Summary of the evidence for associations between dietary factors and serum IGF-I
| Nutrient/food group . | Studies using animal models* . | Cross-sectional studies in humans† . | Experimental studies in humans* . | Conclusion . |
|---|---|---|---|---|
| Energy | Severe energy restriction results in 25% decreased serum IGF-I (review; ref. 75) | No association (8×; refs. 68, 63, 69, 71, 66, 72, 73, 74) | Serum IGF-I is markedly lowered by energy deprivation (review; ref. 60) | Probable; positive association with extremes of energy intake |
| Positive association (2×; refs. 67, 70) | No association within normal range of energy intake | |||
| Protein (total or animal) | Severe protein restriction results in 50% decreased serum IGF-I (review; ref. 61) | No association with total and/or animal protein (6×; refs. 64, 66, 71, 73, 74) | Serum IGF-I is markedly lowered by protein deprivation (review; ref. 60) | Possible; positive association with intake of animal protein |
| Positive association with total or animal protein (5×; refs. 63, 67, 68, 69, 70) | ||||
| Alcohol | Chronic alcohol feeding may increase or decrease serum IGF-I | No association with IGF-I (5×; refs. 64, 67, 68, 72, 74) | Decline in serum IGF-I after drinking alcohol (1 d only) | No association within normal range of alcohol consumption |
| Minor inverse association (66) | ||||
| Positive association (in men only; ref. 62) | ||||
| Minerals (total or zinc) | Mineral (e.g., zinc, copper, and magnesium) deficiency is associated with low serum IGF-I levels | No association with zinc (68) | Zinc supplementation: increased serum IGF-I levels in specific populations (anemic, non-insulin-dependent diabetes mellitus) | Probable; positive association with zinc or a combination of minerals (including supplements) |
| Positive association with zinc (63, 67) | ||||
| Positive association with combined intake of five minerals (including zinc; ref. 69) | ||||
| Dairy products | Only studies relating to IGF-I content in milk (review; ref. 79) | No association with total dairy (64, 66, 68, 71; one positive association with milk only; ref. 71) | Serum IGF-I increased 10% with three servings of dairy daily (77) | Possible; positive association with consumption of dairy/milk |
| Positive association with total dairy (67) | ||||
| Positive association with milk (65, 69) | ||||
| Soy or isoflavones | Decrease in serum IGF-I in animals on soy protein/phytoestrogen diet (93, 94) | Asian populations: no association with soy or isoflavones (72); positive association (in men only; ref. 73) | Soy protein vs. milk protein: stronger increase in IGF-I in soy protein group (99, 100) | Insufficient; association could be positive or inverse, depending on dose and population |
| European population (including vegans/vegetarians): no association with soy protein (borderline positive with soy milk; ref. 68) | Soy protein with vs. without isoflavones: trend towards lower IGF-I with isoflavones (postmenopausal women only; ref. 98) or no difference between groups (101) | Effects of soy protein and isoflavones/phytoestrogens to need be disentangled (possibly opposite effects) | ||
| European population (low intake): no association with soy or isoflavones (74) | ||||
| Tomatoes or lycopene | Increased plasma IGFBP-3 concentration in ferrets receiving lycopene (8) | Tomato products: no association (64, 71, 74; 1× inverse with IGF-I/IGFBP-3 ratio; ref. 71); inverse association (66) | Lycopene intervention: IGF-I decreased in both intervention and control groups (85) | Insufficient; inverse association with IGF (or IGF/IGFBP-3 ratio); positive association with IGFBP-3 |
| Lycopene intake: no association (positive with IGFBP-3; ref. 67) |
| Nutrient/food group . | Studies using animal models* . | Cross-sectional studies in humans† . | Experimental studies in humans* . | Conclusion . |
|---|---|---|---|---|
| Energy | Severe energy restriction results in 25% decreased serum IGF-I (review; ref. 75) | No association (8×; refs. 68, 63, 69, 71, 66, 72, 73, 74) | Serum IGF-I is markedly lowered by energy deprivation (review; ref. 60) | Probable; positive association with extremes of energy intake |
| Positive association (2×; refs. 67, 70) | No association within normal range of energy intake | |||
| Protein (total or animal) | Severe protein restriction results in 50% decreased serum IGF-I (review; ref. 61) | No association with total and/or animal protein (6×; refs. 64, 66, 71, 73, 74) | Serum IGF-I is markedly lowered by protein deprivation (review; ref. 60) | Possible; positive association with intake of animal protein |
| Positive association with total or animal protein (5×; refs. 63, 67, 68, 69, 70) | ||||
| Alcohol | Chronic alcohol feeding may increase or decrease serum IGF-I | No association with IGF-I (5×; refs. 64, 67, 68, 72, 74) | Decline in serum IGF-I after drinking alcohol (1 d only) | No association within normal range of alcohol consumption |
| Minor inverse association (66) | ||||
| Positive association (in men only; ref. 62) | ||||
| Minerals (total or zinc) | Mineral (e.g., zinc, copper, and magnesium) deficiency is associated with low serum IGF-I levels | No association with zinc (68) | Zinc supplementation: increased serum IGF-I levels in specific populations (anemic, non-insulin-dependent diabetes mellitus) | Probable; positive association with zinc or a combination of minerals (including supplements) |
| Positive association with zinc (63, 67) | ||||
| Positive association with combined intake of five minerals (including zinc; ref. 69) | ||||
| Dairy products | Only studies relating to IGF-I content in milk (review; ref. 79) | No association with total dairy (64, 66, 68, 71; one positive association with milk only; ref. 71) | Serum IGF-I increased 10% with three servings of dairy daily (77) | Possible; positive association with consumption of dairy/milk |
| Positive association with total dairy (67) | ||||
| Positive association with milk (65, 69) | ||||
| Soy or isoflavones | Decrease in serum IGF-I in animals on soy protein/phytoestrogen diet (93, 94) | Asian populations: no association with soy or isoflavones (72); positive association (in men only; ref. 73) | Soy protein vs. milk protein: stronger increase in IGF-I in soy protein group (99, 100) | Insufficient; association could be positive or inverse, depending on dose and population |
| European population (including vegans/vegetarians): no association with soy protein (borderline positive with soy milk; ref. 68) | Soy protein with vs. without isoflavones: trend towards lower IGF-I with isoflavones (postmenopausal women only; ref. 98) or no difference between groups (101) | Effects of soy protein and isoflavones/phytoestrogens to need be disentangled (possibly opposite effects) | ||
| European population (low intake): no association with soy or isoflavones (74) | ||||
| Tomatoes or lycopene | Increased plasma IGFBP-3 concentration in ferrets receiving lycopene (8) | Tomato products: no association (64, 71, 74; 1× inverse with IGF-I/IGFBP-3 ratio; ref. 71); inverse association (66) | Lycopene intervention: IGF-I decreased in both intervention and control groups (85) | Insufficient; inverse association with IGF (or IGF/IGFBP-3 ratio); positive association with IGFBP-3 |
| Lycopene intake: no association (positive with IGFBP-3; ref. 67) |
NOTE: “Probable”: The association (or lack of) is consistently shown in cross-sectional studies and confirmed by experimental studies in animals and in humans. “Possible”: Association in some but not all cross-sectional studies; experimental studies in animals consistently show the association; however, no or only one experimental study in humans is available. “Insufficient”: Only very few studies have been conducted and these are not entirely consistent.
References are only included if a review article is available or if literature is limited to one or two studies.
All published cross-sectional studies (n = 13) are included; references of studies including >500 individuals are printed in bold.