Rats were fed normal chow containing 18% protein or a low protein diet with 5% protein. To expose rats to cyanide the drinking water was supplemented with 40 or 80 mM acetonitrile (CH3CN) for up to 4 weeks. Weight gain was monitored and 24-hr urines were collected for analyses of total sulfur, inorganic sulfate, thiocyanate, and 2-aminothiolazine-4-carboxylic acid (ATC). Blood was collected for analyses of cyanide and cyanate. Rats on a normal diet grew throughout the experiment, while those on a low protein diet initially lost weight and then stabilized at a constant weight. Rats exposed to acetonitrile all progressively lost weight, those on a low protein diet at the highest rate. Signs of neurological damage were not observed. Rats not exposed to acetonitrile excreted < 0.2% of sulfur as thiocyanate and those on a low protein diet reduced their total sulfur excretion to one-third that of rats of the normal diet. Rats on the normal diet did not change total sulfur excretion during exposure to acetonitrile, although thiocyanate now contributed more than two-thirds of excreted sulfur. Rats on a low protein diet exposed to acetonitrile increased both total sulfur and thiocyanate excretion to the levels of rats on a normal diet. Rats exposed to acetonitrile had manyfold increases of circulating concentrations of cyanide and cyanate and of urinary excretion of ATC. There was a positive correlation between blood cyanide concentrations and the plasma concentration of cyanate. It is concluded that the rat has a high capacity for detoxification of cyanide. During adaptation to a low protein intake, sulfur is conserved but cyanide detoxification is still possible at the cost of extensive protein catabolism. It is thus possible that subclinical cyanide exposure could interfere with normal growth and development. The observation of a relationship between circulating cyanide on the one hand and circulating cyanate and urinary excretion of ATC on the other highlights the possibility that cyanide metabolites may mediate neurotoxic effects of cyanide.

Age-dependent differences in insulin secretion and intracellular handling of insulin in isolated pancreatic islets of the rat.

Insulin secretory response to glucose changes with age. To elucidate age-dependent differences in pancreatic islet responsiveness to glucose, isolated islets from rats one week, three months and 14 months old were investigated in vitro. At three months of age, islet insulin secretion was increased five-fold by an acute glucose challenge. There was a significantly lower insulin response in both younger and older age groups. Islet insulin biosynthesis, as determined by the rate of incorporation of radioactive leucine into immunoprecipitable insulin, was lower at three months of age than at one week or 14 months. Insulin content was lowest in islets from the youngest rats and increased with age. The capacity for islet intracellular degradation of insulin was estimated according to the disappearance of tritiated leucine-labelled insulin during a 24-hour chase incubation. At a high glucose concentration, virtually no insulin was degraded intracellularly in islets from three-month-old rats, whereas islets from both younger and older animals showed a significant degradative capacity. High activities of a number of lysosomal enzymes in islets from one-week-old rats could account for the high degradative capacity and relatively low insulin content of these islets. Thus, low insulin response to glucose during early development may depend primarily on low insulin stores. However, during ageing, when islets are characterised by high insulin content, low response may depend primarily on impairment of beta-cell stimulus-secretion coupling, with high intracellular degradation of insulin resulting secondarily from an accumulation of insulin in the islets.

Inter-relationship between serum concentrations of glucose, glucagon and insulin during the first two days of life in healthy newborns.

The relationship between serum concentrations of glucose, insulin and glucagon during the first two days of life was studied in healthy newborns. The first capillary blood sample was obtained at 3-15 h of age (median 6 h; day 0) and a second sample approximately 24 h later (day 1). Serum glucose concentrations in the first sample averaged 2.1 +/- 0.07 mmol/l (mean +/- SEM; n = 60) and were positively correlated with postnatal age (p < 0.01). Serum glucagon concentrations in the first sample averaged 570 +/- 32 pg/ml and were inversely correlated with glucose concentrations (p < 0.0001). At the second sampling, serum glucose concentrations had increased to 2.9 +/- 0.07 mmol/l (p < 0.001; n = 57) and serum glucagon concentrations had decreased to 403 +/- 22 pg/ml (p < 0.001). Serum insulin concentrations were 11.7 +/- 0.3 microU/ml and 10.2 +/- 0.3 microU/ml at the two samplings and did not correlate with serum glucose concentrations. The relationship of serum glucose and hormone concentrations to maternal and infant characteristics was studied by stepwise regression analysis. Serum glucose concentration on day 0 was positively correlated with postnatal age (p < 0.01) and birth weight (p < 0.05) but inversely correlated with duration of labour (p < 0.05). Serum glucose concentration on day 1 was positively correlated with birth weight (p < 0.0001) and inversely correlated with maternal prep-pregnancy weight (p < 0.05). Similar analyses of serum hormone concentrations did not demonstrate any relationships with maternal or infant characteristics.(ABSTRACT TRUNCATED AT 250 WORDS)

Diabetes in pregnancy: fetal macrosomia, hyperinsulinism, and islet hyperplasia in the offspring of rats subjected to temporary protein-energy malnutrition early in life.

The effect of temporary, severe protein-energy malnutrition in young female rats on the fetal and maternal outcome of subsequent pregnancy has been investigated. Female rats were fed a low-protein diet between 3 and 6 wk of age. During this period, growth was stunted, glucose tolerance impaired, and insulin secretory response to glucose almost absent. Upon refeeding normal diet, growth was resumed and glucose tolerance normalized, but a subnormal insulin secretory response to glucose and a reduced pancreatic B-cell mass persisted into adult life. During pregnancy, such rats had a glucose tolerance with wide excursions of the serum glucose concentrations and an exaggerated insulin secretory response compared with normal rats. The previously malnourished rats also expanded their B-cell mass during pregnancy to the level of normal rats. The incidence of fetal loss and malformations was not increased in the offspring of previously malnourished rats. At term, the viable offspring were heavier than those of normal rats and had an increased pancreatic insulin content and an increased pancreatic B-cell mass. It is concluded that temporary protein-energy malnutrition in young rats reduces the ability to increase insulin production to meet the needs of pregnancy. The capacity for glucose disposal is decreased and glucose and other nutrients are transferred to the fetus in increased amounts. This stimulates pancreatic B-cell growth and the development of macrosomia in the offspring, well recognized features of diabetic pregnancy. The long-term consequences of malnutrition are thus not restricted to the malnourished but may also be inflicted on subsequent generations.

Persistent reduction of pancreatic beta-cell mass after a limited period of protein-energy malnutrition in the young rat.

Kwashiorkor, the human disease of protein-energy malnutrition, has been implicated in the aetiology of malnutrition-related diabetes mellitus, a form of diabetes not uncommon in developing countries. We have previously demonstrated that temporary protein-energy malnutrition in young rats causes a persisting impairment of insulin secretion. The present study investigates whether this secretory deficiency is accompanied by structural alterations of the endocrine pancreas. Three-week-old rats were weaned onto semi-synthetic diets containing either 15% or 5% protein and these diets were maintained for 3 weeks. From 6 weeks of age all rats were fed a commercial chow containing 18% protein. The endocrine pancreas was investigated by light and electron microscopic morphometry at 3, 6 and 12 weeks of age. In rats not subjected to protein-energy malnutrition there was a progressive increase, with age, of total pancreatic Beta-cell weight and individual Beta-cell size. In 6-week-old rats fed the low protein diet total pancreatic Beta-cell weight and individual Beta-cell size were diminished. In 12-week-old rats previously fed the low protein diet total Beta-cell weight remained lower compared to control rats. It is concluded that protein-energy malnutrition early in life may result in a diminished reserve for insulin production. This may predispose to glucose intolerance or even diabetes in situations with an increased insulin demand.

Functional maturation and proliferation of fetal pancreatic beta-cells.

We review some key aspects of the maturation of stimulus-secretion coupling and the regulation of DNA replication in the fetal beta-cell. During fetal life, the beta-cell shows a poor insulin response to glucose, although it responds to several other nonnutrient stimuli. However, chronic exposure to glucose in excess of basal levels can induce maturation of the stimulus-secretion coupling. Studies of glucose metabolism and the transmembrane flow of K+ and Ca2+ indicate that the attenuated glucose-stimulated insulin release is due to an immature glucose metabolism resulting in impaired regulation of ATP-sensitive K+ channels in the plasma membrane of the fetal beta-cell. In late fetal life, glucose is also a strong stimulus to beta-cell replication, and metabolism of glucose is a prerequisite for this process. Glucose stimulates proliferation by recruiting beta-cells from a resting state into a proliferative compartment composed of cells in an active cell cycle. The proliferative compartment comprises less than 10% of the total islet cell population even at maximal stimulation. The proliferation of fetal beta-cells is also regulated by several peptide growth factors such as growth hormone, insulinlike growth factor I, and platelet-derived growth factor. The observation that glucose can both induce precocious maturation of the stimulus-secretion coupling and stimulate proliferation of the fetal beta-cell explains the intrauterine hyperinsulinemia and beta-cell hyperplasia of the offspring of diabetic mothers with relatively mild hyperglycemia. However, severe hyperglycemia, at least when induced in rats, seems to retard rather than stimulate beta-cell growth.

Tissue and serum insulin-like growth factor I (IGF I) concentrations in rats subjected to temporary protein-energy malnutrition early in life.

Rats subjected to temporary protein-energy malnutrition and subsequent nutritional rehabilitation remain smaller than adequately fed animals, have a subnormal insulin secretion and persisting cellular hypoplasia in several tissues. This investigation studies whether impaired production of insulin-like growth factor I (IGF I) is another persisting consequence of malnutrition. Rats were subjected to severe protein-energy malnutrition between 3 and 6 weeks of age and subsequently fed adequate diet up to 12 weeks of age. Serum and tissue samples for analysis of IGF I were obtained at 12 weeks of age. IGF I concentrations were similar in serum, heart, liver and lung of previously malnourished and control rats. In the kidneys of previously-malnourished rats the IGF I concentration was twice that of control rats. Results suggest that during protein-energy malnutrition and subsequent nutritional rehabilitation IGF I tissue concentrations are primarily regulated by the prevailing plane of nutrition. It is speculated that the temporary protein-energy malnutrition blunts the cellular capacity for IGF I production and, except in the kidney, prevents increased IGF I tissue concentrations and associated compensatory growth.

Long-term effects on glucose tolerance and insulin secretory response to glucose following a limited period of severe protein or energy malnutrition in young rats.

The long-term effects on growth, glucose tolerance and insulin secretory response to glucose of temporary malnutrition early in life have been investigated. Rats were weaned onto either normal diet (18% protein), a protein-restricted diet (5% protein) or a diet adequate in protein but restricted in amount to equal the energy intake of protein-restricted rats ("energy restriction"). From 6 weeks of age and onwards all rats were fed normal diet. Body weight gain was inhibited by both protein and energy restriction but growth was resumed when rats were transferred to normal diet. Protein restriction impaired glucose tolerance and blunted insulin secretory response to glucose. Following refeeding glucose tolerance was normalized but insulin secretory response remained impaired at 12 weeks of age. Energy restriction did not initially affect glucose tolerance and insulin secretion. However, after refeeding male energy restricted rats developed a delayed and exaggerated insulin secretory response to glucose without concomitant deterioration of glucose tolerance. It is suggested that temporary protein restriction at a young age impairs pancreatic B-cell function and decreases peripheral sensitivity to insulin. By contrast, temporary energy restriction does not directly affect B-cell function but confers insulin resistance and compensatory increases of the insulin secretory response to glucose later in life. These models of malnutrition offer possibilities to further study long-term effects of early nutritional insults.