OBJECTIVE. The aim of this study was to investigate the effect of catch-up growth occurring at different stages of childhood on glucose levels and β-cell function at 7 years of age.
METHODS. Oral glucose tolerance tests were performed on 152 7-year-old children. Anthropometric data were available from birth to 7 years of age. Children were split into catch-up, catch-down, and normal-growth groups on the basis of growth rates between birth and 1 year, birth and 5 years, and birth and 7 years. Fasting and 30- and 120-minute blood samples collected during the oral glucose tolerance tests were assayed for glucose, insulin, proinsulin, and des-31,32-proinsulin levels, and area-under-the-curve values were calculated.
RESULTS. Children with catch-up growth between birth and 5 years or birth and 7 years had greater area-under-the-curve insulin levels than the children with catch-down growth. Children with catch-up growth only between birth and 7 years exhibited higher proinsulin levels and a greater insulin secretory response to glucose than those who experienced catch-up growth between both birth and 1 year and birth and 7 years of age. Low birth weight children with no catch-up growth between birth and 7 years had the highest glucose and lowest insulinogenic index levels, whereas children with high birth weight and catch-up growth had the highest insulin levels.
CONCLUSIONS. Extremes of birth weight in conjunction with extremes of postnatal growth are all detrimental to childhood metabolism. The negative metabolic effects of catch-up growth between birth and 7 years may be attenuated if catch-up growth also occurs between birth and 1 year of age.
Catch-up growth is characterized by an increased growth velocity in height and/or weight after the removal of some constraint on normal growth.1 This increased velocity brings a child's height-for-age or weight-for-age status back toward the normal centiles and in the best-case scenario actually returns his or her growth pattern to its preinsult status.2 Catch-up growth has been documented after the removal of an insult in childhood and adolescence and also during infancy. In this latter scenario, the growth of a fetus is considered to have been constrained during gestation and, when freed from this constraint, the affected infant demonstrates rapid growth to reach his or her genetically determined growth canal.3 Although infantile catch-up growth has been studied in children who were small for gestational age (SGA) at birth and who were thus thought to have suffered from intrauterine growth retardation, rapid growth during infancy in children who have not been suffering from intrauterine insult and are not SGA has been the focus of more recent research.4–8 These authors have indicated that rapid catch-up weight gain in infancy is associated with obesity and a more centralized fat distribution later in childhood.7 In an analysis of South African urban children, Cameron4 was able to conclude that the majority of such children were overweight by 2 years of age and more likely to be classified as obese by 9 years of age. In addition, such children did not appear to have any advancement or delay in their skeletal maturity that might explain the differences in growth rate of rapidly and slower growing children.5
The presence of increased risk factors for overweight and obesity would suggest that such children may also be at an increased risk of developing insulin resistance. Indeed, we have demonstrated previously that insulin resistance was apparent in 7-year-old children who experienced rapid growth between birth and 7 years.9 A recent study also demonstrated that children who experienced catch-up growth between birth and 3 years were more insulin resistant than children with normal and catch-down growth at 8 years of age.10 These data suggest that both rapid postnatal weight gain and sustained childhood weight gain can influence insulin sensitivity. However, no study has been performed to separately investigate the effects of catch-up growth in infancy (0–1 year of age), infancy and early childhood (0–5 years), and throughout childhood (0–7 years) on insulin and glucose metabolism. Therefore, the aims of this study were to (1) compare the effects of catch-up growth occurring between birth and 1, 4, 5, and 7 years of age on glucose tolerance and insulin sensitivity in 7-year-old African children, (2) compare the glucose and insulin effects of catch-up growth occurring only between birth and 7 years with catch-up growth occurring between both birth and 1 year and birth and 7 years, and (3) study the effect of the absence and presence of catch-up growth on glucose and insulin levels in children with low or high birth weights.
The sample was selected from the Birth to Twenty birth cohort study, which enrolled voluntary participants from all births that took place between April 23 and June 8, 1990, in the Soweto-Johannesburg conurbation, South Africa.11 A sample of black African children was selected from this group. All had experienced term gestation and had complete data for birth weight and weight and height at 1 and 5 years of age. Two hundred forty participants (120 boys, 120 girls) were randomly selected from the 466 participants who met these selection criteria. Field workers visited each of these 240 families to provide information about the study and request participation. A sample of 152 children (79 boys, 73 girls), which was greater than that required as a result of power analysis (n = 110; β = .80; α = .05), agreed to participate. The study was approved by the University of the Witwatersrand Faculty of Health Sciences Committee for Research on Human Subjects (Medical).
Oral Glucose Tolerance Test
An oral glucose tolerance test (OGTT) was performed on all children at the age of 7 years. The children fasted for 10 to 12 hours before the test. An anesthetic cream was applied to the site of the decubitus vein, and 30 minutes later a cannula was inserted. A fasting blood sample was taken, and 1.75 g·kg−1 of glucose was given to each child orally in the form of Lucozade. Additional blood samples were taken 30 and 120 minutes after the glucose load. Blood samples were immediately centrifuged, and glucose concentrations were measured. Serum samples for insulin, proinsulin, and des-31,32-proinsulin were stored at −70°C until assayed.
Glucose levels were measured by using commercially available enzymatic colorimetric methods (Boehringer Mannheim, Mannheim, Germany). Insulin was measured by using an insulin-specific immunoenzymetric assay (Medgenix, Fleurus, Belgium). The lower limit of sensitivity for the insulin assay was 1.0 pmol/L, and the intraassay and interassay coefficients of variations were 3.0% to 5.3% and 5.6% to 9.8%, respectively. Proinsulin and des-31,32-proinsulin levels were assayed by using a 2-site immunoradiometric assay.12 The lower limit of detection for both assays was 1.0 pmol/L, and the intraassay and interassay coefficients of variations were <5% each.
Measurement of Insulin Resistance and β-Cell Function
Insulin resistance in the fasting state was calculated by using homeostasis model assessment (HOMA).13 β-cell function was assessed by using the insulinogenic index (the ratio between the change in insulin and the change in glucose over the first 30 minutes of the OGTT).14 The total amounts of insulin, proinsulin, des-31,32-proinsulin, and glucose produced during the OGTT was assessed by calculating the areas under the curve (AUC) for these 4 analytes by using the trapezoid rule.
At each measurement occasion, data were collected on height and weight by using standard techniques.15z scores (SD scores) for weight were determined from the Centers for Disease Control and Prevention/World Health Organization/National Center for Health Statistics 2000 database within the statistical package Anthro (Centers for Disease Control and Prevention, Atlanta, GA).
Ong et al7 defined catch-up growth during infancy as an increase in z score of >0.67 between birth and 2 years. This change is equivalent to a deviation in the growth curve on reference charts of 1 centile band width (eg, from the 25th to 50th centile, 50th to 75th centile, etc). Catch-down growth was defined as a decrease in z score of >0.67, and a z-score change of <0.67 in either direction defines “normal” growth. In this study, changes in z score were calculated for the periods birth to 1 year (birth-to-1 group), birth to 4 years (birth-to-4 group), birth to 5 years (birth-to-5 group), and birth to 7 years (birth-to-7 group). These time points were chosen because weight was available for all children at these ages.
The influence of birth weight on the metabolic effects of catch-up growth was analyzed by splitting subjects into groups on the basis of birth weight tertiles. Two groups were produced from the bottom tertile by subdividing them into those who had experienced catch-up growth between birth and age 7 (LBW-CU group) and those who had experienced normal or catch-down growth (LBW-noCU group). The same process was applied to children in the combined middle and top birth weight tertiles to produce the high birth weight with catch-up growth (HBW-CU) and high birth weight with normal or catch-down growth (HBW-noCU) groups.
The effect of early, in combination with sustained, catch-up growth on glucose metabolism was analyzed by comparing a group of subjects who had experienced catch-up growth between birth and 7 years only (CU7) with a group who had experienced catch-up growth between birth and 1 year and birth and 7 years (CU1/7).
Variables that displayed a non-Gaussian distribution were normalized by log transformation or by expressing as reciprocals. One-way analysis of covariance (ANCOVA) was used to compare metabolic and anthropometric differences between groups after adjusting for gestational age, age and weight at time of the OGTT, and gender, with the Tukey test for unequal n numbers used as the posthoc test for comparing paired means. Two-way ANCOVA was used to analyze the interaction between birth weight and catch-up growth on the metabolic variables, with adjustment for gender, gestational age, and age and weight at time of the OGTT. Multiple regression analyses were performed to determine the effect of change-in-weight z score (calculated by subtracting the birth weight z score from weight z scores at age 1, 4, 5, or 7 years) on metabolic variables with gender, gestational age, and age at the time of OGTT included as independent variables within each model. In the text, anthropometric and metabolic data are expressed as means ± SDs.
There were no significant differences between the genders for birth weight or the ages at which assessments were undertaken. The boys were taller (74.5 ± 3.0 vs 72.7 ± 3.0 cm; P < .05) and heavier (9.98 ± 1.26 vs 9.10 ± 1.31 kg; P < .05) than the girls at 1 year of age, but there were no significant differences at other ages; therefore, the boys and girls were analyzed as a single group, and gender was included as a covariate in all statistical analyses.
Children displaying catch-down growth between birth and 1 year, birth and 4 years, birth and 5 years, or birth and 7 years were significantly heavier at birth and had higher gestational ages than the children who displayed normal or catch-up growth over the same time periods (P < .001 for all birth weight comparisons; P < .05 for all gestational age comparisons), whereas the children who displayed catch-up growth over any of these 4 time periods had lower birth weights than the children with normal growth (P < .001 for all birth weight comparisons). At age 7 (the time of the OGTT), children who displayed catch-up growth between birth and 1 year, birth and 5 years, or birth and 7 years were heavier than those who displayed normal or catch-down growth over the same time periods (P < .05 for all comparisons), whereas children who displayed catch-up growth between birth and 4 years had similar weights to those who displayed normal and catch-down growth.
Metabolic Effects of Catch-up Growth Occurring Over Different Time Periods of Childhood
The data in Table 1 show AUC (total) values for des-31,32-proinsulin, proinsulin, insulin, and glucose for the OGTTs. No significant differences in these values were found between the growth subgroups for the birth-to-1 infants, whereas for the birth-to-5 and birth-to-7 subjects, total insulin levels were significantly higher (P < .05) in the catch-up than in the catch-down group. In the birth-to-7 group, total des-31,32-proinsulin levels were also higher in the catch-up than in the catch-down subjects (P < .05). No statistically significant differences were noted between the 3 birth-to-4 growth subgroups; however, the catch-up growth group had consistently higher levels of these metabolic variables than the other 2 groups. The normal and catch-down group subjects, therefore, were combined, and this combined group had lower total glucose levels than the catch-up growth group (613 ± 96 vs 648 ± 89 mmol/L × minutes, respectively; P < .05). The insulinogenic index and HOMA values were not found to be different between any of the growth subgroups.
Multiple regression analysis demonstrated no correlation of change-in-weight z scores between birth and 1 year with any of the metabolic variables. However, change-in-weight z scores between birth and 4, 5, and 7 years all correlated with 30-minute insulin (β = .25, P = .007; β = .28, P = .001; β = .22, P = .02, respectively) and total insulin (β = .37, P = .0002; β = .39, P < .0001; β = .31, P = .001, respectively) levels. The total proinsulin and total des-31,32-proinsulin levels also correlated with change-in-weight z scores between birth and 5 years and birth and 7 years, whereas only total des-31,32-proinsulin levels correlated with the change-in-weight z scores between birth and 4 years.
Catch-up Growth From Birth to 1 Year in the Presence or Absence of Catch-up Growth From Birth to 7 Years
The study subjects were split into 2 groups: the first group (CU7) experienced catch-up growth from birth to 7 years but not from birth to 1 year (n = 18). The second group (CU1/7) experienced catch-up growth from birth to 1 year and from birth to 7 years (n = 34). The CU7 and CU1/7 groups had similar gestational ages and weights at birth and 7 years, but the CU7 subjects were lighter at 1 year and had a higher insulinogenic index (Table 2) than those in the CU1/7 group (assessed by ANCOVA adjusted for weight, age at 7 years, and gender). Total insulin and total proinsulin levels were higher in the CU7 subjects but just failed to reach a statistically significant level (P = .09 for both). However, the 30-minute insulin level was significantly higher in the CU7 subjects (400 ± 297 pmol/L) than in the CU1/7 subjects (274 ± 104 pmol/L) (P = .05), and the fasting proinsulin (3.95 ± 1.64 vs 2.47 ± 1.40 pmol/L; P < .01) and 30-minute proinsulin (12.16 ± 4.69 vs 8.83 ± 5.30 pmol/L; P < .05) levels were also significantly higher in the CU7 subjects.
Effects on Glucose Tolerance of a Lack of Catch-up Growth in Low Birth Weight Subjects
The data shown in Fig 1 demonstrate that subjects in the LBW-noCU group had a raised 30-minute glucose level (7.05 ± 1.24 mmol/L) significantly higher than that of those in the HBW-noCU group (5.84 ± 0.99 mmol/L) (P < .01) but with insulin levels that were not significantly different than those in the other groups. The children in the HBW-CU group had the highest 30-minute insulin levels (383 ± 271 pmol/L), which were statistically significantly greater than those in the HBW-noCU group of children (248 ± 112 pmol/L) (P < .05).
The data in Table 3 show that the LBW-noCU group had the highest total glucose levels and just failed to reach statistical significance with the HBW-noCU group (P = .06). The LBW-noCU group had the lowest insulinogenic index, which was significantly lower than that for the HBW-CU group (P < .01). The HBW-CU group had the highest total insulin levels, which were significantly higher than those of the HBW-noCU group (P < .05). The HOMA did not differ across the groups.
Multiple regression analysis showed that for the subjects in tertile 1 of birth weight, the change in weight z scores between birth and 7 years correlated negatively with 30-minute glucose levels (β = −.36, P = .03) and positively with insulinogenic index (β = .34, P = .03). This measure of catch-up growth correlated with total (β = .32, P = .004), 30-minute (β = .22, P = .03) and 120-minute (β = .26, P = .03) insulin and total des-31,32-proinsulin (β = .32, P = .01) levels in the subjects in birth weight tertiles 2 and 3. Two-way ANCOVA demonstrated that there was a significant interaction between catch-up growth and birth weight for both 30-minute (F = 5.7, P = .02) and total (F = 5.2, P = .02) glucose levels.
To our knowledge, this study is the first to compare the effects of different periods of catch-up growth on glucose tolerance and insulin sensitivity in infants. The results show that catch-up growth between birth and 1 year had no adverse effect on derived insulin or glucose variables measured at age 7; however, sustained catch-up growth (ie, between birth and 7 years) leads to higher insulin levels. Catch-up growth at all ages leads to higher body weight at age 7. These data are similar to those from studies that have shown that catch-up growth between birth and 2 years leads to higher body weight7 and catch-up growth between birth and 3 years leads to greater levels of insulin resistance.10
One study has shown that SGA infants who experience catch-up growth during the first year of life have higher fasting insulin levels than SGA children who did not catch-up and appropriate-for-gestational-age children.16 The insulin levels were measured at 1 year of age; therefore, it is not known whether the children who showed catch-up growth will remain hyperinsulinemic later in childhood. A study of 385 preterm children with birth weights of <1850 g demonstrated that post–glucose-load insulin levels were highest in those children with the greatest increase in weight between 18 months and 9 to 12 years of age, whereas weight gain between birth and 18 months had no effect on insulin levels.17 Therefore, this study17 also suggested that rapid growth during the first year of life does not compromise insulin sensitivity later in childhood, but weight gain over the first 9 to 12 years of life does. However, it must be noted that 2 studies have shown no effect of childhood catch-up growth on insulin sensitivity.18,19
In this study we also investigated the effect of neonatal catch-up growth occurring in combination with sustained catch-up growth (the CU1/7 group). Children in this group had lower postprandial insulin levels and insulinogenic indices than those subjects who experienced catch-up growth between birth and 7 years only (the CU7 group) but similar glucose levels, which suggests higher postprandial insulin sensitivity in the CU1/7 subjects. The CU1/7 and CU7 groups had similar birth weights and weights at 7 years, but the former group had much higher weights at 1 year of age. It is possible that rapid growth between birth and 1 year in the CU1/7 group was a result of higher caloric intake, which leads to metabolic adaptations to an environment in which calorie availability is greater than that experienced in utero. However, in the CU7 group, weight gain may have occurred more slowly over the first year of life as a result of lower caloric intake. The neonates then would have remained adapted to an environment of nutritional thrift in which reduced insulin sensitivity of the peripheral tissues ensured adequate glucose supplies to the essential organs, particularly the brain. An alternative hypothesis would be that children in the CU1/7 group were more insulin sensitive than those in the CU7 group and, therefore, gained weight more rapidly in the first year of life because of the growth-enhancing effects of insulin.20 This theory is supported by a study that showed that postnatal growth immediately after birth correlates positively with insulin sensitivity.21 Also, some studies have shown that catch-up growth is not related to neonatal dietary intake during infancy,22,23 which suggests that catch-up growth is determined by intrauterine and/or genetic factors.
The use of 1-way and 2-way ANCOVA and multiple regression analysis all demonstrated that in subjects in the LBW-noCU group, low postnatal weight gain was associated with higher postprandial glucose levels and lower insulinogenic indices. These associations were not observed in subjects in the top 2 tertiles of birth weight. Similar data were observed when postnatal growth of this same cohort of children was analyzed by dividing subjects into 4 groups on the basis of the median values for birth weight and weight at 7 years of age. In this analysis, subjects with both birth weight and weight at 7 years below the median value had the highest glucose levels and the lowest insulin secretory responses to glucose.24 Our study also demonstrates that higher birth weight in conjunction with catch-up growth produces the highest postprandial insulin and des-31,32-proinsulin levels and the highest insulin secretory responses to glucose. One interpretation of these data is that a lack of catch-up growth in subjects with poor fetal growth leads to relative islet β-cell insulin secretory hypofunction and raises postprandial glucose levels, whereas catch-up growth in subjects with good fetal growth leads to relative β-cell insulin secretory hyperfunction and lower postprandial insulin sensitivity. Alternatively, and as discussed previously, it could be hypothesized that poor fetal growth leads to reduced β-cell function and/or mass, which causes reduced postnatal weight gain, whereas good fetal growth leads to increased β-cell mass and greater postnatal weight gain. Support for this hypothesis comes from a study that showed that intrauterine growth retardation in humans does lead to lower pancreatic β-cell numbers25 and that postnatal weight gain is greatest in neonates with the highest insulin sensitivity.21
Our study demonstrates that in subjects who experienced catch-up growth between birth and 7 years, des-31,32-proinsulin levels were higher than in those who experienced catch-down growth. Furthermore, although proinsulin and des-31,32-proinsulin levels were not abnormally high in subjects who experienced catch-up growth, there was a strong tendency for the concentration of these molecules to rise as the level of catch-up growth increased between birth and 5 years and birth and 7 years but not between birth and 1 year. These data suggest that prolonged catch-up growth may upregulate islet β-cell production of insulin precursors in an attempt to meet the greater demands for insulin synthesis. Studies have shown that elevated proinsulin and des-31,32-proinsulin levels are characteristic of impaired glucose tolerance26 and type 2 diabetes27,28 and that raised proinsulin concentrations predict the future development of type 2 diabetes.29 Thus, catch-up growth between birth and 7 years of age leads to a β-cell secretory pattern that is characteristic of a prediabetic state. Furthermore, postprandial insulin, but not glucose levels, are increased by catch-up growth, which demonstrates that catch-up growth causes postprandial insulin resistance. The elevated insulin levels were mirrored by increased insulin precursor levels, which suggests that the rise in proinsulin concentration was caused by increased secretory demand on the islet β cells resulting from the insulin resistance.30
The insulinogenic index has been shown to be a good marker of β-cell function31,32 and to give an accurate estimate of first-phase insulin secretion,31 and for these reasons, was chosen to be the primary method for determining β-cell function. We assessed insulin resistance by using the HOMA method, which is an indirect measure of insulin resistance; however, it has been shown to correlate well with the euglycemic, hyperinsulinemic clamp method13,33 and represents a safer and less invasive method than the clamp technique for measuring insulin resistance in children. It is interesting to note that in this study, catch-up growth had effects only on postprandial and not fasting glucose and insulin levels, as also observed in a previous study.17
Tertiles of birth weight were used to split subjects into low and high birth weight groups, which allowed us to produce 4 groups of subjects, each with a sufficiently high n number to allow meaningful statistical analysis. Also, the low birth weight group had a mean birth weight of 2.54 ± 0.25 kg, and 43% had birth weights of ≤2.5 kg, a recognized definition of low birth weight. Furthermore, 60% of the subjects who experienced catch-up growth fell into this tertile. The use of birth weight tertiles demonstrated that metabolic differences were apparent in subjects with low birth weight who did not experience catch-up growth, which suggests that this definition of low birth weight, at least in this population, does have some metabolic significance. Whether this technique can be applied to other ethnic groups requires additional investigation.
This study was limited to children aged ≤7 years. This age range was chosen because it covers the prepubertal period, thus avoiding the puberty-associated decrease in insulin sensitivity,34,35 which may have masked any influence of birth weight or postnatal weight gain on insulin secretion or action. Whether these data will also apply to children at a postpubertal age is not known. However, life-course studies have shown that subjects at the greatest risk of developing type 2 diabetes are those who experienced growth faltering between birth and 6 months of age.36 These data support the results from our study and suggests that the absence of catch-up growth in the first year of life can have long-term metabolic effects. In addition, high weight gain between 2 and 11 years of age has also been shown to be associated with an increased risk of type 2 diabetes,36,37 which correlates well with the data from our study that show that sustained weight gain from birth to 7 years was associated with higher insulin levels.
With this study we demonstrated that sustained high weight gain from birth to 4 years and beyond leads to increased postprandial insulin resistance, but this affect can be attenuated if catch-up growth occurs between birth and 1 year of age. We also demonstrated that subjects with low birth weight who display low postnatal weight gain between birth and 7 years are the most glucose intolerant and have the worst β-cell insulin secretory function, whereas children with high birth weight who display rapid postnatal weight gain between birth and 7 years are the most insulin resistant. These data suggest that the manipulation of neonatal and infant growth rates may influence childhood and possibly adult insulin secretion and action.
We thank Johnson & Johnson and SmithKline Beecham for help with funding this study. The Birth to Twenty study received financial and logistic support from the South African Medical Research Council, the Anglo-American Chairman's Fund, the University of the Witwatersrand (Johannesburg, South Africa), and the Wellcome Trust (London, United Kingdom). Dr Crowther received research funding from the National Health Laboratory Service and the South African Medical Research Council.
We thank the Chemical Pathology routine laboratory at Chris Hani-Baragwanath Hospital for performing the glucose assays, Sr Nomsa Ramela for taking the blood samples, and the Birth to Twenty research assistants and study participants.
- Accepted January 2, 2008.
- Address correspondence to Nigel J. Crowther, PhD, Department of Chemical Pathology, NHLS, University of Witwatersrand Medical School, 7 York Rd, Parktown 2193, Johannesburg, South Africa. E-mail:
Dr Gray's current affiliation is Department of Clinical Biochemistry, University Hospital, Durham, United Kingdom.
The authors have indicated they have no financial relationships relevant to this article to disclose.
What's Known on This Subject
Low birth weight in combination with postnatal catch-up growth leads to obesity and insulin resistance in children. One study has also shown that poor fetal growth in combination with poor postnatal growth leads to relative glucose intolerance in children.
What This Study Adds
Catch-up growth between birth and 1 year attenuates the negative effect of sustained rapid growth between birth and 7 years of age on insulin sensitivity. The absence of catch-up growth in children with low birth weight leads to lower glucose tolerance.
- ↵Tanner JM. Growth as a target-seeking function: catch-up and catch-down growth in man. In: Falkner F, ed. Human Growth: A Comprehensive Treatise. New York, NY: Plenum; 1986:167– 179
- ↵Tanner JM. Growth from birth to two: a critical review. Acta Med Auxol (Milano).1994;26 (1):7– 45
- Law CM, Shiell AW, Newsome CA, et al. Fetal, infant, and childhood growth and adult blood pressure: a longitudinal study from birth to 22 years of age. Circulation.2002;105 (9):1088– 1092
- ↵Ong KKL, Ahmed ML, Emmett PM, Preece MA, Dunger DB; ALSPAC Study Team. Association between postnatal catch-up growth and obesity in childhood: prospective cohort study [published correction appears in BMJ. 2000;320(7244):1244]. BMJ.2000;320 (7240):967– 971
- ↵Sobey WJ, Beer SF, Carrington CA, et al. Sensitive and specific two-site immunoradiometric assays for human insulin, proinsulin, 65–66 split and 32–33 split proinsulins. Biochem J.1989;260 (2):535– 541
- ↵Kosaka K, Hagura R, Kuzuya T. Insulin responses in equivocal and definite diabetes with special reference to subjects who had mild glucose intolerance but later developed definite diabetes. Diabetes.1977;26 (10):944– 952
- ↵Cameron N. The Measurement of Human Growth. London, United Kingdom: Croom-Helm; 1984
- ↵Wilkin TJ, Metcalf BS, Murphy MJ, Kirkby J, Jeffery AN, Voss LD. The relative contributions of birth weight, weight change, and current weight to insulin resistance in contemporary 5-year-olds: the EarlyBird Study. Diabetes.2002;51 (12):3468– 3472
- ↵Henderson G, Fahey T, McGuire W. Calorie and protein-enriched formula versus standard term formula for improving growth and development in preterm or low birth weight infants following hospital discharge. Cochrane Database Syst Rev.2005;2 (2):CD004696
- ↵Phillips DIW, Clark PM, Hales CN, Osmond C. Understanding oral glucose tolerance: comparison of glucose or insulin measurements during the oral glucose tolerance test with specific measurements of insulin resistance and insulin secretion. Diabetic Med.1994;11 (3):286– 292
- ↵Haffner SM, Miettinen H, Stern MP. The homeostasis model in the San Antonio Heart Study. Diabetes Care.1997;20 (7):1087– 1092
- ↵Anderson RL, Hamman RF, Savage PJ, et al. Exploration of simple insulin sensitivity measures derived from the frequently sampled intravenous glucose tolerance (FSIGT) tests. The Insulin Resistance Atherosclerosis Study. Am J Epidemiol.1995;142 (7):724– 732
- Copyright © 2008 by the American Academy of Pediatrics