European Journal of Clinical Nutrition (2003) 57, 1079–1088
& 2003 Nature Publishing Group All rights reserved 0954-3007/03 $25.00 www.nature.com/ejcn
ORIGINAL COMMUNICATION Effects of dietary intervention on IGF-I and IGFbinding proteins, and related alterations in sex steroid metabolism: the Diet and Androgens (DIANA) Randomised Trial R Kaaks1*, C Bellati2, E Venturelli2, S Rinaldi1, G Secreto3, C Biessy1, V Pala2, S Sieri2 and F Berrino2 1
Hormones and Cancer Group, International Agency for Research on Cancer, Lyon, France; 2Unit of Epidemiology Istituto Nazionale dei Tumori, Milan, Italy; and 3Unit of Nuclear Medicine, Istituto Nazionale dei Tumori, Milan, Italy
Objective: To assess the effects of a comprehensive change in dietary composition on endogenous hormone metabolism. The specific aim was to examine whether this intervention could lead to favourable changes in insulin sensitivity, levels of IGF-I and IGF-binding proteins (IGFBPs), and total and bioavailable testosterone and estradiol, that would be expected to reduce breast cancer risk. Design: Randomised dietary intervention study; duration of 5 months. Subjects: From a total of 99 postmenopausal women, who had elevated baseline plasma testosterone levels, 49 women were randomly assigned to the dietary intervention arm and the other 50 to a control group. Interventions: Main aspects of the dietary intervention were reductions in the intake of total fat and refined carbohydrates, an increase in the ratio of n-3 over n-6 plus saturated fatty acids, and increased intakes of foods rich in dietary fibre and phytooestrogens. Results: Relative to the control group, women of the intervention group showed a significant reduction of body weight, waist circumference, fasting serum levels of testosterone, C peptide, glucose, and insulin area after glucose tolerance test, and a significant increase of serum levels of sex hormone-binding globulin, IGFBP-1, -2, and growth hormone-binding protein. Serum levels of IGF-I did not change. Conclusion: This comprehensive dietary intervention strategy proved to be successful in inducing changes in endogenous hormone metabolism that might eventually result in reduced breast cancer risk. Additional studies are needed to show whether the dietary intervention and related hormonal changes can be both maintained over longer periods, of at least several years. European Journal of Clinical Nutrition (2003) 57, 1079–1088. doi:10.1038/sj.ejcn.1601647 Keywords: diet; growth hormone; estradiol; intervention study; growth hormone-binding protein; sex hormone-binding
globulin
Introduction Recent epidemiological studies have provided evidence that comparatively elevated plasma levels of IGF-I, either as absolute concentrations or relative to levels of IGFBP-3, IGF’s major plasmatic binding protein, are related to an increased
*Correspondence: R Kaaks, Hormones and Cancer Group, International Agency for Research on Cancer, 150 Cours Albert Thomas, 69372 Lyon Cedex 08, France. E-mail:
[email protected]. Received 1 May 2002; revised 10 July 2002; accepted 21 August 2002
risk of breast cancer (Hankinson et al, 1998; Toniolo et al, 2000). Other studies have shown similar associations of plasma IGF-I levels with the risk of developing cancers of the prostate (Chan et al, 1998; Stattin et al, 2000) or colorectum (Giovannucci et al, 2000; Palmqvist et al, 2002), and ovary (Kaaks & Lukanova, 2001; Lukanova et al, 2002). In addition to these epidemiological relations, there is extensive experimental evidence that IGF-I can enhance tumour development by stimulating cell proliferation, by inhibiting programmed cell death (apoptosis), and by affecting the degree of cellular differentiation (Khandwala et al, 2000).
DIANA randomized trial R Kaaks et al
1080 IGF-I and at least six different IGF-binding proteins (IGFBPs) are synthesised in the liver, which is the origin of more than 80% of these peptides in blood. Circulating IGF-I can thus be considered a classical hormone, exerting endocrine effects on other target tissues than that in which it was produced. The IGFBPs regulate the size of the IGF-I pool in blood and the efflux of IGF-I from blood to target tissues. Besides the liver, IGF-I and IGFBPs are also synthesised by most other tissue types, where the peptides exert paracrine and autocrine effects (Jones & Clemmons, 1995). One key function of IGF-I is the stimulation of anabolic processes and body growth. In addition, IGF-I, and also insulin, have a central role in regulating plasma levels of bioavailable sex steroids. In vitro, IGF-I and insulin both inhibit the synthesis of sex hormone-binding globulin (SHBG) by liver, and stimulate the synthesis of androgens in ovarian and adrenal tissues (Kaaks, 1996; Poretsky et al, 1999). Plasma levels of both insulin and IGF-I generally correlate inversely with levels of SHBG, in both men and women (Pugeat et al, 1991). In women with ovarian hyperandrogenism (polycystic ovary syndrome), plasma levels of insulin correlate directly with levels of D4androstenedione (D4A), total testosterone (T), and free testosterone (fT) unbound to SHBG or albumin (Kaaks, 1996; Poretsky et al, 1999). Circulating levels of both insulin and IGF-I are regulated as a function of available energy and nutrients essential for growth (eg, amino acids, calcium), from the diet, or from body reserves (Thissen et al, 1994), and elevated levels of insulin and/or IGF-I might mediate effects of nutrition on plasma sex steroid profiles. In this report, we present results from the Diet and Androgens (DIANA) study, a nutritional intervention study conducted among postmenopausal women that aimed at investigating the effects of a comprehensive change in diet on levels of endogenous hormones known to be related to breast cancer risk (Berrino et al, 2001a). The intervention diet, consumed ad libitum, was designed to lower plasma insulin levels, by lowering the intake of total fat and foods rich in sugar and refined carbohydrates, by increasing the proportion of monounsaturated and n-3 polyunsaturated fatty acids, and by increasing the consumption of lowglycaemic index foods such as unrefined cereals, legumes, and vegetables. In addition, the diet was rich in phytooestrogens, both in the form of isoflavones and lignans, with the principal scope of reducing levels of endogenous androgens and oestrogens. The ultimate aim of this study was to determine whether this type of diet would be worth investigating in long-term trials to reduce the risk of breast cancer. The effects of this intervention on plasma levels of SHBG, insulin, estradiol (E2), and T were reported previously (Berrino et al, 2001b). In the present report, we describe the effects on plasma levels of IGF-I, IGFBPs-1, -2 and -3, and growth hormone (GH), as well as of growth hormonebinding protein (GHBP), which is a marker of growth hormone receptor levels in liver and other tissues (Baumann, European Journal of Clinical Nutrition
2001). In addition, we examine the interrelations between levels of these peptides, and levels of SHBG and the sex steroids before and after the intervention.
Study subjects and methods Subjects In all, 312 healthy women, age 50–65 y, from the Milan area (northern Italy) volunteered to take part in the study after advertisements had been placed in the local media. Eligibility criteria were that women were postmenopausal for at least 2 y, had at least one ovary, did not use hormonal replacement therapy during the previous 6 months, had no history of cancer, did not follow a vegetarian, macrobiotic or other medically prescribed diet, and did not receive any treatment for diabetes. Written informed consent was obtained from all the women and the Scientific and Ethical Committee of the Milan Cancer Institute approved the study.
Study design The levels of T in the serum of the volunteers were determined (prebaseline), and the 104 women in the upper tertile (T40.38 ng/ml) were selected for the study. With the exception of two close friends, who were allocated to the same group, these women were individually randomised to the intervention and control groups (52 women each), stratified for age (above or under the median of 58 y), prebaseline serum T (three levels), and prebaseline fasting insulin (three levels). Women were selected on the basis of serum T level because its measurement is highly reliable (Muti et al, 1996), and it predicts breast cancer risk equally as well as serum oestrogen levels (Bouchard et al, 1993; Berrino et al, 1996). The women in the intervention group agreed to adhere to the diet described below for 4.5 months (between months 1 and 5 of the study). The control women were not given any information about this diet, nor any dietary instruction, but were advised to increase their consumption of fruit and vegetables according to the cancer prevention decalogue of the Europe Against Cancer program, a leaflet largely available to the general population. Before the start and at the end of the intervention, fasting blood samples and 24-h urine samples were taken and stored at 301C for hormone assays. An oral glucose tolerance test was also performed, involving collection of blood samples 1, 2, and 3 h after the ingestion of 100 g of glucose.
Dietary intervention Women in the intervention group were invited for common meals and cooking classes twice a week for 18 weeks. On each occasion, the menu was different, but mainly based on Mediterranean vegetarian and macrobiotic recipes. We recommended that the same foods should be consumed on a daily basis at home, but did not prescribe menus. However,
DIANA randomized trial R Kaaks et al
1081 we did provide written instructions that indicated how to replace meat, eggs, and dairy products by vegetable sources of essential amino acids, vitamins, and minerals; recommended that meat, eggs, and dairy products should not be eaten more than once a week; urged reducing the consumption of refined carbohydrates (sucrose, white bread, refined flour), replacing these by whole-grain cereal products, and using fruit or fermented cereal as edulcorants; and recommended cooking with little added fat and salt. The women were also encouraged to eat at least one portion of a soy product (soy milk, miso soup, tofu, tempeh, or soy beans) every day, to season moderately with unrefined olive oil and various seeds but not dairy fats, and to consume fish and seaweed. Every week each woman received a 1 kg loaf of bread made from whole-wheat flour and 8% flax seed (half whole seeds and half milled), occasionally mixed with oats or rye, and also a free pack of other recommended products that are not a normal part of the northern Italian diet. Additional details on the types of food consumed by the intervention and control groups have been given previously (Berrino et al, 2001b). In the first month of the study, participants were asked to change their habits gradually in order to prevent adverse reactions because of excessive fermentation in the bowel. The diet was ad libitum, and no advice was given to reduce total food intake or to count calories.
Assessment of dietary intake and anthropometric measurements Before randomisation, all women compiled a food-frequency questionnaire developed for the European Prospective Investigation into Cancer and Nutrition (EPIC) (Pisani et al, 1997). During the study, compliance with dietary recommendations was monitored by 24-h food-frequency diaries, which were filled in 24 times by the intervention group and 10 times by the control women. In the fourth month of the study, all women (ie, also those in the control group) were interviewed and asked to recall everything they had eaten in the preceding 24 h, including quantities. Data were collected with the computerised EPIC 24-h dietary recall system (Slimani et al, 2000), which was then used to estimate absolute intakes of nutrients and energy in the two groups. The system makes use of the Italian food composition database (Salvini et al, 1998), which also includes several foods used in macrobiotic recipes. The average consumption of isoflavonoids and lignans by the intervention and the control groups was estimated from available databases on the phytooestrogen content of foods (Polonsky et al, 1986; Mazur & Adlercreutz, 1988; Thompson et al, 1991; Eberhardt et al, 1994; Eidson et al, 1994; Whelan et al, 1994; Dawson et al, 1995; Reinli & Block, 1996; Nesbitt & Thompson, 1997; Wakai et al, 1999) and from the food-frequency diaries, using as standard portion sizes those derived from the interviews. Height, weight, waist circumferences (at natural waist when clearly identifiable or midway between lower rib and
iliac crest), and hip circumference (at crotch level) were measured at the beginning and at the end of the study.
Compliance, and subjects excluded from statistical analysis Of the 52 women of the intervention group, 50 followed the whole dietary programme. Two women followed only about half of the programme but, following the ‘intention to treat’ principle, were included in all the analyses. Only five women were absent more than five times from the 36 lessons and common meals. Urinary daidzein and equol levels were used as the indicator of compliance with soy consumption. Two women from the intervention group and one woman from the control group were excluded because they received hormonal drugs during the study period. Two other women from the control group were excluded because they did not attend the final examination. Thus, a total of 99 women were analysed: 50 in the intervention group, and 49 controls. Of these, four (two in the intervention group and two controls) had missing values for fasting insulin, and five (one in the intervention and four in the control group) had missing values for the oral glucose tolerance test.
Hormone assays For insulin, samples were assayed within 2 weeks of collection. To reduce the effects of interassay variability, for SHBG, T, E2, IGF-I, IGFBP-1, -2, -3, GH, and GHBP, baseline and final serum samples of the same woman were analysed in the same batch. In each analytical batch, an equal number of samples from women of the intervention group and from women of the control group were measured. Laboratory personnel could not distinguish repeated samples belonging to the same subject, and were blind to their dietary status. Radioimmunoassay (RIA) kits from ORION Diagnostic (Turku, Finland) for T and E2; IRMA (immunoradiometric) kits from Farmos (Oulunsalo, Finland) for SHBG, and MEIA kits from ABBOTT (Abbott Park, IL, USA) for insulin were used. The coefficients of intra- and interassay variation in eight replicates were: 4.2 and 12.5% for a T value of 0.420 ng/ ml; 5.2 and 11.1% for an E2 concentration of 10 pg/ml; 3.5 and 6.7% for an SHBG value of 34.0 nmol/l; and 2.5 and 4.6% for an insulin value of 14.2 mIU/ml. IGF-I, IGFBP-1, IGFBP-3, and GH were measured by double-antibody IRMA kits. IGFBP-2 was measured by RIA, and GHBP by enzyme-linked immuno sorbent assay (ELISA). All the reagents were from Diagnostic System Laboratories (Webster, TX, USA). Total IGF-I was measured after acid– ethanol precipitation of IGFBPs. To control for the quality of these peptide measurements, three standard sera were inserted randomly in each batch. The mean intra- and interassay coefficients of variation were 1.5 and 3.4%, respectively, for IGF-I, 4.9 and 10.5% for IGFBP-1,10.8 and 16.0% for IGFBP-2, 1.5 and 5.1% for IGFBP-3, 4.3 and 8.7% for GH, and 7.2 and 8.9% for GHBP. European Journal of Clinical Nutrition
DIANA randomized trial R Kaaks et al
1082 Free T (fT) and free E2 (fE2) were calculated from total serum T, E2, and SHBG using a set of theoretical equations based on the mass action laws, and validated for measurements of fT and fE2 in serum samples of postmenopausal women. This set of equations relies on the evidence that interindividual variation in the concentrations of fT and fE2 in blood is determined mainly by the interaction between SHBG, total T, and total E2, and that competitive binding of SHBG with other hormones in blood (eg, dihydrotestosterone) does not influence this equilibrium much (Vermeulen et al, 1999; Rinaldi et al, 2002).
Statistical analyses The statistical analysis focused on changes in hormonal and other relevant variables, calculated as the difference between end-of-study and baseline values for each woman. Hormone values were log transformed to obtain approximately normal frequency distributions, and effects of the intervention on hormone levels were examined as geometric mean differences. The statistical significance of mean changes in the intervention group compared to controls was assessed by analysis of variance (ANOVA) (using log-transformed variables for the various hormone and growth factor measurements). All ANOVAs were stratified according to the blocking (stratification) scheme used for the randomisation. As the numbers of observations within the various blocks were not balanced, all ANOVAs employed generalised linear models, using the SAS statistical software package. All P-values are for two-sided statistical tests.
Results Changes in the dietary intake in the intervention group, as compared to the controls, have been described in detail previously (Berrino et al, 2001a). In brief, 24-h recall data indicated that, towards the end of the 18-week intervention period, the total energy intake was about 250 kcal/day lower for the intervention group than for the control group. Women in the intervention group shifted from animal to vegetable sources of protein (animal proteins accounted for 29% of total protein against 60% in the control group) and fats (28% from animal sources against 43% in the control group), from simple to complex carbohydrates (34% of total carbohydrates from simple sugars against 44% among controls), and from low to high dietary fibre (35.5 against 23.3 g/day). Furthermore, a soy product was consumed on average 1.7 times a day, flax seeds (a very rich source of lignans) were eaten every day in bread or cookies, and seaweed (a source of lignans and n-3 fatty acids) was used every other day as an ingredient of various dishes. The women in the control group rarely, if ever, consumed any of these food items. The intervention group also increased their consumption of whole rice and cereal products, nuts and seeds, legumes, cruciferous vegetables and berries, and had significantly higher intakes of these foods than the control European Journal of Clinical Nutrition
group. Other fruits and vegetables were consumed with equal frequencies by the intervention and control groups. The estimated intake of isoflavonoids in the intervention group was around 40 mg/day, against 2 mg/day for the control group. This large difference in estimated isoflavonoid intake was confirmed by measurements in a 24-h urinary sample collected at the end of the intervention period. Between baseline (January: month 1) and the end of the 5month intervention period (June: month 5), the dietary intervention group showed significant decreases in weight for the intervention group compared to the controls ( 4.06 vs 0.56 kg), and concurrent changes in waist and hip circumferences (Table 1). In addition, the intervention led to significant decreases in plasma C-peptide, fasting glucose, and average area under the insulinaemic curve, and significant increases in IGFBP-1 and -2, compared to the control group (Table 2). Plasma GH levels increased strongly (+54.2%) only in the intervention group but, as a result of the large interindividual variation in the measured GH levels (GH has a pulsatile daily secretion pattern), this increase was not statistically significant from changes among the control subjects. Levels of GHBP increased in the intervention group and decreased in the control group, and the difference between these changes in the two groups was statistically significant. The dietary intervention did not have any significant effect on levels of IGF-I and IGFBP-3. In addition to these results, we previously reported a significant increase in SHBG (+25%), and significant decreases in the intervention group of T ( 18%) (Berrino et al, 2001b). Serum E2 levels showed an equal reduction, but this reduction was not statistically significant. Calculated levels of fT and fE2 both decreased significantly in the intervention group, compared to the controls. After adjustment for changes in body weight and/or waist circumference, many of the effects of the
Table 1 Mean anthropometric indices before and after dietary intervention Mean N
January
June
Change
P-value
BMI (kg/m2) Intervention Control
50 49
26.88 27.36
25.26 27.14
1.62 0.22
0.0001
Waist (cm) Intervention Control
50 49
84.02 83.43
80.15 82.94
3.88 0.49
0.0001
Hip (cm) Intervention Control
50 49
102.34 102.67
99.87 102.87
2.47 0.20
0.0001
WHR (cm) Intervention Control
50 49
0.82 0.81
0.80 0.80
0.02 0.01
0.0045
DIANA randomized trial R Kaaks et al
1083 Table 2 Mean hormonal parameters before and after dietary intervention Geometric mean N
January 1996
June 1996
Change (%)
P-value
P-value weight
P-value waist
P-value w w
SHBG (nmol/l) Intervention Control
50 49
36.03 36.32
45.10 37.61
25.2 3.6
0.0001
0.28
0.01
0.29
Testosterone (ng/ml) Intervention Control
50 49
0.41 0.42
0.33 0.39
18.3 6.4
0.004
0.30
0.16
0.33
Estradiol (pg/ml) Intervention Control
50 49
8.62 8.30
7.07 7.84
18.0 5.7
0.13
0.85
0.51
0.86
Free testosterone (ng/ml) Intervention Control
50 49
0.006 0.010
0.005 0.010
28.6 8.2
o0.0001
0.16
0.02
0.18
Free estradiol (pg/ml) Intervention Control
50 49
0.23 0.22
0.17 0.21
23.4 6.8
0.05
0.79
0.39
0.81
Fasting insulin (mIU/ml) Intervention Control
48 47
4.82 4.63
4.31 4.87
10.4 5.1
0.14
0.72
0.80
0.77
Insulin area (mIU/ml) Intervention Control
47 43
7299.95 6878.51
6739.74 7525.41
7.7 9.4
0.0404
0.66 0.64*
0.69 0.76*
0.65 0.62*
Fasting glucose (mg/dl) Intervention Control
50 49
91.66 93.55
86.47 92.44
5.7 1.2
0.026
0.05
0.08
0.06
Glycaemic area (mg/dl) Intervention Control
47 43
20000.38 20291.05
20967.78 21473.88
4.8 5.8
0.85
0.24 0.77*
0.44 0.11*
0.23 0.62*
C-Peptide (ng/ml) Intervention Control
50 47
2.57 2.54
2.08 2.41
19.1 5.3
0.03
0.20
0.06
0.19
GH (ng/ml) Intervention Control
50 49
1.25 1.10
1.92 1.20
54.2 9.5
0.22
0.91
0.41
0.89
GHBP (ng/ml) Intervention Control
50 48
0.94 0.89
1.21 0.65
29.1 26.8
0.002
0.006
0.009
0.006
IGF 1 (ng/ml) Intervention Control
50 49
146.65 135.18
138.00 129.29
5.9 4.4
0.64
0.41
0.26
0.37
IGFBP 1 (ng/ml) Intervention Control
49 47
40.76 48.07
45.73 44.98
12.2 6.4
0.015
0.70
0.49
0.77
IGFBP 2 (ng/ml) Intervention Control
50 47
6.15 6.11
8.02 6.52
30.4 6.6
0.0004
0.21
0.03
0.22
European Journal of Clinical Nutrition
DIANA randomized trial R Kaaks et al
1084
Table 2 (continued) Geometric mean
IGFBP 3 (ng/ml) Intervention Control
N
January 1996
June 1996
50 46
3107.9 3053.51
2968.2 2987.15
Change (%)
4.5 2.2
P-value
P-value weight
P-value waist
P-value w w
0.21
0.66
0.18
0.61
All P-values are adjusted for treatment/intervention status, age class, insulin level, testosterone level. P-value weight: P-value for change in hormonal parameter with additional adjustment for change in weight. P-value waist: P-value for change in hormonal parameter with additional adjustment for change in waist circumference. P-value w wt: P-value for change in hormonal parameter with additional adjustment for changes in both weight and waist circumference. * Insulin area fasting and glycaemic area fasting.
dietary intervention were reduced, and no longer statistically significant. The changes in levels of fasting glucose and GHBP, however, remained statistically significant even after these adjustments. Table 3 shows Spearman’s coefficients of correlation between anthropometric and hormonal variables at baseline (intervention and control groups combined), as well as between the longitudinal changes in these parameters over the intervention period (intervention group only). Indices of insulin resistance (fasting C-peptide, insulin, and areas under the glycaemic and insulinaemic curves after intake of the oral glucose dose) correlated strongly with BMI, waist circumference, and WHR, whereas each of these indices and anthropometric measurements correlated inversely with levels of IGFBP-1 and -2. Anthropometric indices of adiposity, as well as fasting insulin and glucose levels, correlated inversely with levels of GH, but not with IGF-I. Finally, each of the glycaemic and insulinaemic indices, as well as the anthropometric measures of adiposity, correlated directly with serum E2, fT, and fE2, and inversely with SHBG. Longitudinally, and within the intervention group only, changes (in most subjects a decrease) in the anthropometric indices of adiposity (BMI, waist and hip circumferences, and WHR) all showed inverse correlations with changes in SHBG level (significantly so only for BMI). Furthermore, decreases in BMI, and waist and hip circumferences were associated with increases in SHBG, and with reductions in levels of total and bioavailable testosterone (again, significantly only for BMI), but not with changes in estradiol. There were no clear correlations between changes in insulin, C-peptide, or postload glycaemic and insulinaemic areas with changes in SHBG level, although changes in fasting insulin correlated directly with changes in free testosterone.
Discussion This study shows multiple endocrine effects of a comprehensive dietary intervention, based on reduced intakes of refined carbohydrates and total fat, and increased intakes of n-3 fatty acids, dietary fibre, and phytooestrogens (both European Journal of Clinical Nutrition
isoflavones and lignans). Major observations were reductions in fasting glucose, C-peptide, area under the insulinaemic curve, T and fT, and increases in IGFBP-1, IGFBP-2, SHBG and GHBP. In addition, there was a nonsignificant decrease in total E2 and a nonsignificant increase in GH. The dietary intervention had no effect on serum levels of IGF-I and IGFBP-3. The reductions in fasting glucose, (fasting) C-peptide, and insulin area most likely reflect improvements in insulin sensitivity. This improvement in insulin sensitivity appears to be explained, to a large extent, by the reductions in body weight and body fat stores, since the changes related to the intervention were no longer statistically significant after adjustment for changes in weight and/or waist circumference. Besides the decrease in body weight and fat stores, the reduced intake of refined carbohydrates and high-glycaemic index foods (Ludwig, 2002) and the relative increase in the intake of monounsaturated and n-3 fatty acids (Lovejoy, 1999) may also have contributed to the improvement in insulin sensitivity. Energy restriction generally leads to increases in IGFBPs-1 and -2 (Thissen et al, 1994; Kaaks & Lukanova, 2001), and this can be explained by concomitant reductions in insulin levels. Insulin is a key regulator of IGFBP-1 levels, inhibiting its synthesis by liver and other tissues (Brismar et al, 1995), and plasma insulin generally correlates also inversely with circulating IGFBP-2 levels (Kaaks & Lukanova, 2001). These various relations were reflected in our data, which showed inverse cross-sectional correlations of IGFBP-1 and -2 with BMI, waist circumference measurements, and insulin, and increases in IGFBP-1 and -2 after the dietary intervention, in parallel with reductions in insulin levels. The increases in IGFBP-1 and -2 most likely caused a reduction in circulating free IGF-IFa small fraction of circulating IGF-I that is not bound by any IGF-binding protein, and that probably reflects the fraction most rapidly available to tissues for binding to cellular receptors. Other studies (Frystyk et al, 1995; Nam et al, 1997; Nyomba et al, 1997) have shown inverse correlations of plasma IGFBP-1 and -2 with plasma free IGF-I levels. Unfortunately, we had technical problems with an assay kit for measurement of free IGF-I, and thus
Table 3 Cross-sectional and longitudinal relation (Spearman correlations) of indices of weight excess and of glucose and insulin metabolism with insulin-dependent binding proteins (IGFBP1, IGFBP-2 and SHBG), total and bioavailable sex steroids, and GH/IGF-I axis Cross-sectional relation at month 1 (baseline): intervention and control groups combined Indices of excess weight
BMI (kg/m2) Indices of excess weight Waist (cm) WHR(cm) Hip (cm)
Insulin-dependent binding proteins IGFBP1 (ng/ml) IGFBP2 (ng/ml) SHBG (nmol/l)
WHR
Indices of glucose and insulin metabolism Hip
Fasting glucose
Fasting insulin
C-peptide
Glycaemic area
0.90 0.61 0.79
0.82 0.70
0.21
0.45
0.47
0.49
0.25
0.58
0.63
0.60
0.37
0.40
0.55
0.59
0.50
0.41
0.45
0.74
0.33
0.32
0.29
0.22
0.68
0.28
0.38
0.34
0.41
0.39
0.24
0.21
0.67
0.53
0.37
0.43 0.44 0.52
0.50 0.46 0.54
0.49 0.47 0.51
0.26 0.19 0.31
0.37 0.48 0.42
0.38 0.46 0.48
0.37 0.53 0.46
0.29 0.40 0.38
0.14
0.14
0.08
0.02
0.20
0.06
0.47
0.46
0.25
0.28
0.47
0.51
0.30
0.52
0.27
0.57
0.37
0.55
Insulin area
BMI Waist (kg/m2)
WHR
Hip
Indices of glucose and insulin metabolism Fasting Fasting C-peptide Glycaemic Insulin glucose insulin area area
0.59 0.28 0.50
0.79 0.34
0.21
0.13
0.003
0.13
0.15
0.22
0.16
0.03
0.16
0.39
0.06
0.15
0.21
0.02
0.33
0.59
0.13
0.02
0.15
0.33
0.25
0.22
0.18
0.22
0.20
0.12
0.25
0.10
0.20
0.14
0.58
0.43 0.33 0.38
0.34 0.17 0.34
0.14 0.12 0.27
0.12 0.01 0.19
0.37 0.20 0.26
0.05 0.20 0.18
0.01 0.20 0.09
0.15 0.34 0.18
0.21 0.05 0.15
0.43 0.20 0.17
0.14
0.15
0.29
0.20
0.01
0.24
0.06
0.22
0.18
0.08
0.16
0.38
0.34
0.36
0.45
0.27
0.12
0.23
0.04
0.29
0.08
0.19
0.17
0.42
0.31
0.34
0.31
0.04
0.09
0.09
0.13
0.15
0.02
0.19
0.15
0.16
0.33
0.49
0.39
0.38
0.38
0.15
0.13
0.11
0.10
0.06
0.09
0.11
0.03
0.07
GH/IGF-I axis GHBP (ng/ml) GH (ng/ml))
0.10 0.39
0.09 0.40
0.01 0.35
0.14 0.24
0.05 0.33
0.04 0.38
0.12 0.22
0.003 0.17
0.04 0.20
0.11 0.05
0.06 0.06
0.08 0.15
0.17 0.19
0.28 0.20
0.07 0.34
0.25 0.27
0.12 0.02
0.004 0.12
IGF-I (ng/ml) IGFBP3 (ng/ml)
0.15 0.19
0.14 0.21
0.04 0.21
0.19 0.07
0.15 0.32
0.07 0.21
0.11 0.13
0.04 0.34
0.004 0.10
0.04 0.12
0.10 0.11
0.20 0.28
0.06 0.18
0.18 0.04
0.08 0.16
0.09 0.07
0.06 0.19
0.11 0.22
1085
European Journal of Clinical Nutrition
Total and bio-available sex steroid Testosterone 0.20 (T) (ng/ml) Free testosterone 0.48 (fT) (ng/ml) Oestradiol (E2) 0.58 (pg/ml) 0.64 Free oestradiol (fE2 (pg/ml)
Indices of excess weight
DIANA randomized trial R Kaaks et al
Indices glucose and insulin Fasting glucose (mg/dl) Fasting insulin (mlU/ml) C-peptide (ng/ml) Glycaemic area (mg/dl) Insulin area (mIU/ml)
Waist
Correlations of changes owing to intervention: intervention group only
DIANA randomized trial R Kaaks et al
1086 could not reproduce these findings and quantify the change in free IGF-I. The principal stimulus for the synthesis of IGF-I (and also IGFBP-3) in liver and other tissues is provided by GH. However, the capacity of GH to stimulate IGF-I synthesis and to stimulate related growth processes depends strongly on the availability of energy and nutrients (eg, essential amino acids) from diet and body reserves. Reductions in IGF-I and IGFBP-3 generally observed with energy and/or protein restrictions can be explained by the resistance of liver and other tissues against the action of GH. This GH resistance is generally illustrated by a strong increase in circulating GH levels during energy restriction, in contrast to the drops in IGF-I and IGFBP-3 (Thissen et al, 1994; Kaaks & Lukanova, 2001). Our data showed no drop in either IGF-I or IGFBP-3 in the dietary intervention group, but did show a 54% (but statistically nonsignificant) increase in serum GH levels. An alternative explanation for the rise in GH is a reduction in free IGF-I, which exerts negative feedback control over pituitary GH secretion (Tannenbaum et al, 1983; Chapman et al, 1998). Experimental studies with liver tissue in vitro (Baxter & Turtle, 1978; Tollet et al, 1990), and clinical studies in diabetes patients (Hanaire-Broutin et al, 1996) have shown that insulin provides a key stimulus for GH-receptor synthesis. Furthermore, plasma GHBP is reduced in insulindependent diabetics (Mercado & Baumann, 1995; HanaireBroutin et al, 1996), who have a strongly reduced endogenous production and hepatic exposure to insulin, but not in noninsulin-dependent diabetics, who generally have normal or even elevated endogenous insulin levels. GHBP is identical to the external domain of GH-receptors, and plasma GHBP in humans appears to originate mostly or entirely from the cleavage and release of this external domain into the circulation (Baumann, 2001). Contrary to these observations, however, and contrary to some previous intervention studies (Rasmussen et al, 1996), our data did not show a decrease, but an increase, in GHBP with weight loss in the dietary intervention group. The absence in our data of a direct cross-sectional association of serum GHBP with BMI, however, is consistent with at least one other study in postmenopausal women (Bondanelli et al, 2001). Major changes in sex steroid metabolism were a significant decrease in serum T levels and an increase in SHBG. The strong rise in SHBG can be explained by the decrease in insulin, and possibly by a reduction of IGF-I activity within liver caused by the increase in IGFBP-1 and -2 levels. Studies with liver cells in vitro have clearly shown that insulin and IGF-I are both key regulators of SHBG synthesis (Singh et al, 1990; Crave et al, 1995), and in cross-sectional human studies, both serum insulin and IGF-I generally have been found to be inversely correlated with SHBG level (Erfurth et al, 1996; Pfeilschifter et al, 1996). Besides the inhibition of SHBG synthesis, insulin and IGF-I have both been shown to stimulate ovarian, and possibly adrenal, synthesis of androgen production (Kaaks, 1996; Poretsky et al, 1999). The very signiEuropean Journal of Clinical Nutrition
ficant drop in testosterone levels during the intervention may thus be explained by the reductions in insulin, and most likely in free IGF-I (through increases in IGFBP-1 and -2). The effects of the dietary intervention on levels of IGFBP1, -2, SHBG, T, and other endocrine parameters were stronger than one would have expected on the basis of cross-sectional relationships of these endocrine parameters with BMI, waist circumference, or fasting insulin. One possible explanation for this is that alterations in these hormone levels are stronger during the dynamic phase of weight loss or negative energy balance than alterations induced by a reduction in BMI per se. This would imply that some of the observed alterations in hormone levels would disappear when the dynamic phase of weight loss ends, and when body weight stabilises (even if stabilisation occurs at a lower than baseline value). Verification of this hypothesis would have required the collection of additional blood samples, at regular comparatively short intervals (eg, monthly) for some time after termination of the active intervention period; unfortunately, these additional samples were not available. Another possible explanation would be that, in addition to weight loss induced by the dietary intervention, there were other specific effects owing to the change in dietary composition on hormone metabolism. For example, the increase in dietary fibre may have led to higher faecal elimination of conjugated sex steroids excreted with bile into the gut (Goldin et al, 1982). If indeed the endocrine changes observed in this study were sustainable over a longer period of time, the dietary intervention used here might be of interest for future breast cancer prevention studies. Indeed, the change in total and bioavailable plasma T and E2 in the intervention group was of roughly the same magnitude as the mean percentual differences observed between postmenopausal women who develop breast cancer, and control subjects who do not (The Endogenous Hormones and Breast Cancer Collaborative Group, 2002). Additional hormonal changes that might also contribute to reduce breast cancer risk are the reduction in insulin, and increases in IGFBP-1 and -2 (which may reduce IGF-I bioactivity within the breast). Besides estradiol, IGF-I and insulin may exert direct growth-promoting effects on breast epithelium (Kaaks, 1996; Papa & Belfiore, 1996; Chappell et al, 2001; Lai et al, 2001), and there is some evidence that both elevated circulating insulin (Bruning et al, 1992; Del Giudice et al, 1998; Yang et al, 2001) and elevated circulating free IGF-I (Li et al, 2001) may increase breast cancer risk, and worsen breast cancer prognosis (survival time) (Goodwin et al, 2002). An estimation of the breast cancer fraction that might be prevented with the type of diet examined in our study remains difficult, however, as favourable alterations in endogenous hormone levels relatively late in life may not have the same effect on risk as having had a more favourable hormone level lifetime. An intervention study with cancer as an end point (eg, in women at high genetic breast cancer risk) would be useful for this type of evaluation.
DIANA randomized trial R Kaaks et al
1087 In conclusion, this study was conducted to examine the effects of a comprehensive, multifactorial change in diet, aiming at a maximum possible effect on hormones. The aim was to maximally combine different possible dietary intervention effects rather than to identify and disentangle the effects of specific, individual aspects of the diet. We observed a number of important changes in endogenous hormone metabolism after a comprehensive dietary intervention, and these changes (reductions in plasma insulin, T, and E2, and increases in SHBG, IGFBP-1, and IGFBP-2) were all in the direction anticipated, and would all be expected to contribute to a reduction in the risk of breast cancer (Kaaks, 1996), as well as of other forms of cancer (Kaaks & Lukanova, 2001), if it was possible to maintain the diet in the long term. Further studies are needed to examine whether the alterations in hormone levels would persist if the dietary intervention was continued for periods longer than 5 months.
References Baumann G (2001): Growth hormone binding protein 2001. J Pediatr. Endocrinol. Metab. 14, 355–375. Baxter RC & Turtle JR (1978): Regulation of hepatic growth hormone receptors by insulin. Biochem. Biophys. Res. Commun. 84, 350–357. Berrino F, Muti P, Micheli A, Bolelli G, Krogh V, Sciajno R, Pisani P, Panico S & Secreto G (1996): Serum sex hormone levels after menopause and subsequent breast cancer. J. Natl. Cancer Inst. 88, 291–296. Berrino F, Bellati C, Secreto G, Camerini E, Pala V, Panico S, Allegro G & Kaaks R (2001a): Reducing bioavailable sex hormones through a comprehensive change in diet: the diet and androgens (DIANA) randomized trial. Berrino F, Bellati C, Secreto G, Camerini E, Pala V, Panico S, Allegro G & Kaaks R (2001b): Reducing bioavailable sex hormones through a comprehensive change in diet: the diet and androgens (DIANA) randomized trial. Cancer Epidemiol. Biomarkers Prev. 10, 25–33. Bondanelli M, Margutti A, Ambrosio M R, Plaino L, Cobellis L, Petraglia F & degli Uberti EC (2001): Blood growth hormonebinding protein levels in premenopausal and postmenopausal women: roles of body weight and estrogen levels. J. Clin. Endocrinol. Metab. 86, 1973–1980 Bouchard C, Despres JP & Mauriege P (1993): Genetic and nongenetic determinants of regional fat distribution. Endocr. Rev. 14, 72–93. Brismar K, Hilding A & Lindgren B (1995): Regulation of IGFBP-1 in humans. Prog. Growth Factor Res. 6, 449–456. Bruning PF, Bonfrer JM, van Noord PA, Hart AA, de Jong-Bakker M & Nooijen WJ (1992): Insulin resistance and breast–cancer risk. Int. J. Cancer 52, 511–516. Chan JM, Stampfer MJ, Giovannucci E, Gann PH, Ma J, Wilkinson P, Hennekens CH & Pollak M (1998): Plasma insulin–like growth factor–I and prostate cancer risk: a prospective study. Science 279, 563–566. Chapman IM, Hartman ML, Pieper KS, Skiles EH, Pezzoli SS, Hintz RL & Thorner MO (1998): Recovery of growth hormone release from suppression by exogenous insulin–like growth factor I (IGF–I): evidence for a suppressive action of free rather than bound IGF–I. J. Clin. Endocrinol Metab. 83, 2836–2842. Chappell J, Leitner JW, Solomon S, Golovchenko I, Goalstone ML & Draznin B (2001): Effect of insulin on cell cycle progression in MCF–7 breast cancer cells. Direct and potentiating influence. J. Biol. Chem. 276, 38023–38028.
Crave JC, Lejeune H, Brebant C, Baret C & Pugeat M (1995): Differential effects of insulin and insulin–like growth factor I on the production of plasma steroid–binding globulins by human hepatoblastoma–derived (Hep G2) cells. J. Clin. Endocrinol. Metab. 80, 1283–1289. Dawson T & Wynford Thomas D (1995): Does autocrine growth factor secretion form part of a mechanism which paradoxically protects against tumour development? Br. J. Cancer 71, 1136–1141. Del Giudice ME, Fantus IG, Ezzat S, McKeown–Eyssen G, Page D & Goodwin PJ (1998): Insulin and related factors in premenopausal breast cancer risk. Breast. Cancer. Res. Treat. 47, 111–120. Eberhardt MS, Lackland DT, Wheeler FC, German RR & Teutsch SM (1994): Is race related to glycemic control? An assessment of glycosylated hemoglobin in two South Carolina communities. J. Clin. Epidemiol. 47,1181–1189. Eidson M, Becker TM, Wiggins CL, Key CR & Samet JM (1994): Breast cancer among Hispanics, American Indians and non–Hispanic whites in New Mexico. Int. J. Epidemiol. 23, 231–237. Erfurth EM, Hagmar LE, Saaf M & Hall K (1996): Serum levels of insulin–like growth factor I and insulin–like growth factor– binding protein 1 correlate with serum free testosterone and sex hormone binding globulin levels in healthy young and middle– aged men. Clin. Endocrinol. (Oxf). 44, 659–664. Frystyk J, Vestbo E, Skjaerbaek C, Mogensen CE & Orskov H (1995): Free insulin–like growth factors in human obesity. Metabolism 44, 37–44. Giovannucci E, Pollak MN, Plate EA, Willett WC, Stampfer MJ, Majeed N, Colditz GA, Speizer FE & Hankinson SE (2000): A prospective study of plasma insulin–like growth factor–1 and binding protein–3 and risk of colorectal neoplasia in women. Cancer. Epidemiol. Biomarkers. Prev. 9, 345–349. Goldin BR, Adlercreutz H, Gorbach SL, Warram JH, Dwyer JT, Swenson L & Woods MN (1982): Estrogen excretion patterns and plasma levels in vegetarian and omnivorous women. N. Engl. J. Med. 307, 1542–1547. Goodwin PJ, Ennis M, Pritchard KI, Trudeau ME, Koo J, Madamas Y, Hartwick W, Hoffman B & Hood N (2002): Fasting insulin and outcome in early–stage breast cancer: results of a prospective cohort study. J. Clin. Oncol. 20, 42–51. Hanaire-Broutin H, Sallerin-Caute B, Poncet MF, Tauber M, Bastide R, Rosenfeld R & Tauber JP (1998): Insulin therapy and GH–IGF–I axis disorders in diabetes: impact of glycaemic control and hepatic insulinization. Diabetes Metab. 22: 245–250. Hankinson SE, Willett WC, Colditz GA, Hunter DJ, Michaud DS, Deroo B, Rosner B, Speizer FE & Pollak M (1998): Circulating concentrations of insulin-like growth factor, I and risk of breast cancer. Lancet 351, 1393–1396. Jones JI & Clemmons DR (1995): Insulin–like growth factors and their binding proteins: biological actions. Endocr. Rev. 16, 3–34. Kaaks R (1996): Nutrition, hormones, and breast cancer: is insulin the missing link? Cancer Causes Control 7, 605–625. Kaaks R & Lukanova A (2001): Energy balance and cancer the role of insulin and insulin-like growth factor-I. Proc. Nutr. Soc. 60: 91–106. Khandwala HM, McCutcheon IE, Flyvbjerg A & Friend KE (2000): The effects of insulin–like growth factors on tumorigenesis and neoplastic growth. Endocr. Rev. 21, 215–244. Lai A, Sarcevic B, Prall OW & Sutherland RL (2001): Insulin/insulin– like growth factor–I and estrogen cooperate to stimulate cyclin E– Cdk2 activation and cell cycle progression in MCF–7 breast cancer cells through differential regulation of cyclin E and p21(WAFl/ Cipl). J. Biol. Chem. 276, 25823–25833. Li BD, Khosravi MJ, Berkel HJ, Diamandi A, Dayton MA, Smith M & Yu H (2001): Free insulin-like growth factor-I and breast cancer risk. Int. J. Cancer. 91, 736–739. Lovejoy JC (1999): Dietary fatty acids and insulin resistance. Curr. Atheroscler. Rep. 1, 215–220 Ludwig DS (2002): The glycemic index: physiological mechanisms relating to obesity, diabetes, and cardiovascular disease. JAMA 287, 2414–2423.
European Journal of Clinical Nutrition
DIANA randomized trial R Kaaks et al
1088 Lukanova A, Soederberg S, Stattin P, Palmquist R, Lundin E, Biessy C, Rinaldi S, Riboli E, Hallmans G & Kaaks R (2002): Non–linear relationship of insulin-like growth factor(IGF-I) and IGF-I/IGFbinding protein-3 ratio with indices of adiposity and plasma insulin. Cancer Causes Control. 13, 509–516. Mazur W & Adlercreutz H (1988): Naturally occurring oestrogens in food. Pure. Appl. Chem. 70, 1759–1776. Mercado M & Baumann G (1995): Characteristics of the somatotropic axis in insulin dependent diabetes mellitus. Arch. Med. Res. 26, 101–109. Muti P, Trevisan M, Micheli A, Krogh V, Bolelli G, Sciajno R & Berrino F (1996): Reliability of serum hormones in premenopausal and postmenopausal women over a one-year period. Cancer. Epidemiol. Biomarkers. Prev. 5, 917–922. Nam SY, Lee EJ, Kim KR, Cha BS, Song YD, Lim SK, Lee HC & Huh KB (1997): Effect of obesity on total and free insulin-like growth factor (IGF)-1, and their relationship to IGF–binding protein (BP)-l, IGFBP-2, IGFBP-3, insulin, and growth hormone. Int. J. Obes. Relat. Metab. Disord. 21, 355–359. Nesbitt PD & Thompson LU (1997): Lignans in homemade and commercial products containing flax seed. Nutr. Cancer. 29, 222–227. Nyomba BL, Berard L & Murphy LJ (1997): Free insulin–like growth factor I (IGF–I) in healthy subjects: relationship with IGF–binding proteins and insulin sensitivity. J. Clin. Endocrinol. Metab. 82, 2177–2181. Palmqvist R, Hallmans G, Rinaldi S, Biessy C, Stenling E, Riboli E & Kaaks R (2002): Plasma IGF-I, IGF-binding protein-3 and risk of colorectal cancer: a prospective study in Northern Sweden. Gut. 50, 642–646. Papa V & Belfiore A (1996): Insulin receptors in breast cancer: biological and clinical role. J. Endocrinol. Invest. 19, 324–333. Pfeilschifter J, Scheidt–Nave C, Leidig–Brackner G, Woitge HW, Blum WF, Wuster C, Haack D & Ziegler R (1996): Relationship between circulating insulin–like growth factor components and sex hormones in a population–based sample of 50- to 80-year-old men and women. J. Clin. Endocrinol. Metab. 81, 2534–2540. Pisani P, Faggiano F, Krogh V, Palli D, Vineis P & Berrino F (1997): Relative validity and reproducibility of a food frequency dietary questionnaire for use in the Italian EPIC centres. Int. J. Epidemiol. 26(Suppl 1), S152–S160. Polonsky K, Frank B, Pugh W, Addis A, Karrison T, Meier P, Tager H & Rubenstein A (1986): The limitations to and valid use of C-peptide as a marker of the secretion of insulin. Diabetes 35, 379–386. Poretsky L, Cataldo NA, Rosenwaks Z & Giudice LC (1999): The insulin-related ovarian regulatory system in health and disease. Endocr. Rev. 20, 535–582. Pugeat M, Crave JC, Elmidani M, Nicolas MH, Garoscio Cholet M, Lejeune H, Dechaud H & Tourniaire J (1991): Pathophysiology of sex hormone binding globulin (SHBG). relation to insulin. J. Steroid. Biochem. Mol. Biol. 40, 841–849. Rasmussen MH, Ho KK, Kjems L & Hilsted J (1996): Serum growth hormone-binding protein in obesity: effect of a short-term, very low calorie diet and diet-induced weight loss. J. Clin. Endocrinol. Metab. 81, 1519–1524. Reinli K & Block G (1996): Phytoestrogen content of foodsFa compendium of literature values. Nutr. Cancer. 26, 123–148. Rinaldi S, Geay A, Dechaud H, Biessy C, Zeleniuch-Jacquotte A, Akhmedkhanov A, Snore RE, Riboli E, Toniolo P, Kaaks R (2002):
European Journal of Clinical Nutrition
Validity of free testosterone and free estradiol determinations in serum samples from postmenopausal women by theoretical calculations. Cancer Epidemol Biomarkers Prev. 11, 1065–1071. Salvini S, Parpinel M, Gnagnarella P, Maisonneuve P & Turrini A (1998): Banca dati di composizione degli alimenti per studi epidemiologici in Italia. Milano, Ita´lia. Singh A, Hamilton Fairley D, Koistinen R, Seppala M, James VH, Franks S & Reed MJ (1990): Effect of insulin–like growth factortype I (IGF-I) and insulin on secretion of sex hormone binding globulin and IGF-I binding protein (IBP-I) by human hepatoma cells. J. Endocrinol 124, R1–3. Slimani N, Ferrari P, Ocke M, Welch A, Boeing H, Liere M, Pala V, Amiano P, Lagiou A, Mattisson I, Stripp C, Engeset D, Charrondiere R, Buzzard M, Staveren W & Riboli E (2000): Standardization of the 24-hour diet recall calibration method used in the European prospective investigation into cancer and nutrition (EPIC): general concepts and preliminary results. Eur J. Clin. Nutr. 54, 900–917. Stattin P, Bylund A, Rinaldi S, Biessy C, Dechaud H, Stenman UH, Egevad L, Riboli E, Hallmans G & Kaaks R (2000): Plasma insulin– like growth factor I, insulin like growth factor–binding proteins, and prostate cancer risk: a prospective study. J. Natl. Cancer. Inst. 92, 1910–1917. Tannenbaum GS, Guyda HJ & Posner BI (1983): Insulin–like growth factors: a role in growth hormone negative feedback and body weight regulation via brain. Science 220, 77–79. The Endogenous Hormones and Breast Cancer Collaborative Group (2002): Endogenous sex hormones and breast cancer in postmenopausal women: reanalysis of nine prospective studies. J. Natl. Cancer. Inst. 94, 606–616. Thissen JP, Ketelslegers JM & Underwood LE (1994): Nutritional regulation of the insulin like growth factors. Endocr. Rev. 15, 80– 101. Thompson LU, Robb P Serraino M & Cheung F (1991): Mammalian lignan production from various foods. Nutr. Cancer. 16, 43–52. Tollet P, Enberg B & Mode A (1990): Growth hormone (GH) regulation of cytochrome P-450IIC12, insulin-like growth factorI (IGF-I), and GH receptor messenger RNA expression in primary rat hepatocytes: a hormonal interplay with insulin, IGF-I, and thyroid hormone. Mol. Endocrinol. 4, 1934–1942. Toniolo P, Bruning PF, Akhmedkhanov A, Bonfrer JM, Koenig KL, Lukanova A, Shore RE & Zeleniuch-Jacquotte A (2000): Serum insulin-like growth factor–I and breast cancer. Int. J. Cancer. 88, 828–832. Vermeulen A, Verdonck L, Kaufman JM (1999): A critical evaluation of simple methods for the estimation of free testosterone in serum. J. Clin. Endocrinol. Metab. 84, 3666–3672. Wakai K, Egami I, Kato K, Kawamura T, Tamakoshi A, Lin Y, Nakayama T, Wada M & Ohno Y (1999): Dietary intake and sources of isoflavones among Japanese. Nutr. Cancer 33, 139–145. Whelan EA, Sandier DP, Root JL, Smith KR & Weinberg CR (1994): Menstrual cycle patterns and risk of breast cancer. Am. J. Epidemiol. 140, 1081–1090. Yang G, Lu G, Jin F, Dai Q, Best R, Shu XO, Chen JR, Pan XY, Shrubsole M & Zheng W (2001): Population-based, case–control study of blood C-peptide level and breast cancer risk. Cancer Epidemiol. Biomarkers. Prev. 10, 1207–1211.
