In a Mongolian pilot study, levels of prepubertal GH, IGF-I, and IGF-I/IGFBP-3 were markedly higher after one month of drinking milk. In the smaller, shorter Boston pilot study, which contrasted cow's milk to a macronutrient control for cow's milk, levels of GH, IGF-I and IGF-I/IGFBP3 were not elevated, although we observed a small non-significant shift in somatotropin levels in the direction of the Mongolian findings. The rise in somatotropic hormone and growth factor levels was larger in the Mongolian than in the Boston study, perhaps due to the longer intervention, the fact that the Mongolian children consumed less dairy at baseline, the use of whole versus lowfat milk, the older age of the Mongolian participants, or the baseline nutritional status of the Mongolian children.
We conducted the Mongolian feasibility pilot with limited resources, and therefore did not attempt a macronutrient control. During the intervention month, the Mongolian children experienced rapid linear growth (the equivalent of 12 cm/year) compared to an average height velocity of 5–6 cm/year in U.S. children age 10–11 years . Although such rapid growth suggests that the children may have been undernourished at the start of the trial, the increase in GH is unlikely to be a refeeding effect of milk nutrients. The classic hormonal profile in undernutrition is low IGF-I, but stable or even elevated GH levels [30–33]. High GH during fasting may promote the tissue-sparing switch from gluconeogenesis to lipolysis during periods of undernutrition [30, 31]. On refeeding, IGF-I levels increase, but GH levels typically fall [31, 32]. In the Mongolian children, the increase in GH levels after milk consumption was proportionally higher than the increase in either IGF-I or IGF-I/IGFBP-3 levels. This finding is inconsistent with a malnutrition correction effect. As GH is released in a pulsatile fashion, and we had only one GH measurement in these pilot studies, we undoubtedly overestimated the variability of GH, reducing our statistical power to detect a GH response to milk. Future studies should measure GH more frequently to capture GH secretory dynamics and better estimate GH levels.
Our studies suggest an immediate stimulatory impact of milk on the secretion of GH and bioactive IGF-I. Other studies suggest that the impact of milk on IGF-I levels is maintained, and may grow over time. In China, 10 year old girls randomized to consume whole milk had 9% higher IGF-I levels after one year, and 20% higher levels after two years (p < 0.05) . In Britain, 12 year-old girls randomized to drink approximately 300 ml/day of milk had 10% higher IGF-I levels after 6 months of intervention, and 17% higher levels at months 12 and 18 of intervention (p = 0.02) . To our knowledge, ours is the first study to report GH response to dairy intake, which was large and statistically significant among the Mongolian children.
The findings here should be replicated in larger studies with macronutrient and micronutrient control. If consistent findings emerge, it would be important to identify the specific bioactive factor in bovine milk that raises GH and IGF-I levels. Although human and bovine GH are not identical, bovine GH absorbed from milk might cross-react with the human GH assay, and it is conceivable (but as yet untested) that bovine GH might stimulate human IGF-I production. Animal feeding experiments, using radioactively labeled bovine vs. human GH, would be an approach to answering the question of the origin of elevated GH we observed in the children's serum. Alternatively, another factor in bovine milk may stimulate endogenous production of human GH, which would, in turn, provoke IGF-I production. Although bovine and human IGF-I are identical, IGF-I absorbed from milk is unlikely to stimulate human GH secretion, as GH levels should drop in response to negative feedback from IGF-I infusion. Future analysis should focus on identifying factors in milk that could stimulate human GH secretion, such as growth hormone releasing hormone, ghrelin, or a substance that stimulates ghrelin release.