“Associations of Serum Insulin-like Growth Factor (IGF-I) and IGFBP-3 Levels Biomarker-Calibrated Protein, Dairy, and Milk Intake in the Women's Health Initiative”

Abstract

It is well-established that protein-energy malnutrition decreases serum insulin-like growth factor (IGF-I) levels, and supplementation of 30 grams of whey protein daily increased serum IGF-1 levels by 8% after 2 years in a clinical trial⁽¹⁾. Cohort studies provide the opportunity to assess associations between dietary protein intake and the IGF-axis under more typical eating conditions. We studied the associations of circulating IGF-axis protein levels (ELISA, Diagnostic Systems Laboratories) with total biomarker-calibrated protein intake, as well as dairy and milk intake, among postmenopausal women enrolled in the Women's Health Initiative (n=747). Analyses were conducted using multivariate linear regression models that adjusted for age, BMI, race/ethnicity, education, biomarker-calibrated energy, alcohol, smoking, physical activity, and hormone therapy use. There was a positive association between milk intake and free-IGF-1. A 3 serving increase in milk intake per day (~30 grams of protein) was associated with an estimated average 18.6% higher increase in free IGF-1 (95% CI 0.9% to 39.3%). Total IGF-I and IGFBP-3, however, were not associated with milk consumption, nor were there associations between biomarker-calibrated protein intake, biomarker-calibrated energy, and free IGF-I, total IGF-I, or IGFBP-3. This study of postmenopausal women is consistent with clinical trial data suggesting a specific relationship between milk consumption and serum IGF-I levels; albeit, in our dataset, this association was only statistically significant for free, but not total, IGF-I nor IGFBP-3.

Introduction

Recent research by our group and others suggests higher protein intake may be associated with preservation of lean body mass⁽²⁾ and reduced frailty⁽³⁾ in postmenopausal women. Characterizing mechanisms through which higher protein intake may be related to successful aging phenotypes could inform dietary guidelines. The insulin-like growth factor (IGF)-axis constitutes an evolutionary conserved system involved in the regulation of cell growth, proliferation, and survival that affects nearly every organ system in the body. The axis consists of two phylogenetically conserved peptide ligands, IGF-I and IGF-II, with potent anabolic effects and six high-affinity binding proteins (IGFBP-1 to IGFBP-6)⁽⁴⁾.

IGF-I is the primary mediator of the growth effects of growth hormone, and is thought to be the major IGF affecting growth, health and disease following fetal development^{(5; 6)}. IGF-I shares extensive sequence homology and downstream signaling pathways with insulin, and has insulin-like effects on glucose and fat uptake in peripheral tissues. However, IGF-I exhibits stronger mitogenic and anti-apoptotic activity than insulin⁽⁷⁾. Previous research has suggested that circulating IGF-I levels may influence the risk of cancer⁽⁵⁾, diabetes⁽⁶⁾, and other conditions related to healthy aging⁽⁸⁾. Most of the IGF-I in the circulation is produced by the liver and is bound to IGFBPs, with IGFBP-3 binding 75% or more of all IGF-I in blood. Only approximately 1% of total serum IGF-I is unbound, and this free fraction may be the most biologically active component of total IGF-I⁽⁹⁾.

It is well known that diets deficient in energy and/or protein cause substantive reductions in serum IGF-I⁽¹⁰⁾; however, the role of protein and other dietary factors under conditions of weight maintenance are not as well-characterized. Two randomized, controlled trials of 20g¹¹ and 30g¹ of daily protein supplementation versus isocaloric placebo reported significant increases in serum IGF-I (51.5% (95% CI 18.6% to 84.4%) after 6 months and 8% (p=0.016) after 2 years, respectively)^{(1; 11; 12)}. A randomized trial comparing 3 daily servings of milk to control reported a 10% increase in IGF-I (P<0.001)¹². Observational studies have suggested that animal protein⁽¹³⁾, particularly from dairy sources⁽¹⁴⁾, is specifically associated with higher serum IGF-I; such studies can help to discern whether the putative increases in IGF-I are sustained with habitually higher compared to lower protein intake.

We are unaware of any published data that address the relation of protein intake with free IGF-I, and few dietary studies examined IGFBP-3 levels, though some data from two cross-sectional studies reported IGFBP-3 was not associated with protein consumption^{(13; 15)}. The current study also has the potential advantage of having data regarding biomarker-calibrated total protein intake, thought to be a better measure of dietary intake compared to self-report alone, from the Women's Health Initiative (WHI). Therefore, the current analysis examined the cross-sectional associations of total IGF-I, free IGF-I, and IGFBP-3, with biomarker-calibrated total protein intake, along with uncalibrated dairy and milk consumption among 747 women in the WHI – Observational Study (WHI-OS)^{(16; 17)}.

Methods

STUDY POPULATION

The WHI-OS is a prospective cohort study that enrolled 93,676 women ages 50-79 at 40 US clinical centers between 1993 and 1998, as described in detail elsewhere⁽¹⁸⁾. This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving participants were approved by the institutional review boards at each of the WHI clinical centers and at the Clinical Coordinating Center at the Fred Hutchinson Cancer Research Center (Seattle, Washington, USA). Written informed consent was obtained from all participants. Participants in the current investigation were controls from an ancillary nested case-cohort study designed to evaluate the relationship between the IGF axis and colorectal, breast, and endometrial cancer, as previously reported^{(19; 20; 21)}. Of the 816 randomly selected controls from the WHI-OS, the analytic sample consisted of 747 women who reported plausible energy intake on FFQ (i.e., excluding values <2510.4 or >20,920 kJ/day) and having complete information on covariates needed to calculate biomarker-calibrated protein (age, body mass index, smoking, race/ethnicity, education, income, physical activity).

IGF Measurement

Levels of total IGF-I, free IGF-I, and IGFBP-3 were determined in serum samples collected at baseline using commercially obtained ELISA (Diagnostic Systems Laboratories, Webster, TX, USA )^{(19; 20; 21)}. Inter-assay coefficients of variation were 8.2%, 11.2%, and 3.6% for total IGF-I, free IGF-I, and IGFBP-3, respectively.

Dietary Protein Assessment

All participants completed the WHI Food Frequency Questionnaire (FFQ) at WHI-OS baseline. The self-administered FFQ included 122 items for individual foods/food groups, 19 adjustment items, and summary questions⁽²²⁾. Protein intake was characterized as absolute (total intake in grams) and relative to energy intake (as a percentage of total kilocalorie intake (% energy)), and relative to body weight (as a ratio of 0.1 grams of daily protein intake per kilogram of body weight)⁽²³⁾. Milk consumption was captured from FFQ questions of frequency and portion for milk as beverage, on cereal and in coffee or tea. Dairy consumption was computed by summing milk consumption plus twenty items from the FFQ querying weekly servings of cheese, yogurt, and combination foods containing these items (i.e. quesadilla, cream soups).

Calibrated Energy and Protein Estimation

As previously described⁽¹⁷⁾, the WHI Nutritional Biomarkers Study (NBS) used objective biomarkers of total energy expenditure (equivalent to energy intake in weight stable persons) and protein intake to assess the measurement properties of the FFQ. Briefly, 544 women from twelve clinical centers of the Dietary Modification trial participated in a doubly-labeled water protocol to estimate total energy expenditure over a two-week period and a urinary nitrogen protocol to estimate protein consumption over a 24-hour period to be compared with concurrent self-reported dietary intake data. Calibrated energy and protein estimates were obtained by inserting baseline FFQ consumption estimates and other participant characteristics obtained at baseline from the WHI-OS into linear regression calibration equations¹⁹. For example:

$$\begin{gathered} \text{Log Calibrated percent energy from protein}=2.66+0.439\times \text{Log FFQ percent energy from protein}-0.004\times \\ \mathrm{BMI}-0.005\times \text{Age}-0.129\times \text{Current smoker}(1=\mathrm{yes},0=\mathrm{no}). \end{gathered}$$

Covariate Measurements

Demographic characteristics, medical history, and other health-related characteristics were obtained by self-report at WHI-OS baseline. BMI was computed using measured height and weight at baseline (kg / m²).

Statistical Analysis

IGF levels, protein, dairy, and milk intake were ln-transformed both to better approximate a normal distribution and to allow coefficients of regression models to be interpreted as percent difference. Spearman correlations were calculated between levels of total IGF-I, free IGF-I, and IGFBP-3. A series of linear regression models were examined to evaluate the influence of energy intake, BMI, and other potential confounders on observed associations between IGF-axis components and protein, dairy, and milk intake. Age, energy intake, and BMI were modeled as continuous variables⁽²⁴⁾. Exposures of interest were modeled both continuously and categorically. Standard errors for regression coefficients where total protein was the exposure were estimated using a bootstrapping procedure (1,000 replicates) to acknowledge the uncertainty in the regression calibration coefficients as well as that due to the sampling from the study population. To compare results with intervention studies, for the dairy and milk analyses, both the outcome and exposure were log-transformed and scaled to express associations in terms of percent difference in the outcome per three serving increase in the exposure. Logistic regression models were used to measure associations between calibrated protein intake categorized into quartiles and the odds of high serum total IGF-1, free IGF-1, and IGFBP-3 (defined as top 10%). Additional analyses were conducted modeling percent energy from protein sources (animal, vegetable, milk) as the exposure, adjusting for calcium intake, restricted to participants with no history of hormone therapy (HT) use and without prevalent diabetes (n=201). Analyses stratified by BMI (kg/m²) category (BMI <25, BMI 25-29.9, BMI ≥30), protein source (animal/vegetable), and by milk fat content (Whole/2%, 1%, skim) were also conducted, and the P for interaction term was estimated by multiplying the exposure variable by the interaction term.. All analyses were conducted using SAS Version 9.3 (SAS Institute, Cary, NC, USA)

Results

The majority of women were white (85%), had at least a high school education (79%), were current users of postmenopausal hormone therapy at baseline (49%), and were overweight (mean BMI 27.4 kg/m² (SD=5.7) (Table 1). Women who were current HT users had lower free IGF-I levels (P<0.05), as previously reported^{(19; 20; 21)}. Each of the IGF-axis proteins positively correlated with one another, with the strongest association between total IGF-I and IGFBP-3 (r=0.46, p<0.05) (Table 2).

Table 1.

Demographic and Health Characteristics by Free IGF-I tertile (n=747), WHI Observational Study

Characteristic	Overall, mean ± SD or n (column %)	Tertile 1, row (%)	Tertile 2, row (%)	Tertile 3, row (%)	P-value
Age, yr	62.9 ± 7.4	62.5 ± 7.1	63.0 ± 7.5	63.1 ± 7.6	0.34
Body Mass Index (kg/m2)	27.4 ± 5.7	27.2 ± 6.0	27.4 ± 5.7	27.7 ± 5.5	0.37
Ethnicity					0.08
White	638(85)	35	34	32
Black	58 (8)	21	29	50
Hispanic	25 (3)	28	44	28
Other	26 (4)	31	27	42
Education					0.36
≤High School	156 (21)	31	32	37
School after high school	273 (37)	38	33	29
College degree	91 (12)	29	31	41
Graduate school	227 (30)	31	36	33
Hormone Therapy Use					<0.001
Never	216 (29)	23	32	45
Past	158 (22)	30	34	37
Current	360 (49)	41	34	25
Smoking					0.56
Never	401 (54)	31	35	34
Past	298 (40)	35	31	34
Current	48 (6)	40	33	27
Alcohol intake					0.59
Never drinker	78 (11)	23	37	40
Past drinker	138 (19)	36	32	33
< 1 drink/wk	237 (32)	35	33	32
>= drinks/wk	292 (39)	33	33	34
Physical Activity, MET-hrs/week	13.9 ± 14.6	13.8 ± 15.2	12.7 ± 12.6	15.2 ± 15.9	0.27
Dietary Intake, Mean (SD)
Total energy, biomarker-calibrated, kJ/d)	8657 ± 858	8661 ± 100	8632 ± 812	8657 ± 837	0.96
Protein, biomarker-calibrated (% energy/d)	14.3 ± 1.4	14.2 ± 1.3	14.3 ± 1.4	14.3 ± 1.5	0.42
Protein, biomarker-calibrated (g/d)	76.3 ± 11.3	76.7 ± 11.3	76.2 ± 11.2	75.8 ± 11.4	0.40
Protein, biomarker-calibrated (g/kg body weight/d)	1.09 ± 0.20	1.11 ± 0.19	1.09 ± 0.21	1.07 ± 0.20	0.04
Dietary Calcium (diet alone) (mg/day)	847 ± 461	842 ± 445	841 ± 436	859 ± 501	0.68
Dietary Calcium (with supplements) (mg/day)	1264 ± 815	1303 ± 917	1234 ± 760	1255 ± 762	0.51
Dairy (servings/day)	1.9 ± 1.4	1.9 ± 1.3	1.9 ± 1.3	1.9 ± 1.5	0.53
Milk (servings/day)	0.83 ± 0.98	0.80 ± 0.92	0.79 ± 0.93	0.88 ± 1.08	0.35

Table 2.

Geometric Mean Levels and Spearman Correlations between IGF-axis proteins

IGF-axis protein, ng/mL	Geometric Mean (95% CI)	Free IGF	Total IGF	IGFBP-3
Free IGF-I	0.30 (0.28 to 0.32)	*
Total IGF-I	117.4 (114.0 to 120.8)	0.23	*
IGFBP-3	4041 (3984 to 4100)	0.14	0.46	*

Mean calibrated protein intake was 1.1 (SD=0.2) g/kg body weight/day, exceeding the current recommended dietary allowance for protein (0.8 g/kg body weight/day)⁽²³⁾. Expressed as a percentage of calibrated energy, median protein intake was 14.3% across a narrow range (interquartile range 13.3% to 15.2%). Dairy and milk intake also did not vary widely (i.e. median of 1.6 servings of dairy daily (interquartile range 0.9 to 2.6)).

Calibrated protein intake, expressed in absolute terms, relative to calibrated energy intake, or relative to body weight, were not associated with free IGF, IGF-I, or IGFBP-3 (Table 3). However, there was a positive association between free IGF-I and milk intake (Table 4). Specifically, a 3 serving increase in milk intake per day was associated with an estimated average 18.6% increase in free IGF-I (95% CI 0.9 to 39.3%) (Table 4).

Table 3.

Multivariate linear regression associations (β± 95% confidence intervals) between Free IGF-I, Total IGF-1, and IGFBP-3 with Calibrated Protein Intake

	N	Calibrated Protein, %energy	Calibrated Protein, g/d	Calibrated Protein, g/kg body weight/d*
Free IGF, %	732	0.02 (−0.004 to 0.04)	0.003 (−0.003 to 0.007)	−0.02 (−0.03 to 0.02)
Total IGF, %	759	−0.34 (−3.94 to 1.75)	−0.15 (−0.91 to 0.43)	1.9 (−2.7 to 3.0)
IGFBP-3, %	736	6.81 (−43.4 to 50.8)	3.11 (−10.3 to 14.4)	−7.8 (−67.6 to 33.7)

^*Scaled to 0.1 g/kg body weight/d; %energy=percentage of energy intake; g=gram; kg=kilogram; d=day

**adjusted for age, BMI, race/ethnicity, education, calibrated energy, alcohol, smoking, physical activity, hormone therapy use

^Primary exposure and outcome variables ln-transformed to represent expected % difference in IGF per % difference in protein intake

Table 4.

Percent difference in IGF per 3 serving increase in daily Dairy and Milk intake estimated from linear regression models (% Difference ± 95% Confidence Intervals)

	Total Dairy	Milk
Free IGF, %	11.0 (−1.3 to 25.3)	18.6 (0.9 to 39.3)
Total IGF, %	4.7 (−0.8 to 10.5)	4.4 (−3.0 to 12.3)
IGFBP-3, %	0.7 (−2.1 to 3.5)	−1.1 (−4.7 to 2.7)

**Bold indicates p<0.05.

Estimated from ln-transformed outcome and primary exposure variables in multivariate linear regression models adjusted for age, BMI, race/ethnicity, education, energy intake, alcohol, smoking, physical activity, and hormone therapy use. Estimates were scaled (2.58 and 4.3 ^β for dairy and milk, respectively) to reflect intervention amount of 3 servings (average dairy intake is 1.9 servings per day so 3 serving increase per day = 158% increase, and average milk intake is 0.8 servings per day so 3 serving increase = 375% increase)

IGFBP-3 was not associated with dairy or milk intake in continuous or categorical analyses. Results from logistic regression models between calibrated protein intake, dairy, and milk categorized into quartiles and the odds of high serum total IGF-1, free IGF-1, and IGFBP-3 (defined as top 10%) were consistent with linear regression models (data not shown). The results of analyses that modeled the exposure as a categorical or continuous variable were similar, and there was no statistically significant or discernible difference in estimates restricted to (i) never hormone therapy users and (ii) women without prevalent diabetes. Nor were the results meaningfully affected by stratification by BMI (kg/m²) category (BMI<25, 25-29.9, ≥30), protein source (percent energy from animal vegetable, and milk protein), or milk fat content (Whole/2%, 1%, skim) (all P>0.05 for interaction, data not shown). Analyses including uncalibrated instead of calibrated energy and protein intake were also similar (data not shown). Lastly, we assessed whether the relation of milk consumption with free IGF-I might be explained by calcium intake. There was no association between Free IGF-I and calcium intake (β=0, 95% CI (−0.002 to 0.0003) for calcium from food alone as well as including supplements β=0, 95% CI (−0.001 to 0).

Discussion

This study supports evidence from clinical trials and observational studies of a specific relationship between milk consumption and IGF-I; albeit, in our study this relationship was only statistically significant for free IGF-I. To our knowledge, this is the first study of the relationship between protein intake and milk consumption and free IGF-I. Our data, however, did not support reports from other epidemiological studies suggesting a positive relationship between higher overall total protein intake and higher total IGF-I levels^{(1; 11; 12; 25)}. Biomarker-calibration of protein intake did not appreciably affect results.

Prior observational studies ^{(13; 26; 27; 28)} also support a positive association between protein intake, particularly from dairy sources, and total IGF-I levels^{(1; 11; 12; 25)} . Data from the European Prospective Investigation into Cancer and Nutrition (EPIC, N=4,731 men and women), reported a 2.5% and 2.4% increase in IGF-I per 3% and 2% increase in energy from total protein and dairy, respectively⁽²⁶⁾. Given each serving of dairy equates to 601 kJ⁽²⁹⁾ and mean energy intake was 9,876 kJ in EPIC, 2% energy equates to 198 kJ, or approximately 1/3 dairy serving. Assuming a linear dose response, this would equate to a 22.5% increase in IGF-I with a 3 serving dairy increase per day, which is consistent with the strength of the association observed in the current study (17.3%) and in clinical trials (range 8%-52%). Moreover, of the protein sources examined (plant, meat, dairy, fish, eggs) in EPIC, only dairy was significantly associated with IGF-I levels. The Nurses’ Health Study, examining a comprehensive list of dietary correlates, reported 9.3% and 7.7% higher IGF-I levels comparing highest versus lowest quintiles of total and dairy protein, respectively⁽¹³⁾. Similar to our study, prior studies examining correlates of IGFBP-3 levels^{(15; 26; 27)} reported no significant associations with either total dietary protein intake or consumption of dairy-related protein.

In prior clinical trials, an additional 20-30 grams of protein, often whey protein, or 2-3 milk servings, daily resulted in increases in IGF-I of 8-10% that were sustained for up to two years^{(1; 11; 12; 25)}. This is well above typical average daily milk consumption of 0.7 servings for the US population aged 2 and older (NHANES 2003-2008 data)⁽²⁹⁾. In our study, mean daily milk consumption was 0.8 (SD=1.0) servings, less than 25% of women reported consuming 3 milk servings/day, and the upper limit of intake was 7.2 servings/day. Thus, the magnitude of the effect size observed in the current study is in keeping with that found in clinical trials, and for the first time suggests that dairy intake within the upper range of the typical American diet increases free IGF-I levels. Furthermore, a meta-analysis of eight randomized trials reported a weighted mean difference of circulating IGF-I levels between milk intervention and control groups of 13.8 ng/mL (95% CI=6.1, 21.5) , and postulated components of milk that may increase IGF-I concentrations include its nutrient composition of high protein, fat, and calcium, and/or contaminants or hormones in milk such as estrogens and insulin-like growth factor ⁽³⁰⁾.

Collectively, the current study and prior reports are supportive of a particular association between the levels of IGF-I and milk consumption, which may involve components of milk other than protein. That is, the consistent positive association between IGF-I levels and milk consumption might be explained by a single non-protein nutrient that, nonetheless, may be correlated with total protein or milk protein (since these two factors were variably significant in some prior studies but not in the current investigation or several others), or possibly a combination of milk components that include protein.. For example, the Singapore Chinese Health Study reported a positive association between calcium from food and supplements with IGF-I and IGFBP-3⁽³¹⁾. However, calcium intake, both from foods alone and in combination with calcium supplement use, was not significantly associated with free IGF-1, total IGF-1, or IGFBP-3 in this study.

Limitations of our study should be considered in interpreting these data. Diet was characterized using a FFQ, which is limited by both differential and non-differential measurement error⁽³²⁾. For example, some items that could contribute to protein intake, such as protein supplements, were not queried. .However, using biomarkers of energy and protein intake, we were able to mitigate some important sources of measurement error associated with energy and total protein intake⁽¹⁷⁾. Similar to representative samples of the US population⁽³³⁾, the range of protein, and particularly dairy and milk intake, was not wide in this population. Since the dairy, cheese, and yogurt food group variables are a summary of several FFQ items having different weights and protein content, we could not compute with certainty the protein for dairy, cheese, and yogurt groups. However, our analyses of IGF-axis levels and their relation with protein consumption was null even when examined separately by source (animal, vegetable, or milk protein), suggesting protein consumption alone is not responsible for the observed positive association between milk consumption and free IGF-I.

In conclusion, these data suggest a 3 serving increase in milk consumption per day is associated with an 18.6% (95% CI= 0.9% to 39.3%) increase in free IGF-I levels. These data provide new insight into the factors that influence levels free IGF-I, presumably the most bioactive component of IGF-I; a hormone peptide that is the primary mediator of the effects of growth hormone, and which has been associated with several major health conditions, including diabetes, cardiovascular disease, and cognitive function in the elderly.