Pavel Fort

Fear of obesity: A cause of short stature and delayed puberty

We evaluated 201 children for short stature or delayed puberty or both. Fourteen of them (nine boys and five girls, aged 9 to 17 years) fit a pattern of growth failure due to malnutrition, which was the result of a self-imposed restriction of caloric intake arising from a fear of becoming obese. All 14 patients underwent a complete history, physical examination, diagnostic laboratory evaluation, and psychiatric assessment. They were all below the fifth percentile for weight, and 11 of them were also below the fifth percentile for height. The deficit of weight for height ranged from 5 to 23 per cent. Seven of the older patients also had delayed puberty. All 14 patients had deteriorating linear growth, which was preceded by at least one to two years of inadequate weight gain. They ingested only 32 to 91 per cent of the recommended caloric intake for their age and frequently skipped meals. No gross psychiatric disease or anorexia nervosa was found; on the whole, they were good students with rather compulsive, shy personalities observed in an open-ended interview. The Diagnostic Interview for Children and Adolescents, which was conducted with seven patients, also revealed no psychiatric disease. After nutritional and psychiatric counseling, the patients resumed an adequate caloric intake for their age, and recovery occurred, as demonstrated by increased linear growth and sexual development.

Impaired Somatomedin Generation Test in Children with Insulin-dependent Diabetes Mellitus

Recent studies have suggested a partial block in somatomedin (SM) production or growth hormone (GH) action in IDDM. Twelve well-nourished diabetic children (9 males and 3 females with a mean age of 11.2 ± 3.3 yr), six with an HbA1c of 7.9–11.2% (group A) and six with an HbA1c of 12.5–15.6% (group B), were studied as follows: the GH response after 100 μg of oral clonidine and the SM generation capacity after i.m. administration of 0.2 U/kg/dose of human growth hormone (hGH) for 4 days. Group B diabetic subjects had a significantly higher mean ± SD GH increase after clonidine than did group A patients (Δ of 17.4 ± 4.9 versus 5.7 ± 6.0 ng/ml, P < 0.01); the basal GH of both groups were similar (1.6 ± 0.7 versus 2.3 ± 1.4 ng/ml). In contrast, the SM response to hGH was significantly decreased in group B children as compared with those in group A (Δ of 0.3 ± 0.3 versus 1.2 ± 0.4 U/ml, P ± 0.01). The basal SM levels of both groups were normal for age. GH and SM correlated with HbA1c levels (r = +O.80, P < 0.01; r = −0.79, P < 0.01, respectively); there was no correlation with plasma and urine glucose or serum cholesterol, cortisol, and transferrin. Our data indicate a blunted SM response to hGH in group B diabetic subjects; this defect in SM generation is apparently not present in group A subjects. A biologically inactive GH molecule and poor nutrition seem unlikely, but circulating inhibitory factors not picked up by our radioimmunoassay, the degree of diabetes control, or a still unclear metabolic derangement may be contributing to this defect.

Magnesium Metabolism in Experimental Diabetes Mellitus

One of the consequences of uncontrolled diabetes mellitus (DM) is an increased urinary loss of magnesium. Some of the factors determining this electrolyte imbalance were studied in growing rats with experimentally induced DM. Two sets of experiments were performed in which dietary magnesium intake was the variable. One group each of diabetic (DB) and control (C) animals was fed a complete diet including magnesium ad libitum; the other set was given the same diet as the first two groups but the magnesium was provided daily through a stomach tube in amounts equal to those eaten by non-DB rats of the same size. All DB animals had significantly higher than normal urinary excretion of glucose, magnesium, calcium, and phosphorus. There was a positive correlation between urinary glucose and magnesium excretion. DB animals fed magnesium ad libitum became hypomagnesemic, but their bone magnesium content was increased over that of C rats. Plasma magnesium correlated negatively with both glucose and cholesterol. However, alterations in plasma cholesterol induced in C rats by dietary means had no effect on magnesium levels. When magnesium intake was restricted to the physiologic requirements of C animals, hypomagnesemia was more pronounced and occurred with concomitant depletion of magnesium in bone, suggesting possible risk of electrolyte imbalance in uncontrolled DM. The results indicate that hypomagnesemia without osseous tissue magnesium depletion may occur in experimental DM. Hyperphagia in DB animals fed ad libitum can prevent bone magnesium depletion, but a “normal” intake that does not compensate for losses may be conducive to a marked deficit in the intracellular pool of magnesium.

HYPOTHALAMIC HYPOSOMATOTROPINISM IN CHILDREN WITH SHORT STATURE

Hyposomatotropinism can be due to a pituitary defect of growth hormone (GH) secretion, a hypothalamic defect of GH-releasing hormone (GHRH) secretion or to a Neurosecretory Defect (NSGHD). Twenty-nine patients, aged 5-17 years, with height below the 5th percentile, delayed bone age and no organic disease were studied. GH status was evaluated by [1] a combined hormonal stimulation test (CHST) of sequentially administered insulin, GnRH, TRH, and L-Dopa, [2] 8-hour overnight GH test (CGH) and [3] IV (1-44)hp GHRH-NH2 infusion. Fourteen patients with growth rate (GR) of 3.7±1.1 cm/yr had a normal GH response on CHST (12.9 ± 4.5 ng/ml.) and GHRH (24.6±12.6) but subnormal OGH (2.6±0.7) compatible with NSGHD. Five patients with GR of 2.8±0.6 cm/yr had a subnormal GH response on CHST (2.6±1.4) and OGH(1.6±0.3) but a normal GH response to GHRH (11.3±5.7) compatible with GHRH deficiency. Two patients with a GR of 3.0 cm/yr had subnormal responses on CHST (2.3), OGH (1.3) and GHRH (1.6) compatible with pituitary GH deficiency. Eight patients with GR of 4.9±1.3 cm/yr had normal responses on all 3 tests, deficiency is mainly hypothalamic in origin. The most cammon form is the NSGHD with a clinical spectrum from a virtually normal growth rate down to severe growth stunting. GHRH deficiency is a less common, but is more common than a pituitary GH deficiency. This study demonstrates that combined physiologic and pharmacologic tests of GH reserve can determine the origin of the disorder.

483 PITUITARY FUNCTION IN PRIMARY MALNUTRITION DUE TO ANOREXIA NERVOSA AND OTHER EATING DISORDERS

The effects of nutritional state on pituitary function were studied using a combined pituitary stimulation test in 3 groups of 10 patients each. 1-Classical Anorexia Nervosa (AN) with a weight deficit for height (WDH) of 26.0±8.1%, aged 12–19 years. 2-Non-AN eating disorders with a WDH of 12.5±8.6%, aged 12–17 years. 3-Familial or constitutional short stature without pathological WDH and no nutritionally based growth disorder, aged 10–16 years. Each patient received seqentially IV TRH, LRH, regular insulin and PO L-Dopa. Serum TSH, glucose, cortisol, GH, LH and FSH, and prolactin were sampled periodically over 2 hrs. Basal T3, T4, Somatomedin-C and other chemistries were obtained. The peak TSH response was blunted and delayed in direct relation to the WDH. The ratio of the integrated TSH response from 0 min. to 45 min. divided by the 45 min. to 120 min. response was correlated to the WDH (p<0.001) and to recovery from hypoglycemia (p<0.001). Also, the recovery from hypoglycemia (p<0.001), the basal cortisol (p<0.05), and transferrin (p<0.05), all correlated to the WDH. The transferrin correlated to the T3 (p<0.01). The most underweight patients had the lowest T3, basal glucose and Somatomedin-C levels. This study shows the newly described integrated TSH ratio to correlate well to the WDH and the ability to recover from stress. Also demonstrated was that the responses noted were independant of the etiology of the eating disorder which resulted in the malnourished state.

IMPAIRED SOMATOMEDIN (SM) GENERATION TEST IN POORLY CONTROLLED INSULIN DEPEDENT DIABETES MELLIITUS (IDDM)

Recent studies have suggested a partial block in SM production or growth hormone (GH) action in IDDM. Twelve diabetic children (9 males and 3 females with a mean age of 11.2 ± 3.3 years), six in good control (HbA1C 7.9 - 11.2%) and six in poor control (HbA1C 12 - 15.6%) were studied as follows: the GH response following 100 ug of oral clonidine and the SM generation capacity after i.m. administration of 0.2 u/kg/dose of hGH for 4 days. Poorly controlled diabetics had a significantly higher mean ± SD GH increase after clonidine than did well controlled patients (17.4 ± 4.9 vs 5.7 ± 6.0 ng/ml; p<0.004); the basal GH of both groups were similar (1.6 ± 0.7 vs 2.3 ± 1.4 ng/ml). In contrast the SM response to hGH was significantly decreased in poorly controlled children as compared to well controlled (Δ of 0.3 ± 0.3 vs 1.2 ± 0.4 U/ml, p<0.002). The basal SM levels of both groups were normal for age. Δ GH and Δ SM correlated with HbA1C levels (r = + 0.80, p<0.01;r = -0.79, p < 0.01, respectively);there was no correlation with plasma and urine glucose or serum cholesterol, cortisol and transferin. Our data indicate a blunted SM response to hGH in poorly controlled diabetes; this defect in SM generation is apparently not present in well controlled IDDM. A biologically inactive GH molecule seems unlikely, but circulating inhibitory factors or a still unclear metabolic desangement may be contributing to this defect.

LOW DOSE ORAL CLONIDINE: A SIMPLE AND RELIABLE GROWTH HORMONE (GH) SCREENING TEST

Clonidine has been shown to be an effective GH releasing agent in prepubertal children at a dose of 150 micrograms/M2 body surface area, but at this dose it can produce significant side effects. We have evaluated the efficacy and side effects of low dose oral clonidine on GH, cortisol and blood pressure in 24 healthy short children; ten received 100 micrograms (Group A), and fourteen 50 micrograms (Group B). A 100 micrograms of oral clonidine induced higher, more sustained GH levels and fewer failures to testing than did 50 micrograms. The mean ± SD GH peak at 60 minutes in group A was 14.5 ± 6.3 vs 11.6 ± 6.1 ng/ml in group B (NS). Failure rate (GH < 10 ng/ml) was 10% in group A and 36% in group B. Children who responded to clonidine in group A (GH≥10 ng/ml) had significantly higher GH levels than did group B responders (p<0.01). A similar, but significant drop in cortisol was seen in both groups (from 14.5 ± 6.3 to 7.8 ± 2.6 ug/dl in group A and from 16.9 ± 6.7 to 12.8 ± 8.4 ug/dl in group B). Blood pressure did not change significantly with either dose used and only mild somnolence was noted in both groups.A single oral dose of 100 micrograms of clonidine, followed by 1 blood sample 60 minutes later can be used to effectively and safely screen short children for GH deficiency.

HEMOGLOBIN LEVELS IN HYPOTHYROID INFANTS

Children and adults with long-standing hypothyroidism frequently develop anemia, often of a macrocytic nature. The incidence of anemia has been reported to be between 21 and 61% in adults. We investigated newborns with hypothyroidism for the presence of anemia. All infants were initially determined to have abnormal thyroid function by the N.Y. State Screening Program with heel stick blood specimens obtained by day 3 of life. Newborns with elevated TSH levels (>20 μU/ml) or low T4 levels (<8 ug/dl) were included in the study. Repeat thyroid studies were performed at 10 to 55 days of life, and blood counts with RBC indices were measured by Coulter Counter. All children with other causes of anemia or polycythemia were excluded from analysis. Of 23 infants who fit these criteria, none were found to be anemic, nor did any have macrocytic indices. Surprisingly, 6 children (26%) were discovered to have polycythemia, 4 with significantly elevated hemoglobins as high as 23 g/dl. All children with polycythemia had normal red cell indices. The hemoglobin showed no correlation with T4 or TSH levels. Anemia in patients with hypothyroidism is likely to be a result of chronically abnormal thyroid function, and as such, would not appear to be helpful in screening infants for thyroid dysfunction.

Insulin reserves in children with impaired glucose tolerance(IGT) and subsequent development of diabetes mellitus (DM)

The validity of the maximum stimulation test (MIST) and oral glucose tolerance test(OGTT) as screening procedures for the development of DM was studied in 33 children with IGT. They were administered MIST(p.o.glucose;i.v.tolbutamide and glucagon) and OGTT on 2 consecutive days.Insulin response was compared with the development of DM for up to 5½yrs.MIST was a better indicator of the chance to develop DM:3 non-obese subjects with poor MIST insulin response (glucose/insulin ratio> 4)developed DM. One child who was on Orinase at the time of studies,progressed to DM despite a good MIST insulin release.All others had a good MIST insulin response and did not develop DM.The predictive value for the development of DMwith a poor MIST insulin response was 75%.The predictive value for non-development of DM with a good MIST insulin response was 96.2%.In contrast,OGTT elicited poor insulin responses in 9 of 30 children;only 3 of them progressed to DM.The predictive value for development of DM with a poor OGTT insulin response at 1 hr was 33.3% and the predictive value for non-development of DM with a good OGTT insulin release was 100%.Thus, the ability to release insulin during MIST may be a better prognostic indicator of the chance to develop DM in children with IGT,whereas OGTT may help in assessing the decreased risk of DM.

405 SCREENING FOR DIABETES MELLITUS IN CHILDREN WITH IMPAIRED GLUCOSE TOLERANCE

The estimation of the validity of the maximum insulin stimulation test (MIST) and oral glucose tolerance test(OGTT) as predictive screening procedures for the development of overt diabetes mellitus(DM) was studied.Thirty children with various degrees of asymptomatic impaired glucose tolerance(IGT)were administered MIST(p.o.glucose;i.v.glucagon and tolbutamide)and OGTT on 2 consecutive days following an overnight fast.Insulin response during both tests was validated by the development of DM during a period of up to 5½ yrs.MIST appeared to be a better prognostic indicator of the chance to develop DM:3 non-obese subjects with poor MIST insulin response(glucose/insulin ratio>4)developed DM. However,one child who was on Orinase at the time of studies,had a good MIST insulin release and developed DM.All other patients had a good MIST insulin response and did not progress to DM.The predictive value for development of DM with a poor MIST insulin response was 75%.The predictive value for non-development of DM with a good MIST insulin response was 96.2%.In contrast,OGTT elicited poor insulin responses in 9 of 30 children;only 3 of them progressed to DM.The predictive value for development of DM with a poor OGTT insulin response at 1 hr was 33.3% and the predictive value for non-development of DM with a good OGTT insulin release was 100%.Thus, in a high risk population of children with IGT the ability to release insulin during MIST may be a better prognostic indicator of the chance to develop DM,whereas OGTT may be more useful in assessing the decreased risk of DM.