HbA_1c in Nondiabetic Dutch Infants Aged 8–12 Months: The GECKO-Drenthe birth cohort study

The study population consisted of 86 Dutch infants participating in the Groningen Expert Center for Kids with Obesity (GECKO)-Drenthe study. This population-based birth cohort study within the GECKO was designed to examine risk factors for developing childhood obesity (2). A random subgroup (N = 100) of this study population, aged about 8 months, was invited to participate in the current study. Eighty-seven parents agreed to participate, and in eighty-six infants, an HbA_1c value could be assessed. This study was approved by the Medical Ethics Committee of the University Medical Center Groningen.

HbA_1c was measured in a capillary blood sample using a turbidimetric inhibition immunoassay on a Cobas Integra 800 CTS analyzer (Roche Diagnostics, Nederland BV, Almere, the Netherlands). This method has been standardized against the reference method of the International Federation of Clinical Chemistry and Laboratory Medicine (IFCC). Results (mmol HbA_1c/mol Hb) were converted to units (% HbA_1c) traceable to the Diabetes Control and Complications Trial/National Glycohemoglobin Standardization Program (DCCT/NGSP) using the Roche master-equation. Between-batch imprecision (coefficient of variation) was 2.1% for a mean HbA_1c of 5.5% and 1.9% for a mean HbA_1c of 10.6%.

Anthropometric measurements were performed by trained staff at Well Baby Clinics (2). Forty-eight percent of the infants did not visit the Well Baby Clinic within 2 weeks around blood sampling. For these infants, weight, length, and waist circumference were determined by linear interpolation. Weight, waist circumference, and weight-for-length z-scores were calculated using Growth Analyzer 3.5 (Growth Analyzer B.V., Rotterdam, the Netherlands), based on Dutch reference values (3,4). Growth velocity was defined as increase in weight between birth and time of blood sampling, in grams per week.

Data on gestational age, birth weight, infant feeding, gestational diabetes, maternal BMI, and parental educational level were obtained through questionnaires, and missing data was obtained from Well Baby Clinic files. Small for gestational age (SGA) and large for gestational age (LGA) were defined as birth weight below the 10th percentile and above the 90th percentile, respectively, for gestational age compared with Dutch reference values by parity and sex (5).

We used ANOVA to test for differences in mean HbA_1c between groups and linear regression to test the relationship between continuous variables and HbA_1c. For all analyses, a level of significance of P < 0.05 was applied. Statistical analyses were performed using SPSS 16.0 for Windows (SPSS, Chicago, IL).

Of the 86 infants included, 43 were girls and 43 were boys, with a mean (SD) age of 9.42 months (1.14). Mean (SD) HbA_1c was 5.38% (0.24), range was 4.8–6.0%, with a skewness (SE) of 0.215 (0.260) and a kurtosis (SE) of −0.004 (0.514). According to IFCC values, mean (SD) HbA_1c was 35.29 mmol/mol (2.65), range 29.1–42.1 mmol/mol (Fig. 1).

HbA_1c was unrelated to age or sex. Birth weight, growth velocity, anthropometric measurements, duration of breastfeeding, gestational diabetes, maternal BMI, and parental educational level were all unassociated with HbA_1c. HbA_1c did not differ between infants born SGA and infants not born SGA, or between infants born LGA and those not born LGA.

In this nondiabetic infant population, aged 8–12 months, HbA_1c was normally distributed, with a mean (SD) HbA_1c of 5.38% (0.24), range 4.8–6.0%.

The mean HbA_1c of 5.38% observed in the current study is higher than that found by studies in adults and older children (6–9). Because HbA_1c below 6% is considered normal, the clinical relevance of the difference between an HbA_1c level of 4.9% seen in adults and children, compared with 5.4% in our study population, is arguable. However, because the rate of formation of HbA_1c is directly proportional to ambient glucose concentration, our results could indicate higher glucose levels in 8- to 12-month-old infants. The predominant fuel for human cells is glucose, and survival of the brain depends on a continuous supply, yet the brain cannot synthesize glucose nor store more than a few minutes supply as glycogen. The infant brain is large relative to body mass and, especially in this period, rapid brain growth and differentiation takes place. To meet the high demand for glucose, the rate of glucose production in infants and young children is two to three times that of older children and mature adults (10). This high demand for glucose by the brain and the subsequent higher rate of glucose production might explain the higher HbA_1c levels observed in this study. Another explanation for the relatively high HbA_1c levels might be the higher percentage of infant fat mass compared with older children; fat mass is known to be positively related to insulin resistance (11,12). Unfortunately, glucose and insulin levels were not assessed, so we could not test these hypotheses.

At birth, between 55 and 65% of total hemoglobin synthesis consists of HbF. After birth, the production of the γ-chain declines to values of <5% by the age of 6 months; normal adult HbF values of <1% are usually reached by age 1 year (13). Because our study population’s age ranged from 8.1–12.3 months, somewhat higher levels of HbF might be expected. Glycated HbF is not detected by the turbidimetric inhibition immunoassay we used; HbF, however, is included in total hemoglobin determination. Because HbA_1c is expressed as a proportion, infants with high amounts of HbF (>10%) may have lower than expected HbA_1c values (14). However, this does not explain the relatively high HbA_1c levels found in our study.

The life span of a fetal erythrocyte is approximately 60–80 days, and the life span of erythrocytes in infants is still less than the 120-day life span of erythrocytes in adults (15). A shorter life span for erythrocytes should show lower HbA_1c levels instead of the higher levels we found.

To our knowledge, this is the first study describing the distribution of HbA_1c in nondiabetic infants. Compared with known HbA_1c levels in older children, we found a relatively high mean HbA_1c of 5.38%. No significant association between known risk factors for type 2 diabetes and HbA_1c was found.

No potential conflicts of interest relevant to this article were reported.

H.J. researched data, wrote and edited the manuscript, and contributed to the discussion. H.G.H. researched data, wrote and edited the manuscript, and contributed to the discussion. S.S. and R.P.S. contributed to the discussion and wrote, reviewed, and edited the manuscript. P.J.J.S. contributed to the discussion.

Parts of this study were presented in poster form at the 45th Annual Meeting of the European Diabetes Epidemiology Group, Porto Heli, Greece, 15–18 May 2010.

The authors are indebted to the children and their parents for their cooperation. The authors thank Ulf Ekelund and his team at the MRC Epidemiology Unit, Institute of Metabolic Science Cambridge, United Kingdom, for supportive collaboration.