Skip to main content

Childhood obesity: early programming by infant nutrition


The CHOP Project, designed at testing the protein hypothesis of later overweight, allotted formula-fed infants from 5 European Countries to two formulas (high vs low protein content, HP and LP groups, respectively) up to 12 months of life, with a parallel reference group of breast-fed infants. We have performed a statistic analysis about: energy and macronutrient intakes in the first two years of life and the results are here reported. For this analysis we have obtained the results of the dietary assessment. - Details of the dietary assessment: A weighed food record over three consecutives days (two week days and one weekend day) has been used for formula fed and breastfed infants throughout the intervention. Parents and caregivers recorded the exact amount of milk or other solid foods consumed by infants in monthly intervals up to the age of 9 months and at 12 months of age (T1, T2, T3, T4, T5, T6, T7, T8, T9, T12). The weighed food record, from T1 to T9, was structured into two sections: 'milk preparation and milk intakes' and 'solid food intakes'. The accuracy of recording was verified with the caregivers either during scheduled face-to-face visits or, if returned by mail, by telephone. - Data entry and storing. For data entry and primary data checking an own program was developed ('Nutrcalc'). To standardize nutritional data entry into Nutrcalc and to cope with problems in reporting accuracy, standard operating procedures (SOPs) were created. Completeness and accuracy for each day of the food diaries was recorded on a scale with 5 levels. Data of each centre were regularly synchronised with a central aggregation database. - Nutritional items considered in the project were: energy, animal and vegetal protein, amino acids (valine, leucine, isoleucine, tryptofhane, methionine, phenylalanine, tyrosine, cysteine, arginine, lysine, threonine), alcohol, fats and fatty acids (saturated, mono-unsaturated, poly-unsaturated, cholesterol, oleic , linoleic, linolenic, lauric, myristic, palimitic, stearic acid), vitamins (B12, folates, A, D), minerals (sodium, potassium, calcium, phoshor, iron, zinc, magnesium, iodine) carbohydrate (total, lactose, saccharose, starch, fibres soluble, fibres insoluble, monosaccharides, disaccharides, polysaccharides). Result Energy intakes. The two feeding groups analysed through 12 months shows similar intakes of total energy, except at 6 months, when the LP group ingested on a whole 25 kcal more than the HP counterpart (P = 0.07) and 6 kcal more than HP on a pro/kg basis (P< 0.0001) At 12 months the total energy intake of the BF group is higher that the two formula groups, but not on a pro/kg basis. Protein and macronutrient intakes. As expected, the HP group is associated with higher protein intakes at all the time points between recruitment and 12 months compared to the LP group that shows, conversely, higher fat intakes through the first 12 months, as expected per protocol (the energy equivalent of protein was substituted with fats in both the types of formulas). At 24 months the intakes of proteins in the two groups was similar. The reference, BF group, is similar to the LP group at T12 through T24, while shows lower protein intakes than the HF group at T12 (also on a pro kg basis) and T18. The BF group shows lower total fat intakes compared to both formula groups at 12 months. Finally, at no time point there are differences in carbohydrate intakes between the two formula groups, but the breastfed group shows lower intakes of CHO at 12 months compared to the HP formula group. Breastfed infants remained a unique model, with lower energy and fat, and higher carbohydrate intakes, compared to both formula groups, while protein intakes where similar to the LP group, at 12 months of age. During the complementary feeding period the introduction of solids raised the intake of proteins in both formula groups in a similar fashion. In conclusion, the basic hypothesis of the CHOP project has been successful in producing two different populations of formula fed infants as far as protein intakes in the first 12 months of age.
Total energy expenditure (TEE) has been measured at 6 months of live by doubly labelled water method. A report on the effect of feeding infants with breast milk or infant formula with low or high protein content on their total energy expenditure at 6 months-old, has been developed. A total of 66 energy expenditure measurements were completed successfully (children from Spain and Germany). Total energy expenditure per body weight (TEE/kg) was higher in the breastfed (BF) group and lower in the high-protein formula fed (FF) group but differences were not statistically significant. Data from this prospective study suggested that protein intake during the firsts months of life could be inversely related to adiposity at 6 month-old. We found that total fat mass (TFM) was lower for BF than for low-protein FF infants and, in turn, in those, TFM was lower than for high-protein FF infants. It should be stressed, however, that results from multiple linear regression analyses were only marginally significant. If TFM is truly associated with protein intake, from our data we could infer that an infant would have at 6 months of age, on average, a 14% or 21% increase in TFM if fed with a low- or high-protein formula, respectively, than if BF at least during the first 3 months of life. Given that we have not yet analyzed dietary data to check if total protein intake was lower in BF infants, we hypothesized that it could be plausible since, even protein concentration in low-protein formula (7.3%) is lower than that of breast milk (9% on average), total milk volume in FF tend to be higher than in BF infants. Another possible explanation could be that amino acid composition of proteins from breast milk could a lower impact on IGF-1 plasma concentration than that of formula milk. In our study, no differences in Lean Body Mass (LBM) were observed between BF and high-protein FF infants. This finding is consistent with the null association that we found between TEE and feeding mode since LBM is the main determinant of TEE. We found LBM to be lower in infants fed with low-protein formula than BF. It could be hypothesized that, even total daily protein intake in low-protein FF infants would be similar or even higher than in BF ones, proteins from formula present less bioavailability and produce less LBM accretion than breast milk proteins. Consistently with previous findings, LBM and TEE was higher in boys than in girls and TFM was higher in girls than in boys. Strengths of this study are its prospective design, which minimize bias, control for confounding and measurement of adiposity by a highly precise method (DLW). Possible weaknesses are its low power to detect differences that could really exist due to the limited sample size and the need of further follow-up. In conclusion, findings of this study support the hypothesis that early high protein intake in infancy is associated to adiposity at 6 month-old. Additional follow-up of the cohort to see if this observation is maintained over time would be desirable.
Laboratory parametres were investigated in children from 4 countries (Germany, Spain, Belgium, Poland) where local ethics committees approved urine and blood sampling for hormonal and biochemical determinations. Written informed consent was obtained from parents of the infants studied. Blood samples were taken at the age of 6 months and urine samples were taken twice- at the age of 3 months and 6 months. RESULTS Serum concentrations of total IGF-1, free IGF-1 were increased in infants fed high protein formula, but IGF-BP2 concentrations were decreased when compared to results obtained in infants receiving low-protein formula. We did not find any significant difference in IGF BP3 serum concentrations between low- and high-protein groups. We also observed increased urine C-peptide/creatinine ratio in children on high protein diet and decreased serum glucose concentrations. Serum levels of several amino acids, with special respect to leucine, isoleucine and valine (branched chained amino acids - BCAA) were measured. Each of these amino acids as well as total sum of BCAA was found to be increased in children fed with high protein diet. 18 amino acids were measured in blood of each infant. The concentration of all essential amino acids (EAA) was significantly above the recommended minimum. For most of the EAA the distributions were right skewed. Percent participation of each AA in the total pool of amino acids was calculated in LP and HP group. Relative difference in each AA between study groups was calculated later on. Calculation took into account the difference (percentage value) of each single AA in the pool of amino acids. This method of data analysis allowed to identify the highest difference of 3 amino acids: isoleucine, leucine, valine (>25% more in HP than in LP group). There were significant correlations found between BCAA and IGF-1 (total and free) as well as C-peptide/creatinine. Urine/ creatinine concentrations did not differ significantly between low and high protein group. Discussion/ Conclusions The anthropometric results from our study confirm previous epidemiological observations of increased weight/height ratio and BMI in children fed formula in infancy. Consistent results of insulin like growth factor axis stimulation point to the important role of this hormonal regulation of body weight. In fact, free IGF-1 increases in obesity that was documented in many studies. High concentration of free-IGF-1 may be also partially explained by lower levels of IGF-BP-2 , whose production is inhibited by insulin, usually increased with increasing weight. In other populations studied IGF-BP-2 decreased in obesity and the results are very consistent. Furthermore, IGFBP-3 has a complementary stimulatory activity as IGF-1 and this protein is increased in the infants studied, but the difference was not statistically significant. The results of our study seem to be very consistent to point out stimulatory effect of high protein formula feeding on insulin-like growth factor axis. Glucose concentration was decreased in our high protein group opposite to urine C-peptide/creatinine ratio. One can speculate that higher insulin production (and C-peptide) is not stimulated by glucose but other metabolic factors such as insulin releasing amino acids. It seems from our results that insulin production reflected by urine C-peptide/creatinine is increased in infants consuming higher amounts of protein. We can speculate that sustained elevation of insulin synthesis may lead to accelerated growth. Insulin was proposed as a growth promoting factor during the fetal period but also children with insulin deficiency are retarded in growth. Our results confirm earlier findings of higher amino acid serum concentrations with higher protein intake. Both of the formulae studied appeared to be safe as the essential amino acid concentrations exceeded the minimum values. It is also indicated by the right skewed distributions. Even if it seems biologically plausible for the organism to eventually tolerate higher levels of substrates it cannot synthesize on its own, some untowords effects related to the tolerance mechanisms may be expected. Amino acids seem to play an important role in the insulin related metabolic activity and thus can be related to the obesity risk. Leucine is known as a stimulator of insulin release. Therefore these branch chain amino acids, whose concentrations increased to the greatest extend, were carefully
In a number of observational studies protein intake in infancy has been associated with both, rapid early weight gain and with later obesity. In the European Childhood Obesity Project more than thousand healthy, term, formula-fed babies were randomized to receive either a higher or a lower protein formula during the first year of life. The infants were enrolled in study centres Belgium, Spain, Poland, Italy and Germany. Weight, length, weight-for-length, body-mass-index and further anthropometric parameters were determined at inclusion and at 3, 6, 12 and 24 months of age. Anthropometric measures were transformed into standard deviation-scores based on the World Health Organization growth standards 2006 and were compared between the study groups. Infantile diet was evaluated from 3-day dietary records completed by the parents in monthly intervals during the first year of infantile life and at infantile age 24 months. Six hundred eighty five infants could be followed until 24 months of age in the intervention group, and additionally more than three hundred children in a non randomized reference group of breastfed infants were studied. Randomization to a higher protein formula was associated with a significantly higher protein intake during the first year of life, while the energy intake did not differ with formula allocation. There was no difference in length between the intervention groups at any time. Standard-deviation-scores for weight were significantly higher in the higher protein group at 3, 6, and 12 months of age. At 24 months the standard-deviation-score for weight, weight-for-length, and body-mass-index was found to be significantly higher in the higher protein formula group than in the lower protein formula group, respectively. From these data it may be concluded, that reducing the protein content in infant formula to approximate human milk was associated with a lower weight, weight-for-length and body-mass-index in the first 2 years of life. The low protein content in human milk might explain the lower risk for childhood obesity in breastfed children.