| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Unit of Paediatric and Perinatal Epidemiology (I.R., P.E.), Department of Community-Based Medicine, University of Bristol, Bristol BS8 1TQ, United Kingdom; Department of Social Medicine (C.M., D.G.), University of Bristol, Bristol BS8 2PR, United Kingdom; Department of Paediatrics (D.D.), University of Cambridge, Addenbrookes Hospital, Cambridge CB2 2QQ, United Kingdom; and Division of Surgery (J.H.), University of Bristol, Bristol, BS8 2HW, United Kingdom
Address all correspondence and requests for reprints to: Imogen Rogers, Unit of Paediatric and Perinatal Epidemiology, Department of Community-Based Medicine, University of Bristol, 24 Tyndall Avenue, Bristol BS8 1TQ, United Kingdom. E-mail: imogen.rogers{at}bristol.ac.uk.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Participants: A total of 675 children participated in the Avon Longitudinal Study of Parents and Children.
Results: IGF-I was strongly positively associated with growth in height in both sexes. Among boys, IGF-I was strongly associated with subsequent growth in both leg and trunk length, but there was no evidence that IGF-I was more strongly associated with one component of growth than the other. Among girls, IGF-I was strongly positively associated with growth in trunk but not leg length, although there was only weak evidence that these two associations differed in strength (P = 0.058).
Conclusions: These results support the contention that the associations between height and cancer may be mediated by variation in childhood IGF-I. However, they provide no evidence to support the hypothesis that leg length is a better biomarker of childhood IGF-I levels than trunk length.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In the prepubertal period, a greater proportion of growth in total height results from growth in leg rather than trunk length (12); it has therefore been suggested that adult leg length may be a marker of childhood IGF-I levels. In keeping with this suggestion and the possible link between height-cancer and IGF-cancer associations, there is some evidence that leg length is the component of height most strongly associated with cancer risk (13). No previous studies have investigated the association of childhood IGF-I with leg and trunk length and the subsequent rates of growth of these two components of height. In adults, height, leg length, and trunk length are not strongly associated with circulating IGF-I, although there is weak evidence that the ratio of adult leg to trunk length is associated with higher levels of bioavailable IGF-I (14). The lack of consistent association between IGF-I and adult height is unsurprising because final adult height is determined by not only the rate of childhood growth, which may be dependent on IGF-I, but also its duration, which depends on rate of maturation. Children who develop early may manifest higher growth rates in childhood but attain a lower final height if they go through puberty earlier than their peers and therefore have a curtailed growth period with early fusion of the epiphyses. Hence, if it is childhood exposure to IGF-I that is most strongly related to cancer risk in adulthood, adult height will be an imperfect measure of IGF-I exposure during the sensitive period.
The aim of the current study was to relate measures of IGF-I levels in healthy prepubertal children to subsequent growth in height, leg length, and trunk length. In addition, we investigated how associations of IGF-I with growth in height were influenced by IGFBP-3.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The children attended research clinics at the ages of 5, 7–8, and 9–10 yr. At each age, anthropometric measurements were taken, and at 5 and 7–8 yr, blood samples were taken for measurement of IGF-I.
IGF-I and IGFBP-3 measurements
Two different IGF-I assays were used. As part of a previous study (16) at age 5 yr, serum IGF-I levels were measured after acid ethanol extraction by a competitive solid-phase immunoassay. This was modified from the method of Kratzsch et al. (17) by the use of biotin for labeling of IGF-I and streptavidin-Europium (Wallac, Inc., Turku, Finland) for the detection of labeled molecules by time-resolved fluorescence. The limit of sensitivity was less than 15 ng/ml, and intra- and interassay coefficients of variation were less than 10% in the range 100–500 ng/ml.
At age 7–8 yr, serum concentrations of IGF-I were determined by RIA using a monoclonal antibody (Blood Products; Elstree, Hertfordshire, UK) and recombinant peptide (Pharmacia, Stockholm, Sweden) for standard and tracer, after iodination using the chloramine-T method. Samples were analyzed after acid-acetone extraction to remove the IGFBPs with an excess of IGF-II added to the extract to saturate any residual binding proteins (18). Serum levels of IGFBP-3 were determined by RIA using an in-house polyclonal antibody raised against recombinant non-glycosylated IGFBP-3. The assay was calibrated against recombinant glycosylated IGFBP-3 (Dr. C. Maack, Celtrix, Santa Clara, CA). The average coefficients of variation for intraassay variability for IGF-I and IGFBP-3 were 6.7 and 3.6% and interassay coefficients of variation were 12 and 14%, respectively.
Assessment of anthropometry at 5, 7–8, and 9–10 yr
Height and sitting height were measured to the last completed millimeter using a Harpenden stadiometer, a sitting height table, and an anthropometer (19). Leg length was obtained by subtracting sitting height (less the height of the stool) from total height. Weight was measured using a Seca 835 scale at 5 yr and a Tanita weighing scale (Tanita U.K. Ltd., Uxbridge, UK) at 7–8 and 9–10 yr. BMI was calculated from weight (kilograms)/height (meters)2.
Growth was calculated as change per week in height, leg length, and sitting height between the 5- and 7- to 8-yr clinics and between the 7- to 8-yr and 9- to 10-yr clinics.
Confounders
A questionnaire sent to the mother at 32 wk gestation provided information on the following: highest maternal educational level (grouped as CSE or no qualifications, vocational qualification, O level or equivalent, A level or equivalent, degree), housing tenure (grouped as council rented, i.e. government housing, other rented, owned, or mortgaged). CSE and O levels were, respectively, lower and higher levels of qualifications taken at around 16 yr of age. A levels were the standard qualifications taken at around 18 yr of age. In addition, birth weight, sex of the child, and gestation were obtained from hospital records. Gestational age was assessed on the basis of date of last menstrual period, ultrasound assessment, and other clinical indicators. Birth weight was adjusted for gestational age using the residuals method and converted to sex-specific z scores. BMI at 5 and 7 yr was converted to sex-specific z scores.
Paternal social class, maternal smoking in pregnancy, and the child’s self-reported Tanner stage of puberty were also initially considered as confounders but had a minimal effect on the observed associations between growth and IGF once the above confounders had been controlled and so were not included in further analyses.
Statistical analyses
Children providing at least one IGF-I measurement with anthropometry for the subsequent growth period and complete information on confounding variables were included in the analysis. Children from multiple births and nonwhite ethnic groups were excluded from analyses because there is evidence that they have distinctive growth patterns, and there were not enough children from these groups to analyze separately. One hundred fifty-five children provided IGF measures at the first time point only, 257 at the second time point only, and 390 at both time points. All analyses were performed for boys and girls separately because both IGF-I levels and growth patterns differ significantly between the sexes. Compared with children with no IGF data (n = 458), children with at least one IGF measure (n = 802) were less likely to have a mother whose highest educational qualification was CSE or less (11.4 vs. 16.5%). Among girls, mean leg length at 5 yr was higher (P = 0.020) for those with IGF measurements (48.7 cm, SD 2.4 cm, n = 321), compared with those without (48.0 cm, SD 2.4 cm, n = 88). There was no difference in leg length according to presence of IGF data at either 5 or 7 yr in boys or in height at either age in either sex. To allow comparison between the study sample and the general U.K. population, z scores for height and BMI were calculated using the British 1990 growth reference centiles (see Table 1
and Ref. 20).
|
IGFBP-3 measurements were available at 7–8 yr only. Analyses examined the role played by IGFBP-3 in growth in the second period only (7–8 to 9–10 yr), separating the independent effects of IGF-I and IGFBP-3.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Of 1260 white singleton children invited to attend CIF, 545 and 647 provided IGF measurements at 5 and 7–8 yr, respectively (one boy with a reduction of IGF-I of > 4 ng/ml/wk between 5 and 7–8 yr was excluded from analyses). Overall, 675 children provided at least one measurement of IGF and had complete data on confounding variables. Characteristics of these children are given in Table 1
. As expected, in both time periods (5 to 7–8 yr and 7–8 to 9–10 yr), the rate of growth in leg length exceeded that for trunk length in both sexes. IGF-I values did not increase between the ages of 5 and 7–8 yr as would have been expected from other surveys (2). This probably reflects the use of different assay methods at the two time points. However, IGF-I values tracked between the two time points [r = 0.48, P < 0.001 in boys (n = 205) and 0.38, P < 0.001 (n = 163) in girls].
Cross-sectional associations of IGF-I and anthropometry
Table 2 shows the cross-sectional associations of IGF-I at each age with height, leg length, trunk length, and BMI. At 5 yr IGF-I was positively associated with total height and leg and trunk length in both sexes. There was no evidence that the relationship of IGF-I with leg length was stronger than that with trunk length in either boys (P = 0.3) or girls (P = 0.8).
|
IGF-I levels and subsequent growth in height and leg and trunk length
Table 3 shows associations between IGF-I and subsequent growth. There were strong positive associations between IGF-I and subsequent rate of growth in height in both sexes, which were only slightly attenuated on controlling for confounders. There was no evidence that the strength of the association between IGF-I and growth in height differed between the sexes (P for interaction = 0.69).
|
Among girls there was a strong positive association between IGF-I and growth in trunk length, which was only slightly attenuated on adjustment for confounders. There was no apparent association between IGF-I and growth in leg length. However, there was only weak evidence that IGF-I was more strongly associated with growth in trunk than leg length in girls (P = 0.058 for adjusted associations).
There was no evidence that the strength of the association between trunk length and IGF-I differed between boys and girls (P interaction = 0.17); however, there was weak evidence that the association between IGF-I and growth in leg length was stronger in boys than girls (P interaction = 0.057).
The role of IGFBP-3 in growth in the second period
IGFBP-3 was measured only in the blood samples taken at age 7–8 yr. The correlation between IGF-I and IGFBP-3 at 7–8 yr was 0.47 in boys (P < 0.001, n = 298) and 0.47 in girls (P < 0.001, n = 257).
Table 4 shows associations of IGF-I at 7–8 yr with growth in height in the second period only along with the effect of adjusting for IGFBP-3 and the association of IGFBP-3 with growth on adjusting for IGF-I. Among boys, the positive association of IGF-I with subsequent growth in height was attenuated on controlling for IGFBP-3, and there was evidence of an independent positive association of IGFBP-3 with height growth. Among girls also the association of IGF-I with height growth was considerably attenuated on controlling for IGFBP-3, and IGFBP-3 had a significant independent association with height growth. There was no evidence that the association between IGFBP-3 and height growth differed between the sexes (P interaction = 0.21).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We know of only one other study of the relationship between childhood IGF-I and subsequent growth in height, which showed a positive association between IGF-I and growth in the subsequent year in 121 prepubertal children (22). We believe ours is the first study to relate IGF-I to subsequent growth in the leg and trunk components of height. We reported previously a cross-sectional study of the associations of IGF-I with height and the components of height in middle-aged men. We found that neither IGF-I nor IGFBP-3 was associated with total height, leg length, or trunk length, although the IGF-I to IGFBP-3 ratio was weakly positively associated with the leg to trunk ratio (14). Furthermore, in a recent analysis of the Boyd Orr cohort, there was no evidence that childhood or adult height or their components were associated with IGF-I or the molar ratio (IGF-I to IGFBP-3) measured in old age, although childhood trunk length was weakly positively associated with adult IGFBP-3 (23).
There is also considerable clinical evidence that circulating IGF-I is a strong determinant of childhood growth. Children with pathological disturbances of GH secretion have disturbed growth rates concomitant with altered circulating IGF-I levels (24), although because there is evidence that GH directly stimulates local IGF-I expression in the growth plates (25), this does not prove the role of circulating IGF-I. A child with a deletion of the IGF-I gene was reported to be severely growth retarded, but this could also have been due to local effects in the growth plate (26). The strongest evidence has possibly been from the systemic application of IGF-I, which results in an increase in circulating IGF-I concentration and linear growth rate (27), although again this has largely been in children with pathological growth disturbances. Our data indicate that variations in circulating IGF-I concentration within normal children are strongly associated with their linear growth over the subsequent 2 yr.
There are more data on the cross-sectional associations of IGF-I with total height, weight, and BMI. Positive associations between IGF-I and height have been consistently demonstrated in studies of prepubertal children in both industrialized (2, 3, 4, 5, 28) and developing (4) countries, although in two of these, the positive association with height was abolished after controlling for weight (3, 4). The results of studies in adults are more mixed. Most have been conducted on middle-aged subjects, and no association between IGF-I and height was observed in any of the studies that included women (23, 29, 30, 31) or in two of the studies of men (14, 30). However, positive associations between IGF-I and height were observed in studies of young (32), middle-aged (31), and elderly males (33).
Although the association of IGF-I with subsequent growth is examined in only one study (22), several studies have related IGF-I levels to previous growth. Among 9-yr-old children in Salisbury, greater catch-up in birth length/height SD score between birth and 9 yr was positively associated with IGF-I, as was greater height velocity between 7 and 9 yr (34). Similarly, in a study of 4-yr-old Indian children, the highest IGF-I concentrations were found in those who were small at birth and large at 4 yr (4), and a similar association with centile crossing from birth to 7–8 yr was found in Australian children (3). However, in a study of young adults in Wales, although catch-down growth from birth to 1 yr was associated with the lowest IGF-I levels, there was no evidence that catch-up growth in the first 5 yr of life was associated with elevated IGF-I in adulthood (32). In this study IGF-I levels at both ages were inversely associated with birth weight in contrast to the positive associations with contemporaneous anthropometric measures.
IGF-I to IGFBP-3 has been considered a crude indicator of IGF-I bioavailability (35). As such we expected increasing IGFBP-3 levels would attenuate any growth-promoting effect of IGF-I. However, the opposite appeared to be the case; indeed, we even found some evidence for an independent growth-promoting effect of IGFBP-3. There are relatively few other data on the associations between IGFBP-3 and growth or anthropometry in normal subjects, but positive associations among IGFBP-3, height, and BMI have been observed in both adults (31, 32) and children (2), sometimes in the absence of similar associations between IGF-I and anthropometry (2, 32), although not all studies found significant associations between IGFBP-3 and anthropometry (14).
Conclusions
We have found strong positive associations between IGF-I levels and subsequent growth in height in prepubertal children, consistent both with observed associations between previous growth and current IGF-I and with the hypothesis that adult height may be a biomarker of childhood IGF-I. As expected, growth in leg length was greater than in trunk length, highlighting its use as an indicator of prepubertal influences on growth. However, there was no evidence that IGF-I was more strongly associated with growth in leg than trunk length, casting doubt on the hypothesis that associations between adult cancer risk and leg length are mediated by variations in childhood IGF-I. There was some evidence of sex differences in the association of IGF-I with the components of height, associations of IGF-I with leg growth being stronger in boys than girls. Contrary to expectations, we found evidence that IGFBP-3 levels are independently positively associated with linear growth, and these unexpected results need to be replicated.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Summary: I.R., C.M., D.G., and P.E. have nothing to declare. D.D. is on a scientific advisory board for Ipsen and has received lecture fees from Pfizer and Novo Nordisk. J.H. has received consultancy fees from Roche, Aventis, and Diaganostics Systems Laboratories.
First Published Online May 2, 2006
Abbreviations: ALSPAC, Avon Longitudinal Study of Parents and Children; BMI, body mass index; CIF, Children in Focus; CSE, Certificate of Secondary Education; IGFBP, IGF binding protein.
Received February 21, 2006.
Accepted April 25, 2006.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  Y. Lu, J. Prescott, J. Sullivan-Halley, K. D. Henderson, H. Ma, E. T. Chang, C. A. Clarke, P. L. Horn-Ross, G. Ursin, and L. Bernstein Body Size, Recreational Physical Activity, and B-Cell Non-Hodgkin Lymphoma Risk Among Women in the California Teachers Study Am. J. Epidemiol., November 15, 2009; 170(10): 1231 - 1240. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Q. Wang and E. Seeman Growth-Related Cortical Fragility at Metaphyseal Regions IBMS BoneKEy, November 1, 2009; 6(11): 429 - 433. [Full Text] [PDF]
|
|
|
|
 |
|
 K. K Ong, M. Langkamp, M. B Ranke, K. Whitehead, I. A Hughes, C. L Acerini, and D. B Dunger Insulin-like growth factor I concentrations in infancy predict differential gains in body length and adiposity: the Cambridge Baby Growth Study Am. J. Clinical Nutrition, July 1, 2009; 90(1): 156 - 161. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Zuccolo, R. Harris, D. Gunnell, S. Oliver, J. A. Lane, M. Davis, J. Donovan, D. Neal, F. Hamdy, R. Beynon, et al. Height and Prostate Cancer Risk: A Large Nested Case-Control Study (ProtecT) and Meta-analysis Cancer Epidemiol. Biomarkers Prev., September 1, 2008; 17(9): 2325 - 2336. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 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]
|
|
|
|
|
|
V. M. Chia, L. C. Sakoda, B. I. Graubard, M. V. Rubertone, S. J. Chanock, R. L. Erickson, and K. A. McGlynn Risk of Testicular Germ Cell Tumors and Polymorphisms in the Insulin-Like Growth Factor Genes Cancer Epidemiol. Biomarkers Prev., March 1, 2008; 17(3): 721 - 726. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Perrin, H. Chen, D. E. Sandberg, D. Malaspina, and A. S. Brown Growth trajectory during early life and risk of adult schizophrenia The British Journal of Psychiatry, December 1, 2007; 191(6): 512 - 520. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
