Epidemiological studies have shown a link between poor fetal growth and increased risk of developing type 2 diabetes. These observations are highly reproducible in many populations worldwide although the mechanisms behind them remain elusive.The 'Thrifty Phenotype Hypothesis' was proposed to explain the underlying causes of these relationships. Animal models of poor intrauterine nutrition have been utilised to help to define the causal factors and identify the molecular mechanisms. Programmed changes in beta cell function and insulin action have been a common feature of animal models of poor intrauterine nutrition. Fundamental underlying mechanisms are starting to emerge, including changes in the epigenotype and mitochondrial function.

The metabolic and cardiovascular effects of recombinant human IGF-I were compared to insulin in six normal subjects. Subjects were studied twice and intravenously received an infusion of [6,6-2H2]glucose (0-480 min) and in random order either IGF-I 20 micrograms kg-1 h-1 (43.7 pmol kg-1 min-1 or insulin 0.5 mU kg-1 min-1 (3.4 pmol kg-1 min-1) with an euglycaemic clamp. One subject was withdrawn following a serious adverse event. During the IGF-I infusion glucose appearance rate (Ra) decreased from 1.79 +/- 0.13 at baseline (150-180 min) to 0.35 +/- 0.26 mg kg-1 min-1 (P < 0.01) at 360 min, and glucose utilization rate (Rd) increased from 1.79 +/- 0.28 to 4.17 +/- 0.84 mg kg-1 min-1 (P < 0.01). There was no change in free fatty acids (FFA) and an increase (percentage change from pre-infusion mean) in cardiac output +l37.3% +/- 9% (P < 0.01), heart rate +13% +/- 2% (P < 0.01) and stroke volume +21% +/- 7% (P < 0.05). During the insulin infusion glucose Ra decreased from 1.89 +/- 0.13 to 0.34 +/- 0.33 mg kg-1 min-1 (P < 0.01) and FFA from 0.546 mmol l-1 to 0.198 mmol l-1 (P < 0.01), glucose Rd increased from 1.89 +/- 0.18 to 5.41 +/- 1.47 mg kg-1 min-1 (P < 0.01) and there were no significant changes in the cardiovascular variables.

Abbreviations: AUC, area under the curve; AUCR a , area under the curve calculated for glucose R a ; AUCR d , area under the curve calculated for glucose R d ; B1-T4-Ins, N ␣B L-thyroxyl-insulin; CV, coefficient of variation; FPLC, fast protein liquid chromatography; GINF, exogenous glucose infusion; IRI, immunoreactive insulin; MCR, metabolic clearance rate; NEFA, nonesterified free fatty acid; R a , glucose production; R d , glucose disposal; RIA, radioimmunoassay; TBG, thyroxine binding protein; THBP, thyroid hormone binding protein; TSH, thyroid-stimulating hormone.A table elsewhere in this issue shows conventional and Système International (SI) units and conversion factors for many substances. Novel Hepatoselective Insulin AnalogStudies with a covalently linked thyroxyl-insulin complex in humansOBJECTIVE -To test whether a thyroxyl-insulin analog with restricted access to receptor sites in peripheral tissues displays relative hepatoselectivity in humans.RESEARCH DESIGN AND METHODS -Five normal human subjects received a subcutaneous bolus injection of either N ␣B1 L-thyroxyl-insulin (B1-T4-Ins) or NPH insulin in random order. Insulin kinetics, relative effects on hepatic glucose production, and peripheral glucose uptake were studied using euglycemic clamp and stable isotope [D-6,6-2 H 2 ]glucose) dilution techniques. Blood samples were taken for the determination of total immunoreactive insulin/analog concentrations and for liquid chromatography to assess the protein binding of the analog in the circulation.RESULTS -After subcutaneous administration, B1-T4-Ins was well tolerated and rapidly absorbed. The analog had a long serum half-life and was highly protein bound (ϳ86%). Its duration of action, as judged by the duration of infusion of exogenous glucose to maintain euglycemia, was similar to that of NPH insulin. The effect of the analogs on hepatic glucose production was similar to that of NPH insulin, indicating equivalent hepatic potency. The analog demonstrated less effect on peripheral glucose uptake than NPH insulin (P = 0.025), had no effect on metabolic clearance rate of glucose, and exhibited a reduced capacity to inhibit lipolysis (P Ͻ 0.05).CONCLUSIONS -When injected subcutaneously into normal human subjects, B1-T4-Ins is well tolerated, quickly absorbed, and highly protein bound, resulting in a long plasma halflife. This analog appears to have a hepatoselective action, and, therefore, has the potential to provide more physiological insulin action than the insulin preparations currently used. E m e r g i n g T r e a t m e n t s a n d T e c h n o l o g i e s