help button home button Endocrine Society JCEM
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Journal of Clinical Endocrinology & Metabolism , doi:10.1210/jc.2005-1722
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bray, I.
Right arrow Articles by Martin, R. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bray, I.
Right arrow Articles by Martin, R. M.
Related Collections
Right arrow Neuroendocrinology and Pituitary
Right arrow Pediatric Endocrinology
Right arrow Metabolism
The Journal of Clinical Endocrinology & Metabolism Vol. 91, No. 4 1382-1389
Copyright © 2006 by The Endocrine Society

Associations of Childhood and Adulthood Height and the Components of Height with Insulin-Like Growth Factor Levels in Adulthood: A 65-Year Follow-Up of the Boyd Orr Cohort

Isabelle Bray, David Gunnell, Jeff M. P. Holly, Nicos Middleton, George Davey Smith and Richard M. Martin

Department of Social Medicine (I.B., D.G., N.M., G.D.S., R.M.M.), University of Bristol, Bristol BS8 2PR, United Kingdom; and Clinical Sciences at North Bristol (J.M.P.H.), Southmead Hospital, Bristol BS10 5NB, United Kingdom

Address all correspondence and requests for reprints to: Dr. Isabelle Bray, Defence Analytical Services Agency Health, Spur 7, Beckford Block, Ensleigh, Bath BA1 5AB, United Kingdom. E-mail: issy.bray{at}bristol.ac.uk.


   Abstract
Top
 Abstract
Introduction
 Patients and Methods
 Results
 Discussion
 References
 
Context: Taller individuals with longer legs have a higher risk of cancer but a lower risk of coronary heart disease.

Objective: We investigated whether childhood height and its components are associated with the IGF system in adulthood.

Design and Participants: We analyzed data from 429 participants of the Boyd Orr cohort, for whom height measured in childhood (mean age, 7.4 yr) in 1937–1939 could be related to levels of IGF-I, IGF-II, IGF binding protein (IGFBP)-2, and IGFBP-3 in adulthood (mean age, 71.1 yr). In 385 participants, measured height in adulthood could be related to IGF levels.

Results: In fully adjusted models (controlling for age, sex, socioeconomic factors, lifestyle, and body mass index), childhood height and its components were not associated with adult circulating IGF-I, IGF-II, or IGFBP-2 levels. IGFBP-3 was 85.5 ng/ml higher (95% confidence interval, –11.6 to 182.5; P = 0.08) per SD increase in childhood trunk length and 83.6 ng/ml lower (95% confidence interval, –10.3 to 177.5; P = 0.08) per SD increase in childhood leg/trunk ratio. Height in adulthood was not associated with IGF-I, IGF-II, or IGFBP-3 and was inversely associated with IGFBP-2 (P = 0.05) after additionally controlling for childhood height.

Conclusion: There was no evidence that associations of childhood height with cancer and coronary heart disease risk are mediated by IGF-I in adulthood. The anthropometric associations with IGFBP-2 and IGFBP-3 could be chance findings but warrant additional investigation. IGF levels in childhood may be more important determinants of long-term disease risk than adult levels.


   Introduction
Top
 Abstract
 Introduction
Patients and Methods
 Results
 Discussion
 References
 
OBSERVATIONAL EVIDENCE POINTS toward positive associations of height with risk of some cancers (e.g. prostate, breast, and colon) (1) and inverse associations with insulin resistance and coronary heart disease (CHD) (2, 3). Although the biology underlying these observations is unclear, indirect evidence suggests that IGFs may underlie associations between height and these disease outcomes (4, 5, 6). The IGF system comprises IGF-I and IGF-II, which are mainly synthesized in the liver in response to GH, and six binding proteins, IGF binding protein (IGFBP)-1 to IGFBP-6, which regulate the biological activity of circulating IGFs. The GH-IGF axis plays a key role in controlling prepubertal growth (7). IGF-I is associated positively with childhood height (5, 8) and cancer risk (9, 10, 11) and inversely with insulin resistance (12) and CHD risk (13). IGFBP-3 controls IGF-I bioavailability and is associated inversely with some cancers (14) and positively with CHD (13). IGF-II has been positively associated with cancer (15) and atherosclerosis (16), and low levels of circulating IGFBP-2 may indicate insulin resistance, but their role in disease etiology is not well understood (17, 18, 19). IGFBP-1, IGFBP-4, IGFBP-5, and IGFBP-6 are difficult and expensive to measure, and these binding proteins are not studied in the current report.

Leg length is associated with socioeconomic conditions, nutrition, and energy intake in prepubertal children (20, 21, 22), whereas trunk length may be a marker for factors influencing pubertal growth (22). Because a greater proportion of prepubertal growth is attributable to an increase in leg rather than trunk length, leg length may act as a better indicator of exposures influencing prepubertal growth than total height (23, 24, 25). For this reason, it has been suggested that associations of stature with cancer and CHD may be stronger for leg length than for overall height (1, 3, 26, 27). Some (6, 28, 29), but not all (4, 30, 31, 32), cross-sectional studies indicate positive associations of height with IGF levels in adulthood, but there is no convincing evidence that leg length is more strongly associated than trunk length with IGF (4). In childhood, height (8, 33, 34) and both leg and trunk growth (Rogers, I., C. Metcalfe, D. Gunnell, P. Emmett, D. Dunger, J. Holly, and ALSPAC Study Team, submitted for publication) are positively associated with concurrently measured IGF-I. Few studies, however, have examined the association of childhood anthropometry, other than birth weight, with adult IGF levels (5, 35), and none have related childhood leg or trunk length with IGF levels in adulthood. Most reports are restricted to investigating IGF-I and IGFBP-3 associations, although recent work suggests that IGFBP-2 may contribute to biological mechanisms underlying height-disease associations (4). A greater understanding of relationships between childhood stature and the IGF system in adulthood may help elucidate mechanisms linking nutrition and growth in early life with increased cancer and reduced CHD risks in adulthood. The Boyd Orr cohort has uniquely detailed records of measured childhood height, leg length, trunk length, and important dietary and socioeconomic confounding variables, among participants followed up for 65 yr. Adulthood follow-up data include measured height, leg length, trunk length, and circulating levels of IGF-I, IGF-II, IGFBP-2, and IGFBP-3. We report on relationships of height, leg length, and trunk length measured in childhood and adulthood with circulating IGF and IGFBP levels in adulthood in a cohort in which childhood height and leg length have been associated positively with cancer and inversely with CHD (36, 37). There have been too few follow-up events since blood sampling in 2002–2003 to directly relate IGFs with cancer and CHD risk.


   Patients and Methods
Top
 Abstract
 Introduction
 Patients and Methods
Results
 Discussion
 References
 
The study is an historical cohort based on the Carnegie (Boyd Orr) Survey of Diet and Health in Pre-War Britain, 1937–1939 (38). A total of 4999 children from 1343 families were surveyed at 16 centers in England and Scotland. Detailed physical examination was performed on 3762 of the children (75%) from 14 centers. Of the original 4999 participants, 4397 (88%) have been traced and flagged using the National Health Service Central Register at Southport and its equivalent in Edinburgh.

Measures in childhood

Measurements of height and leg and trunk length taken on 2997 children aged 2 to 14 yr 9 months are available. Childhood standing height was measured to the nearest millimeter with a portable measuring stand (38). Childhood leg length was measured with a steel tape measure and recorded as the distance from the ground to the summit of the iliac crest (38). Trunk length was calculated by subtracting leg length from overall height.

Adult follow-up

Between 2002 and 2003, a total of 1295 surviving participants were invited to participate in a follow-up study, as described previously (39). Blood samples and measured adult height were obtained from 728 of 1295 (56%) potentially eligible subjects, either at a research clinic (n = 405) or by the general practitioner (GP) who sent the samples for analysis by mail in approved post office packaging (n = 323) (Fig. 1Go). GP- or clinic-measured adult heights in study members with complete data for multivariable analysis are available for 682 (53%) subjects. However, GPs measured height only, without any specific instruction on measurement techniques. In contrast, height and sitting height were measured in the research clinic by one of two trained researchers with a Harpenden stadiometer using a standardized protocol. Clinic-measured adult height in study members with complete data for multivariable analysis are available for 385 (30%) of the subjects. We based our analysis relating adult height with IGFs, therefore, on the participants with clinic measures to reduce measurement error. In a sensitivity analysis, associations of IGF with adult height (measured by either the GP or in the research clinic) were repeated using all 682 respondents with complete data for multivariable analyses. The number of participants for whom childhood anthropometry was available was 429 (33%), and this is the subset of the cohort on which the main analyses relating childhood height with IGF levels in adulthood are undertaken.


Figure 1
View larger version (22K):
[in this window]
[in a new window]
  		FIG. 1. Description of study numbers: target population, respondents with blood samples, numbers with measures of height and their components in childhood and adulthood, and numbers with complete data for multivariable analysis that were included in the final analyses. aOne subject with a blood sample did not have enough serum for IGF measures.

 
Exposure variables in childhood and adulthood

As in previous analyses of this cohort relating childhood stature with cancer and CHD (36), internally derived age- and sex-standardized Z scores for height (n = 429), leg length (n = 418), trunk length (n = 418), and the leg/trunk ratio (n = 418) measured at one point in time in the children (aged between 2 and 14 yr 9 months) were the childhood exposure variables of interest. Internally age- and sex-standardized Z scores were computed using cubic polynomial regression for boys and girls separately within 6-month age and sex bands (21). SDs were modeled within 6-month age bands (21). Associations of the IGF system with anthropometry in adulthood were based on Z scores internally standardized for sex (n = 385). The magnitude of anthropometry-IGF associations in childhood are thus directly comparable with those in adulthood.

The leg/trunk ratio is an indicator of leg length in relation to total stature. An increased leg/trunk ratio suggests relatively long legs in relation to overall height, and associations of this ratio with IGFs would establish the importance of body proportion (leg relative to trunk length) in relation to the IGF system. During childhood, the leg/trunk ratio provides a measure of prepubertal growth that occurs more in the legs than the trunk (21, 25). In adults, it may also be a marker of delayed puberty, which leads to a prolonged period of leg growth.

IGF measurements

The IGF outcomes were IGF-I, IGF-II, IGFBP-2, IGFBP-3, and the molar ratio IGF-I/IGFBP-3 [a measure of free, biologically active IGF-I (29)]. Among the 727 subjects with IGF measurements, 405 blood samples were taken at research clinics, and 91% of these were obtained fasted for 6 h or more; bloods taken by the GPs were nonfasted (n = 322). The clinic samples were spun and frozen to –20 C within 1 h and then transferred within 3 wk to a –80 C freezer. Seventy-nine percent of bloods sent by mail arrived the next day, and only 1% took 5 d or more to arrive. Adjusting for age and sex, there is no evidence of any differences in mean levels of IGF-I, IGF-II, IGFBP-2, or IGFBP-3 between those that arrived the next day and those that took 2 or more days to arrive. The median (interquartile range) number of days that serum samples were stored before analysis (n = 727), in ascending order, were 355 (interquartile range, 291–442) for IGFBP-3, 409 (348–518) for IGFBP-2, 545 (429–655) for IGF-I, and 723 (562–833) for IGF-II. Adjusting for age and sex, there is no evidence of a linear trend in mean levels of IGF-I, IGFBP-2, or IGFBP-3 across quintiles of storage time, but IGF-II levels were higher in those samples that had been stored the longest. The median IGF-II levels in the shortest and longest quintiles of storage time were 481 and 669 ng/ml, respectively. This trend was not explained by age, sex, or any one of three conditions of sample measurement (whether fasted/not fasted, binary variable; laboratory technician undertaking the assay, indicator variables; or clinic vs. posted blood, binary variable). Hence, we repeated all IGF-II analyses to determine whether controlling for length of storage time made any difference to the results.

Serum IGF-I, IGF-II, and IGFBP-3 levels were measured using in-house double-antibody RIAs as described previously (40). Total levels of IGFBP-2 were measured by one-step sandwich ELISAs (DSL-10-7100; Diagnostic Systems Laboratories, Webster, TX). Results for IGF-I, IGF-II, and IGFBP-3 concentrations were based on the average of three measures and for IGFBP-2 on the average of two measures. The average coefficients of variation for intraassay variability for IGF-I, IGF-II, IGFBP-3, and IGFBP-2 were 6.7, 10, 3.9, and 5% and for interassay variation were 9.7, 14, 8.1, and 7.1%. Based on the molecular weight of IGF-I (7500) and IGFBP-3 (40,000, mean of glycosylated variants), we calculated the molar ratio of IGF-I/IGFBP-3 by multiplying the ratio by 5.33 (40,000:7500).

Statistical analyses

Relationships of childhood or adulthood height and its components with the IGF system in adulthood were assessed using multiple linear regression. The regression coefficients show the change in IGF and IGFBP levels per SD increase in stature. IGFBP-2 values were highly positively skewed and were loge transformed; thus, the percentage change in IGFBP-2 per SD increase in stature is given.

Associations were adjusted for age (continuous variable), sex, clinic-obtained vs. posted blood sample, per capita household food expenditure in childhood [four categories: <£0.25 (equivalent 2002 monetary value, based on average earnings (41), <£36), £0.25 to £0.34 (£36-£49); £0.35 to £0.44 (£50-£65); and >£0.44 (>£65)], social class of head of household when a child in 1937–1939 (four categories: I/II, professional and managerial; III, skilled; IV/V, partly skilled and unskilled; other, unemployed and unclassifiable), social class in adulthood [three categories: I/II, III, IV/V/other (armed forces, unemployed, unclassifiable)], and lifestyle factors in adulthood [smoking (four categories based on pack-years), alcohol (three categories indicating frequency of consumption over the past 12 months: at least weekly, occasionally, and never), exercise (four-level score based on frequency and type of activity validated in the British Regional Heart Study (42), and measured adult body mass index (BMI) (continuous variable)]. We calculated robust SE values to account for lack of independence between observations within families (43). Additional models investigated the impact on observed associations of controlling for key components of diet previously associated with IGFs: milk intake in childhood [daily per capita total milk and milk products consumption (in grams)], dairy intake in adulthood [daily all dairy intake (in grams)], and total energy intake in adulthood [from all foods, including milk and alcohol (in kilocalories)]. We also investigated whether controlling for IGFBP-3 influenced associations of height and its components with IGF-I and IGF-II and whether the inclusion of homeostasis model assessment (HOMA) measures of insulin resistance (44) influenced anthropometry-IGF associations (45).

We performed likelihood ratio tests for interactions with sex and age [defined a priori as in previous analyses (21, 36) as <8 or ≥8 yr] at childhood measurement to investigate whether associations differed in those children who were prepubertal (<8 yr) compared with those who may have entered puberty (≥8 yr). We also controlled for childhood stature in models relating adult height with the IGF system to assess whether growth trajectory since childhood, independent of childhood height, was an important predictor of IGF in adulthood. All analyses were conducted using Stata 8 (Stata, College Station, TX).


   Results
Top
 Abstract
 Introduction
 Patients and Methods
 Results
Discussion
 References
 
Mean age at childhood examination of the study participants was 7.4 yr, and mean age at time of blood sample was 71.1 yr (45.6% male). Males were taller than females, but the leg/trunk ratio was similar (Table 1Go). Mean IGF-I levels and the molar ratio IGF-I/IGFBP-3 were higher in men than women, but IGF-II and IGFBP-3 levels were lower in men than women.


View this table:
[in this window]
[in a new window]
  		TABLE 1. Characteristics of study participants

 
Representativeness of the study population

There was no evidence that the 728 people who provided blood samples in the follow-up study differed from those eligible participants who did not (n = 567) in terms of sex, birth weight, childhood BMI, adult height, or adult BMI. However, those who provided blood samples were more likely to have been younger when originally surveyed (6.5 vs. 7.6 yr), to have fathers in social class I–III (33 vs. 24%), to have been taller in childhood (Z scores: 0.15 vs. –0.03 SDs), to be in social class I–II in adulthood (33 vs. 17%), to consume alcohol at least weekly (52 vs. 43%), and to have never smoked (40 vs. 30%).

Height and components of height in childhood

Table 2Go gives the regression coefficients for models assessing relationships of height and its components in childhood with IGF and IGFBP levels in adulthood, with three sequential levels of adjustment for potential confounding factors as shown. No associations were seen with IGF-I, IGF-II, or the molar ratio IGF-I/IGFBP-3. There was weak evidence that childhood trunk length was positively associated (P = 0.08) and that childhood leg/trunk ratio was inversely associated (P = 0.08) with adulthood levels of IGFBP-3. The association of IGFBP-3 with trunk length was similar after leg length was included in the fully adjusted model [coefficient, 91.7 ng/ml; 95% confidence interval (CI), –5.06 to 188.51; P = 0.06]. Neither HOMA insulin resistance nor dietary factors (milk intake in childhood and dairy and total energy intake in adulthood) confounded the above associations (data not shown). Associations of IGFBP-3 with childhood trunk length and leg/trunk ratio, however, were not independent of circulating IGF-I levels (fully adjusted coefficients additionally controlling for IGF-I were, for trunk length, 38.3 ng/ml; 95% CI, –31.8 to 108.4; P = 0.28; and for leg/trunk ratio, –42.9 ng/ml; –114.5 to 28.8; P = 0.24). When we additionally adjusted height-IGF-I and height-IGF-II associations for IGFBP-3, our conclusions did not change (data not shown). IGFBP-2 was weakly positively associated with height (P = 0.07) and leg length (P = 0.06) in simple models, but associations disappeared in models controlling for socioeconomic and lifestyle factors. Anthropometry-IGFBP-2 associations were similar when additionally adjusted for IGF-I. Height-IGF associations did not differ by sex (all P values for interaction >0.4, except for IGF-II, P = 0.08) or age (< 8 vs. ≥ 8 yr) at time of childhood examination (all P values for interaction > 0.35).


View this table:
[in this window]
[in a new window]
  		TABLE 2. Associations of measured height and its components in childhood with IGFs and IGFBPs measured in adulthood

 
Height and components of height in adulthood

In fully adjusted models, neither measured adulthood height nor its components were associated with circulating IGF-I, IGF-II, IGFBP-2, or IGFBP-3 (Table 3Go). Analyses based on all 682 respondents with adult height, measured by either the GP or in the research clinic, were similar to those based only on those with clinic height. Among 243 participants with both childhood and clinic-measured adulthood heights, there was little evidence that change in height Z score between childhood and adulthood (to assess whether growth trajectory since childhood influences the IGF system in adulthood) was associated with circulating IGF-I, IGF-II, or IGFBP-3 (all correlation coefficients < 0.1). IGFBP-2 levels, however, were lower among those whose height Z score was greater in adulthood relative to the childhood measure (correlation coefficient between change in height Z score and (ln)IGFBP-2, –0.14; P = 0.02). Controlling for childhood stature in the fully adjusted model relating adult height with (ln)IGFBP-2, the regression coefficient implied a change of –7.2% per SD increase in adult height (95% CI, –14.0 to 0.0%; P = 0.05), and the association persisted after controlling for HOMA insulin resistance. There was no evidence that adult height was associated with IGF-I, IGF-II, or IGFBP-3 in fully adjusted models additionally controlling for childhood stature (all P values > 0.2).


View this table:
[in this window]
[in a new window]
  		TABLE 3. Associations of measured height and its components in adulthood with current IGFs and IGFBPs

 
Associations for both childhood and adulthood anthropometry were little altered when controlled for –80 C freezer storage time.


   Discussion
Top
 Abstract
 Introduction
 Patients and Methods
 Results
 Discussion
References
 
Contrary to our predictions, we found little evidence that either height or leg length, measured at one point in time in childhood or adulthood, was associated with circulating levels of IGF-I, IGF-II, or IGFBP-2 in adulthood. Nevertheless, there was weak evidence that IGFBP-3 was higher in participants who, as children, had longer trunks and a lower leg/trunk length ratio, an indicator of relatively long trunk length in relation to total stature. Height in adulthood was inversely associated with IGFBP-2 but only after additionally controlling for childhood height.

Implications

Previous analyses of the Boyd Orr cohort have documented that taller children, specifically those with longer legs, were at increased risk of dying from cancer and at reduced risk of CHD death (36, 37). Epidemiological studies also indicate that exposure to increased IGF-I levels is associated with greater cancer (9, 10, 11) and lower insulin resistance (12) and CHD (13) risk. We had speculated, therefore, that childhood height and leg length associations with chronic disease may be mediated by IGF-I, because circulating levels of this growth factor in early life influence childhood growth (1, 5, 8, 34, 35, 46). Our study indicates that associations of childhood height and leg length with later disease risk are not reflected strongly by associations of childhood anthropometry with adult levels of IGF-I. Two studies that have examined associations of childhood stature with adult IGF levels concur with our results (5, 35), although they did not specifically investigate the leg and trunk length components of height. In the Barry Caerphilly Growth study (5), no association was found between height measured at age 5 yr and IGF-I, IGFBP-3, and their molar ratio measured among men and women in their 20s, although a positive association between current IGF-I levels and adult height was found. Likewise, in a cohort of Finnish men and women in their 60s and 70s, height at 7 yr was not associated with IGF-I measured in old age (35). Others have found, however, that patterns of childhood growth in relation to birth weight (catch-up growth or rate of height gain) are associated with IGF-I levels in childhood (8, 47, 48) and early adulthood (5). Downward centile crossing at any time in childhood was associated with lower levels of IGF-I in adulthood, and the highest levels of IGF-I were observed in young adults who were tall throughout early life (5). We could not investigate growth in childhood because childhood stature was measured only once in most members of our cohort, but we found little evidence that change in height between childhood and adulthood was associated with IGF-I, IGF-II, or IGFBP-3 in adulthood. The inverse association of height in adulthood with IGFBP-2 after controlling for childhood height could indicate that upward growth trajectory between the childhood measure and adulthood predicts lower adult IGFBP-2 levels, because the main determinant of adult height after controlling for the earlier childhood measure is likely to be a composite of the duration and rate of growth during puberty (5).

Mixed findings emerge from studies examining the association of adult stature with circulating growth factor levels. Some (6, 28, 29, 49), but not all (4, 30, 31), previous studies show positive associations of adult height with IGF-I and/or IGFBP-3, but few report on associations with the components of height. In keeping with the findings reported here and by others (30, 31), one study examining cross-sectional associations of both adult height and leg length with IGF-I found no evidence of any relationships, although it did show a weak inverse association between leg/trunk length ratio and IGFBP-3 (P = 0.13) and a positive association between leg/trunk ratio and the molar ratio IGF-I/IGFBP-3 (P = 0.06) (4). These findings are in line with our report, because they could indicate that relatively long trunk length in relation to total stature is associated with higher IGFBP-3 levels. Circulating levels of IGFBP-3 have been positively associated with height in childhood (29) and adulthood (50) and with prepubertal growth in height independent of serum IGF-I (Rogers, I., C. Metcalfe, D. Gunnell, P. Emmett, D. Dunger, J. Holly, and ALSPAC Study Team, submitted for publication). Our study, along with our previous investigation (4), provides limited evidence that differences in childhood body proportions (leg/trunk ratios) may act as indicators of levels of IGFBP-3 in adulthood and, speculatively, may thus be one mechanism linking childhood stature to adult chronic disease risk.

Although our findings do not support the hypothesis that childhood height, measured at one point in time, is a marker for adult IGF levels, it remains possible that childhood IGF levels, for which childhood stature is a marker (8, 47, 51), influence adult chronic disease, either by a direct effect on early pathophysiological processes underlying the later development of chronic disease (52) or via a long-term programming effect on the function of tissues and organs (53). Investigating associations of chronic disease outcomes with IGF genetic polymorphisms that alter lifelong exposure to biologically active IGF and IGFBP levels could further our understanding of this issue (54). The duration and rate of growth in childhood, which influence final adult height, may be a better marker of IGF levels than height in childhood measured at one point in time (5, 46). We found no evidence, however, that height measured in adulthood was associated with circulating levels of IGF-I and IGFBP-3.

Strengths and limitations

A major strength of our study over previous cross-sectional analyses in adulthood (4) is that associations of childhood stature with IGF levels in adulthood will not be influenced by shrinkage, particularly in trunk length, accompanying age-related osteoporotic vertebral collapse (55). Shrinkage attributable to age and comorbidity does not affect stature measured in childhood. Our analysis is based on height, leg length, and trunk length measured in both childhood and 65 yr later in adulthood in a cohort in which childhood height-disease associations have been documented previously. Furthermore, we have detailed information on a range of possible confounding factors, again recorded in both childhood and adulthood, thereby decreasing the likelihood of residual confounding as an explanation for the observed associations. There are four main limitations to our analysis. First, our sample size is relatively small, limiting our power to detect relatively small but possibly biologically important associations. Second, associations of childhood height and its components with the IGF system may have been obscured by either the known age-induced decline in the activity of the GH-IGF axis (30), by coexisting morbidity in the study population, or because the study population was based on a selective sample of the survivors in the cohort. Third, there was an apparent positive association between storage time and levels of IGF-II, but associations were little altered when controlled for storage time. Finally, we tested multiple hypotheses, so observed associations could have arisen by chance. No formal statistical approaches to account for multiple hypothesis testing were used, but we have quoted exact rather than threshold P values.

Conclusions

As far as we are aware, this is the only study to date that relates measured childhood height, leg length, and trunk length to levels of IGF-I, IGF-II, IGFBP-2, and IGFBP-3 in old age. No studies to date have examined associations of the IGF system in childhood with adult cancer or CHD; studies have related IGF levels measured in adulthood with these outcomes. Our data suggest that neither a child’s stature measured at one point in time and compared with his or her peers of the same age nor final height (a marker for the total duration and rate of growth) is strongly associated with the IGF system in adulthood. Levels of IGF in childhood may be more important determinants of long-term disease risk than adult levels. Additional studies should seek to elucidate the causal relationships between GHs, rate of childhood growth, and tumor generation, the origins of which may begin in early life (7).


   Acknowledgments
 
We are very grateful to the cohort members who participated so willingly in the follow-up study. Prof. John Pemberton is thanked for information concerning the conduct of the original survey. We also acknowledge all of the research workers in the original survey in 1937–1939. We thank Prof. Peter Morgan (Director of Rowett Research Institute, Aberdeen, Scotland, UK) for the use of the archive and, in particular, Walter Duncan (honorary archivist to Rowett Research Institute). Susie Potts is thanked for all of her hard work in providing secretarial and administrative support to this study. Simone Watson was the research nurse. We thank all of the general practitioners who participated in the study and the following for providing clinic space in 2002: Dr. Nigel Williams (Clarkson Surgery, Wisbech, Cambridgeshire, UK); Prof. Andrew Morris (Diabetes Centre, Ninewells Hospital, Dundee, Scotland, UK); Dr. Ali Jawad (Royal London, Mile End, London, UK); Sylvia Hey (Human Nutrition Unit, The Rowett Research Institute); Prof. Frank Sullivan (Mill Practice, Dundee, Scotland, UK); and Prof. Philip Hannaford (Aberdeen University, Aberdeen, Scotland, UK). We acknowledge advice from Roger Harbord on Stata programming. Jane Carter and Paul Savage performed the IGF analyses.


   Footnotes
 
This study was funded by The World Cancer Research Fund Grant 2001/31 and UK Survivors. R.M.M. was funded by a Wellcome Trust Research Training Fellowship GR063779FR in clinical epidemiology.

None of the authors have anything to declare.

First Published Online January 24, 2006

Abbreviations: BMI, Body mass index; CHD, coronary heart disease; CI, confidence interval; GP, general practitioner; HOMA, homeostasis model assessment; IGFBP, IGF binding protein.

Received August 1, 2005.

Accepted January 17, 2006.


   References
Top
 Abstract
 Introduction
 Patients and Methods
 Results
 Discussion
 References
 

  1. Gunnell D, Okasha M, Smith GD, Oliver SE, Sandhu J, Holly JM 2001 Height, leg length, and cancer risk: a systematic review. Epidemiol Rev 23:296–325
  2. Williams SR, Jones E, Bell W, Davies B, Bourne MW 1997 Body habitus and coronary heart disease in men. A review with reference to methods of body habitus assessment. Eur Heart J 18:376–393[Free Full Text]
  3. Davey Smith G, Greenwood R, Gunnell DJ, Sweetnam P, Yarnell J, Elwood P 2001 Leg length, insulin resistance, and coronary heart disease risk: the Caerphilly Study. J Epidemiol Community Health 55:867–872[Abstract/Free Full Text]
  4. Gunnell D, Oliver SE, Donovan JL, Peters TJ, Gillatt D, Persad R, Hamdy FC, Neal DE, Holly JMP 2004 Do height-related variations in insulin-like growth factors underlie the associations of stature with adult chronic disease? J Clin Endocrinal Metab 89:213–218[Abstract/Free Full Text]
  5. Ben-Shlomo Y, Holly J, McCarthy A, Savage P, Davies D, Gunnell DJ, Davey Smith G 2003 An investigation of fetal, postnatal and childhood growth with insulin-like growth factor I and binding protein 3 in adulthood. Clin Endocrinol (Oxf) 59:366–373[CrossRef][Medline]
  6. Helle SI, Ekse D, Holly JM, Lonning PE 2002 The IGF-system in healthypre- and postmenopausal women: relations to demographic variables and sex-steroids. J Steroid Biochem Mol Biol 81:95–102[CrossRef][Medline]
  7. Holly J 2004 Physiology of the IGF system. In: Bock G, Goode J, eds. Biology of IGF-1: its interaction with insulin in health and malignant states. Chichester, UK: John Wiley; 19–26
  8. Fall CHD, Pandit AN, Law CM, Yajnik CS, Clark PM, Breier B, Osmond C, Shiell AW, Gluckman PD, Barker DJ 1995 Size at birth and plasma insulin-like growth factor-I concentrations. Arch Dis Child 73:287–293[Abstract/Free Full Text]
  9. Renehan AG, Zwahlen M, Minder C, O’Dwyer S, Shalet SM, Egger M 2004 Insulin-like growth factor (IGF)-1, IGF binding protein-3, and cancer risk: systematic review and meta-regression analysis. Lancet 363:1346–1353[CrossRef][Medline]
  10. Pollak MN, Schernhammer ES, Hankinson SE 2004 Insulin-like growth factors and neoplasia. Nat Rev 4:505–518
  11. Davey Smith G, Gunnell DJ, Holly JMP 2000 Cancer and insulin-like growth factor-I. Br Med J 321:847–848[Free Full Text]
  12. Sandhu MS, Heald AH, Gibson JM, Cruickshank JK, Dunger DB, Wareham NJ 2002 Circulating concentrations of insulin-like growth factor-I and development of glucose intolerance: a prospective observational study. Lancet 359:1740–1745[CrossRef][Medline]
  13. Juul A, Scheike T, Davidsen M, Gyllenborg J, Jorgensen T 2002 Low serum insulin-like growth factor I is associated with increased risk of ischemic heart disease: a population-based case-control study. Circulation 106:939–944[Abstract/Free Full Text]
  14. Yu H, Rohan T 2000 Role of the insulin-like growth factor family in cancer development and progression. J Natl Cancer Inst 92:1472–1489[Abstract/Free Full Text]
  15. Oliver SE, Gunnell D, Donovan J, Peters TJ, Persad R, Gillatt D, Pearce A, Neal DE, Hamdy FC, Holly J 2004 Screen-detected prostate cancer and the insulin-like growth factor axis: results of a population-based case-control study. Int J Cancer 108:887–892[CrossRef][Medline]
  16. Zaina S, Nilsson J 2003 Insulin-like growth factor II and its receptors in atherosclerosis and in conditions predisposing to atherosclerosis. Curr Opin Lipidol 14:483–489[CrossRef][Medline]
  17. Bereket A, Lang CH, Wilson TA 1999 Alterations in the growth hormone-insulin-like growth factor axis in insulin dependent diabetes mellitus. Horm Metab Res 31:172–181[Medline]
  18. Kaaks R, Lundin E, Rinaldi S, Manjer J, Biessy C, Soderberg S, Lenner P, Janzon L, Riboli E, Berglund G, Hallmans G 2002 Prospective study of IGF-I, IGF-binding proteins, and breast cancer risk, in northern and southern Sweden. Cancer Causes Control 13:307–316[CrossRef][Medline]
  19. van den Beld AW, Blum WF, Pols HA, Grobbee DE, Lamberts SW 2003 Serum insulin-like growth factor binding protein-2 levels as an indicator of functional ability in elderly men. Eur J Endocrinol 148:627–634[Abstract]
  20. Martin RM, Davey Smith G, Mangtani P, Frankel S, Gunnell DJ 2002 Association between breast feeding and growth: the Boyd-Orr cohort study. Arch Dis Child Fetal Neonatal Ed 87:F193–F201
  21. Gunnell DJ, Davey Smith G, Frankel SJ, Kemp M, Peters TJ 1998 Socio-economic and dietary influences on leg length and trunk length in childhood: a reanalysis of the Carnegie (Boyd Orr) survey of diet and health in prewar Britain (1937–39). Paediatr Perinat Epidemiol 12(Suppl 1):96–113
  22. Wadsworth MEJ, Hardy RJ, Paul AA, Marshall SF, Cole TC 2002 Leg and trunk length at 43 years in relation to childhood health, diet and family circumstances; evidence from the 1946 national birth cohort. Int J Epidemiol 31:383–390[Abstract/Free Full Text]
  23. Gunnell DJ, Oliver SE, Peters TJ, Donovan JL, Persad R, Maynard M, Gillatt D, Pearce A, Hamdy FC, Neal DE, Holly JMP 2003 Are diet-prostate cancer associations mediated by the IGF axis? A cross-sectional analysis of diet, IGF-I and IGFBP-3 in healthy middle-aged men. Br J Cancer 88:1682–1686[CrossRef][Medline]
  24. Leitch I 1951 Growth and health. Br J Nutr 5:142–151[CrossRef][Medline]
  25. Gunnell D 2002 Can adult anthropometry be used as a "biomarker" for prenatal and childhood exposures? Int J Epidemiol 31:390–394[Free Full Text]
  26. Gunnell D, Whitley E, Upton MN, McConnachie A, Davey Smith G, Watt GC 2003 Associations of height, leg length, and lung function with cardiovascular risk factors in the Midspan Family Study. J Epidemiol Community Health 57:141–146[Abstract/Free Full Text]
  27. Lawlor DA, Taylor M, Davey Smith G, Gunnell D, Ebrahim S 2004 Associations of components of adult height with coronary heart disease in postmenopausal women: the British women’s heart and health study. Heart 90:745–749[Abstract/Free Full Text]
  28. Signorello LB, Kuper H, Lagiou P, Wuu J, Mucci LA, Trichopoulos D, Adami HO 2000 Lifestyle factors and insulin-like growth factor 1 levels among elderly men. Eur J Cancer Prev 9:173–178[CrossRef][Medline]
  29. Juul A, Dalgaard P, Blum WF, Bang P, Hall K, Michaelsen KF, Muller J, Skakkebaek NE 1995 Serum levels of insulin-like growth factor (IGF)-binding protein-3 (IGFBP-3) in healthy infants, children, and adolescents: the relation to IGF-I, IGF-II, IGFBP-1, IGFBP-2, age, sex, body mass index, and pubertal maturation. J Clin Endocrinol Metab 80:2534–2542[Abstract]
  30. Goodman-Gruen D, Barrett-Connor E 1997 Epidemiology of insulin-like growth factor-I in elderly men and women. The Rancho Bernardo Study. Am J Epidemiol 145:970–976[Abstract/Free Full Text]
  31. Sandhu MS, Gibson JM, Heald AH, Dunger DB, Wareham NJ 2004 Association between insulin-like growth factor-I: insulin-like growth factor-binding protein-1 ratio and metabolic and anthropometric factors in men and women. Cancer Epidemiol Biomarkers Prev 13:166–170[Abstract/Free Full Text]
  32. 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[Abstract/Free Full Text]
  33. Garnett S, Cowell C, Bradford D, Lee J, Tao C, Petrauskas V, Fay R, Baur LA 1999 Effects of gender, body composition and birth size on IGF-I in 7- and 8-year-old children. Horm Res 52:221–229[CrossRef][Medline]
  34. Hoppe C, Udam T, Lauritzen L, Molgaard C, Juul A, Michaelson K 2004 Animal protein intake, serum insulin-like growth factor I, and growth in healthy 2.5-yr-old Danish children. Am J Clin Nutr 80:447–452[Abstract/Free Full Text]
  35. Kajantie E, Fall CH, Seppala M, Koistinen R, Dunkel L, Yliharsila H, Osmond C, Andersson S, Barker DJ, Forsen T, Holt RI, Phillips DI, Eriksson J 2003 Serum insulin-like growth factor (IGF)-I and IGF-binding protein-1 in elderly people: relationships with cardiovascular risk factors, body composition, size at birth, and childhood growth. J Clin Endocrinol Metab 88:1059–1065[Abstract/Free Full Text]
  36. Gunnell DJ, Davey Smith G, Frankel S, Nanchahal K, Braddon FEM, Pemberton J, Peters TJ 1998 Childhood leg length and adult mortality: follow up of the Carnegie (Boyd Orr) survey of diet and health in pre-war Britain. J Epidemiol Community Health 52:142–152[Abstract]
  37. Gunnell D, Davey Smith G, Holly JMP, Frankel S 1998 Leg length and risk of cancer in the Boyd Orr cohort. Br Med J 317:1350–1351[Free Full Text]
  38. Rowett Research Institute 1955 Family diet and health in pre-war Britain. Report. Dunfermline, UK: Carnegie United Kingdom Trust
  39. Martin RM, Gunnell DJ, Pemberton J, Frankel S, Davey Smith G 2005 The Boyd Orr cohort: an historical cohort study based on the 65 year follow-up of the Carnegie Survey of Diet and Health (1937–9). Int J Epidemiol 34:742–749[Free Full Text]
  40. Cheetham TD, Holly JM, Baxter RC, Meadows K, Jones J, Taylor AM, Dunger DB 1998 The effects of recombinant human IGF-I administration on concentrations of acid labile subunit, IGF binding protein-3, IGF-I, IGF-II and proteolysis of IGF binding protein-3 in adolescents with insulin- dependent diabetes mellitus. J Endocrinol 157:81–87[Abstract]
  41. Officer LH 2004 What is its relative value in UK pounds? Economic History Services (http://www.eh.net/hmit/ukcompare/).
  42. Shaper AG, Wannamethee G 1991 Physical activity and ischaemic heart disease in middle-aged British men. Br Heart J 66:384–394[Abstract/Free Full Text]
  43. Rogers WH1993 Regression standard errors in clustered samples. Stata Tech Bull Rep 3:88–94
  44. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC 1985 Homeostasis model assessment: insulin resistance and ß-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 28:412–419[CrossRef][Medline]
  45. Holt RI, Simpson HL, Sonksen PH 2003 The role of the growth hormone-insulin-like growth factor axis in glucose homeostasis. Diabet Med 20:3–15[CrossRef][Medline]
  46. Juul A, Bang P, Hertel NT, Main K, Dalgaard P, Jorgensen K, Muller J, Hall K, Skakkebaek NE 1994 Serum insulin-like growth factor-I in 1030 healthy children, adolescents, and adults: relation to age, sex, stage of puberty, testicular size, and body mass index. J Clin Endocrinol Metab 78:744–752[Abstract]
  47. Ong K, Kratzsch J, Kiess W, Dunger D 2002 Circulating IGF-I levels in childhood are related to both current body composition and early postnatal growth rate. J Clin Endocrinol Metab 87:1041–1044[Abstract/Free Full Text]
  48. Fall CH, Clark PM, Hindmarsh PC, Clayton PE, Shiell AW, Law CM 2000 Urinary GH and IGF-I excretion in nine year-old children: relation to sex, current size and size at birth. Clin Endocrinol (Oxf) 53:69–76[CrossRef][Medline]
  49. Gapstur SM, Kopp P, Chiu BCH, Gann PH, Colangelo LA, Liu K 2004 Longitudinal associations of age, anthropometric and lifestyle factors with serum total insulin-like growth factor-I and IGF binding protein-3 levels in black and white men: the CARDIA Male Hormone Study. Cancer Epidemiol Biomarkers Prev 13:2208–2216[Abstract/Free Full Text]
  50. DeLellis K, Rinaldi S, Kaaks RJ, Kolonel LN, Henderson B, Le Marchand L 2004 Dietary and lifestyle correlates of plasma insulin-like growth factor-I (IGF-I) and IGF binding protein-3 (IGFBP-3): The Multiethnic Cohort. Cancer Epidemiol Biomarkers Prev 13:1444–1451[Abstract/Free Full Text]
  51. Rogers I, Emmett P, Gunnell D, Dunger D, Holly JMP 2006 Milk as a food for growth? The IGF link. J Pub Health Nutr, in press
  52. Le Roith D 1997 Insulin-like growth factors. N Engl J Med 336:633–640[Free Full Text]
  53. Lucas A 1994 Role of nutritional programming in determining adult morbidity. Arch Dis Child 71:288–290[Free Full Text]
  54. Davey Smith G, Ebrahim S 2005 What can mendelian randomisation tell us about modifiable behavioural and environmental exposures? BMJ 330:1076–1079[Free Full Text]
  55. Friedlaender JS, Costa Jr PT, Bosse R, Ellis E, Rhoads JG, Stoudt HW 1977 Longitudinal physique changes among healthy white veterans at Boston. Hum Biol 49:541–558[Medline]



This article has been cited by other articles:


Home page
 J. Clin. Endocrinol. Metab.Home page
R. M. Martin, D. Gunnell, E. Whitley, A. Nicolaides, M. Griffin, N. Georgiou, G. Davey Smith, S. Ebrahim, and J. M. P. Holly
Associations of Insulin-Like Growth Factor (IGF)-I, IGF-II, IGF Binding Protein (IGFBP)-2 and IGFBP-3 with Ultrasound Measures of Atherosclerosis and Plaque Stability in an Older Adult Population
J. Clin. Endocrinol. Metab., April 1, 2008; 93(4): 1331 - 1338.
[Abstract] [Full Text] [PDF]

Home page
 J. Clin. Endocrinol. Metab.Home page
R. M. Martin, J. M. P. Holly, G. Davey Smith, and D. Gunnell
Associations of Adiposity from Childhood into Adulthood with Insulin Resistance and the Insulin-Like Growth Factor System: 65-Year Follow-Up of the Boyd Orr Cohort
J. Clin. Endocrinol. Metab., September 1, 2006; 91(9): 3287 - 3295.
[Abstract] [Full Text] [PDF]

Home page
 J. Clin. Endocrinol. Metab.Home page
I. Rogers, C. Metcalfe, D. Gunnell, P. Emmett, D. Dunger, J. Holly, and and the Avon Longitudinal Study of Parents and Chi
Insulin-Like Growth Factor-I and Growth in Height, Leg Length, and Trunk Length between Ages 5 and 10 Years
J. Clin. Endocrinol. Metab., July 1, 2006; 91(7): 2514 - 2519.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bray, I.
Right arrow Articles by Martin, R. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bray, I.
Right arrow Articles by Martin, R. M.
Related Collections
Right arrow Neuroendocrinology and Pituitary
Right arrow Pediatric Endocrinology
Right arrow Metabolism

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology Endocrine Reviews J. Clin. End. & Metab.
Molecular Endocrinology Recent Prog. Horm. Res. All Endocrine Journals