Objective: Insulin-like growth factor binding protein (IGFBP)-1 levels increase overnight, being inversely related to changes in insulin. With prolonged fasting IGFBP-1 levels increase further. In animal studies high IGFBP-1 levels increase plasma glucose levels possibly by regulating the insulin-like actions of 'bio-available' plasma IGF. Following prolonged fasting, there is an increase in insulin requirement. A proportion of this reversible insulin resistance may be due to inhibitory effects of high IGFBP-1 levels on IGF action. This study examined the regulation of IGFBP-1 in the presence of reversible insulin resistance.
Subjects: Nine normal adult volunteers, seven female and two male (mean age 27.6 +/- SD 2.6 years, range 21.7-46.0 years) of normal body mass index were studied.
Methods: Subjects fasted from 2200 h day 0 to 0900 h day 3 (59 hours), the fast being completed with a 75-g glucose meal. At least one week later, an 11-hour overnight fast was performed, followed by a repeat glucose meal. Blood samples were taken at regular intervals from 0900 h day 1 and for 5 hours during both glucose meal studies via an indwelling cannula.
Measurements: Serum levels of IGFBP-1, insulin, GH, glucose, IGF-I and cortisol were measured at varying intervals during the fast and both glucose meal studies.
Results: Following the initial 11-hour overnight fast IGFBP-1 levels rose from (mean +/- SEM) 32 +/- 5 micrograms/l to reach a maximum of 144 +/- 24 micrograms/l after 32 hours of fasting. IGFBP-1 levels then fluctuated, falling in the morning (93 +/- 8 micrograms/l) and then rising overnight (126 +/- 9 micrograms/l), but not regaining the initial peak levels. The increase of IGFBP-1 from overnight fasting levels was associated with a fall in plasma insulin from 5.7 +/- 0.7 to 2.2 +/- 0.2 mU/l. In comparison, 30 minutes after termination of the fast with the glucose meal, IGFBP-1 levels fell from 120 +/- 11 to 24 +/- 2 micrograms/l within 4 hours. After an overnight fast IGFBP-1 levels fell from 35 +/- 5 to 13 +/- 2 micrograms/l within 3 hours. There was glucose intolerance and increased insulin levels following the glucose meal preceded by the 59-hour fast when compared with the overnight fast. The fall of IGFBP-1 levels after the glucose meal was best expressed, taking into account subject variation, by the following regression equations: Glucose meal preceded by 11-hour fast: log [IGFBP-1] = 1.64-0.255 log [1 h previous insulin] (R2 0.51); Glucose meal preceded by 59-hour fast: log [IGFBP-1] = 1.41-0.265 log [1 h previous insulin] + 0.557 log [current glucose] (R2 0.82).
Conclusion: In man, insulin appears to regulate circulating IGFBP-1 levels in all circumstances, this regulation being unaffected by the resistance to insulin action induced by prolonged fasting. The high IGFBP-1 levels were statistically related to the higher glucose levels and may have directly contributed to the increased insulin requirement observed after prolonged fasting.