Insulin Resistance and Body Composition in Preterm Born Children During Prepubertal Ages

Feyza Darendeliler; Firdevs Bas; Ruveyde Bundak; Asuman Coban; Ozlem Sancakli; Sema Kabatas Eryilmaz; Banu Kucukemre; Rian Disci; Gulbin Gokcay; Semih Aki; Zeynep Ince; Nurten Eskiyurt


Clin Endocrinol. 2008;68(5):773-779. 

In This Article


Birth data of the children as seen in Table 1 showed the expected differences between the groups. Premature children had significantly lower gestational age than term children, by definition, and SGA children had lower birth weight and length than AGA children. However, birth weight and length SDS of the preterm SGA and term SGA children, and of preterm AGA and term AGA children were not different from each other.

At the time of the study, there was no pathology in physical examination of the children. All children were prepubertal and had no pubarche by inclusion criteria. As seen in Table 1 , all of the anthropometric parameters of the preterm SGA and term SGA children were similar at similar ages. Δ height SDS (1·5 ± 0·3) and Δ weight SDS (1·3 ± 0·2) of the preterm SGA children were similar to that of term SGA children (1·4 ± 0·3 and 1·7 ± 0·2, respectively), showing that they had made a similar CUG in magnitude both in height and weight, and reached a similar BMI and corrected height for their parents at similar ages.

Comparison of the preterm AGA and term AGA children showed that preterm AGA children were slightly older (P < 0·001), and their weight SDS and height SDS were slightly lower than that of term AGA children (P < 0·001) but they had similar BMI SDS and corrected height for parental height at investigation ( Table 1 ).

Table 2 shows that basal insulin and HOMA-IR were significantly lower in preterm SGA children than in term SGA children (P = 0·001 and P < 0·001). IGFBP-1 was significantly lower in term SGA children than in preterm SGA children (P = 0·005). Otherwise there were no differences with respect to IGF-1 and IGFBP-3, leptin and fat mass between the preterm SGA and term SGA children.

Comparison between preterm AGA and term AGA children as seen in Table 2 yielded lower IGF-1 (P = 0·001) in preterm AGA children. Fasting glucose was lower in preterm AGA children (P < 0·001). Insulin, HOMA-IR leptin levels and fat mass on DEXA were similar in preterm AGA and term AGA children. Basal insulin levels were within the expected ranges reported for children around these ages which are around between 14 and 40 pmol/l.[8,9,15,16] Leptin levels were also analysed in girls and boys separately. Leptin levels in preterm AGA girls was 2·6 ± 1·3 µg/l and in boys 1·9 ± 0·4 µg/l; in preterm SGA girls 2·6 ± 1·3 µg/l and boys 2·5 ± 1·9 µg/l. In term AGA girls, leptin levels were 3·2 ± 0·8 µg/l and in boys 2·2 ± 0·2 µg/l; in term SGA girls, it was 2·4 ± 3·5 µg/l and in boys 2·0 ± 0·4 µg/l. There were no significant differences with respect to gender in the preterm group (both AGA and SGA) and term SGA group. However, only term AGA girls had significantly higher leptin levels than term AGA boys (P = 0·018). There were no significant differences in leptin levels between the different groups with respect to gender.

There was no significant difference between premature AGA and premature SGA children with respect to several neonatal characteristics including feeding regimen (days of parenteral nutrition, duration of breast feeding and age at start of formula feeding), therapeutic intervention (duration and percentage of mechanical ventilation, days on O2 therapy), and antenatal and postnatal glucocorticoid use, except that days until full oral feeding instituted and days of hospitalization were longer in the premature SGAs than premature AGAs. The details of the study group with respect to these parameters are being published elsewhere.[31] There were no significant differences between preterm SGA and preterm AGA children with respect to height, weight, BMI and IGF-1 and IGFBPs, insulin levels, HOMA-IR and body composition at investigation.

There was no noteworthy pathology in the medical history of the term SGA children. Term SGA children had significantly higher insulin levels and HOMA-IR than term AGA children (P = 0·007 and P = 0·018) and lower IGFBP-1 (P = 0·013).

There were no significant differences between the groups with respect to cholesterol, triglyceride levels, HDL and either in the other lipid fractions (data not shown). fT4 and TSH were in normal ranges and were not different between the groups. fT4 and TSH were, respectively, 20·8 ± 0·7 pmol/l and 1·8 ± 0·3 mU/l in the preterm SGA group, 20·4 ± 0·8 pmol/l and 1·9 ± 0·3 mU/l in the preterm AGA group, 20·9 ± 0·7 pmol/l and 1·8 ± 0·2 mU/l in the term SGA group, and 20·0 ± 0·4 pmol/l and 1·8 ± 0·9 mU/l in the term AGA group.

The exclusion of the twins and triplet has not caused a major change in the insulin values.

There was no significant difference between the groups with respect to level of parental education (data not shown).

Insulin showed positive correlation with weight (r = 0·320, P = 0·003), Δweight (r = 0·243, P = 0·025), leptin (r = 0·249, P = 0·022) and IGF-I (r = 0·381, P < 0·001), and negative correlation with IGFBP-1 (r = -0·260, P = 0·017) in all preterms. It did not show correlation with birth weight. Insulin showed positive correlation with gestational age only in preterm SGA children (r = 0·428, P = 0·023). The correlation persisted when adjusted for birth weight (r = 0·419, P = 0·03). HOMA-IR showed similar correlations as insulin.

Truncal fat showed positive correlation with weight (r = 0·688, P < 0·001), Δweight (r = 0·424, P = 0·01) and leptin (r = 0·490, P = 0·003).

Univariate analysis of variance yielded a significant effect of gestational age (P < 0·001) and birth weight (P = 0·025) on insulin levels. Adjusted for BMI, these effects remained significant (P = 0·001 and P = 0·01, respectively).


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.