Effects of Age and Time on Energy and Macronutrient Intake in German Infants and Young Children

Results of the DONALD Study

Abstract

Objectives: 

To describe effects of age and time on the intake of energy and macronutrients in infants and young children.

Patients and Methods: 

The energy and macronutrient intake of 720 DONALD study participants between 1989 and 2003 was evaluated by analysing 3027 three-day weighed dietary records (including test weighing of breast milk consumption) at 3, 6, 9, 12, 18, 24 and 36 months. Mixed models and logistic regression analyses were used to investigate the effects of age and time on intake.

Results: 

Age affected macronutrient patterns (percentage of energy intake, E%) more often in infants than in young children. The percentage of full and partial breast-fed infants increased between 1989 and 2003. Almost no time trends were found for energy intake. Macronutrient patterns changed with time; for example, protein intake decreased throughout infancy (between −0.05 and −0.17 E%/y). An increase in fat intake (0.23 and 0.40 E%/y) was compensated by a decrease in carbohydrates (−0.18 and −0.29 E%/y) in young infants (3 and 6 months). Opposite trends in macronutrient patterns were observed in infants and young children.

Conclusions: 

Increasing breast milk consumption over time was the most obvious factor influencing long-term trends in macronutrient patterns. In all of the age groups examined here, macronutrient patterns have come closer to the references over time.

INTRODUCTION

Physiological requirements for energy and most nutrients as well as neuromotor eating skills change fundamentally during early childhood. Dietary habits and, thus, nutrient intake may also change over time. In addition, specific nutrients, food patterns, for example, breastfeeding, or food-related behaviours in early childhood, may exert long-lasting effects on health later in life (^1-3). To gain a meaningful insight into the dietary patterns of infants and young children, it is mandatory to examine food and nutrient intakes at narrow age intervals and, ideally, in a longitudinal manner.

However, investigation of the dietary intake of infants and young children has mostly been confined to cross-sectional surveys (^4-7), only a few of which consider narrow age margins. So far, no study has examined the influence of time trends in breastfeeding on nutrient patterns at various age points.

The DONALD (DOrtmund Nutritional and Anthropometric Longitudinally Designed) study is an (open) cohort study that collects dietary data from infants and children using 3-day weighed dietary records. The study structure offers the unique opportunity of analysing the effects of age and time on food and nutrient intake in narrow intervals (⁸). Here we present data on the effects of age and time on the intake of energy and macronutrients in infants and young children ages 3 to 36 months during a 15-year study period, 1989 to 2003. We include measured consumption of breast milk and differentiate between breast-fed and nonbreast-fed groups to examine the effects of milk feeding.

PATIENTS AND METHODS

The DONALD Study

The DONALD study, started in 1985, focuses on the investigation of diet, metabolism, growth and development of healthy infants, children and adolescents between the ages of 3 months and 18 years. Details of the study structure and the results of various cross-sectional and longitudinal evaluations of the dietary data have been reported previously (^8,9).

Data Collection

Body weight is measured (to the nearest 0.1 kg) by using an infant weighing scale (Mettler PS 15) or, alternatively, a scale for subjects in a standing position (Seca 753 E). Recumbent length is measured (to the nearest 0.1 cm) using a measuring table for infants younger than 2 years, or height using a digital telescopic wall-mounted stadiometer (⁸).

Nutrition Survey

In the DONALD study, food consumption is assessed by using 3-day weighed dietary records (⁹). Here we present data from a total of 3027 dietary records that were collected between 1989 and 2003 from 720 subjects (on average, 4.2 records per subject) ages 3, 6, 9, 12, 18, 24 and 36 months.

Energy and nutrient intake was calculated using our continuously updated nutrient database LEBTAB. Here we report on intakes of energy, protein, fat and carbohydrates.

To control for underreporting, a ratio of reported energy intake to predicted basal metabolic rate below 0.97 was used as a cutoff for implausible individual 3-day intake (¹⁰). As a result, 1.6% of the records were excluded.

Data Analysis and Presentation

SAS procedures (version 8.2; Statistical Analysis System, Cary, NC) were used for data analysis. Nutrient intake was calculated as the individual mean of the 3 recorded days. Here energy intake was expressed per kilogram body weight per day (MJ/kg_bw/d) and intake of nutrients as percentage of energy intake (E%) to adjust for age-dependent variations in absolute intake.

Breastfeeding was categorised as proposed by the World Health Organisation where appropriate (¹¹). Exclusive (no other fluids) and predominant breastfeeding (including fluids, no milk) were combined into the category "full breastfeeding." Supplementary feeding of formula and/or complementary feeding of weaning foods were combined into the category "partial breastfeeding." To analyse time trends for breastfeeding by feeding categories, logistic regression for categorical data was used (PROC CATMOD).

The effects of age were examined on the basis of all available records for the 7 specific age groups.

To analyse the effect of milk feeding, infants in the breastfeeding group (BF) were those who received any amount of breast milk at the time of measurements. Infants in the non-breastfeeding group (non-BF) were those who did not receive breast milk at the time of measurements. Effects of milk feeding, expressed as differences between BF and non-BF groups, were examined in age groups up to 12 months.

To analyse the influence of time on the outcome variables (age-specific intake of energy and macronutrients), mixed linear and nonlinear models were used, when appropriate, in which the means of the data, the covariance structure (children in the family) and the effect of repeated measurements were modelled (PROC MIXED). P values less than 0.05 were considered significant. Trend results were noted as β: increase (+) or decrease (−) per age group. Because the relative intakes reported here (per kg_bw, E%) were similar for boys and girls, combined results for both sexes are presented.

RESULTS

Overall, body height and weight in the various age groups were similar to reference data from the Euro-Growth Study (¹²) (Table 1).

TABLE 1: Overall anthropometric characteristics (mean, SD) in 3- to 36-month-old infants and young children of the RONALD study from 1989 to 2003

Breastfeeding

Age affected breastfeeding rates such that the overall proportions of fully breastfed infants were 51% at 3 months, 9% at 6 months and 0% at 9 months, whereas the proportions of partially breastfed infants were 13% at 3 months, 44% at 6 months, 24% at 9 months and 9% at 12 months. Long-term time trends in full breastfeeding increased approximately 2-fold (3 months) and 10-fold (6 months) between 1989 and 2003, with similar results in partial breastfeeding (Fig. 1).

FIG. 1: Time trends in full and partial breast-feeding in infants ages 3, 6, 9 and 12 months in the DONALD study from 1989 to 2003.

Effects of Age

Age affected energy intake such that in absolute terms, energy intake (MJ/d) increased 2-fold between the ages of 3 and 36 months, whereas in relative terms, energy intake (MJ/kg_bw/d) decreased by approximately 20%. Overall protein intake increased to a maximum at the age of 18 months, whereas fat intake (E%) decreased during infancy but increased in childhood. Carbohydrate intake increased until the age of 9 months and was stable afterwards.

Effects of Milk Feeding

Effects of milk feeding were observed at the ages of 3 and 6 months when relative energy intake (MJ/kg_bw/d) was higher in non-BF than in BF infants. Also, in most infant age groups, the intake of protein and carbohydrates was higher in non-BF infants than in BF infants, whereas fat intake was always lower in the non-BF groups (Table 2).

TABLE 2: Overall intake of energy and macronutrients (mean, SD) according to age and milk feeding effect of time trends in 3- to 36-month-old infants and young children of the DONALD study from 1989 to 2003

At the age of 3 months, 5% of BF infants, compared with 33% of non-BF infants, were receiving some solid food, whereas almost all infants received solids at the age of 6 months (97% vs 80% in BF vs non-BF groups). For both groups, vegetables and fruits were the most frequent first solid foods.

Time Trends

Time trends in intake of energy and macronutrients are presented in Table 2. We found no significant time trends for energy intake (MJ/kg_bw/d), except for a slight decrease in infants age 6 months. Protein intake decreased in every infant age group (Fig. 2), whereas fat intake showed different directions depending on the age group (Fig. 3). A linear increasing trend was found for fat intake in infants ages 3 and 6 months. At the ages of 9 and 12 months, nonlinear time trends for fat intake were found in the form of a decrease until 1995 and an increase thereafter. In young children, total fat intake showed a linear decrease over time. Time trends for carbohydrate intake were contrary to trends in fat intake in most age groups (Fig. 4).

FIG. 2: Time trends in protein intake (percent of energy, E%) in infants ages 3, 6, 9 and 12 months in the DONALD study from 1989 to 2003.

FIG. 3: Time trends in the intake of fat (percent of energy, E%) in infants and young children ages 3, 6, 12, 18, 24 and 36 months in the DONALD study from 1989 to 2003.

FIG. 4: Time trends in the intake of carbohydrates (percent of energy, E%) in infants and young children ages 3, 6, 12, 18, 24 and 36 months in the DONALD study from 1989 to 2003.

DISCUSSION

This is the first study that longitudinally examines, in the same cohort, the effects of age and time on dietary characteristics in early childhood at narrow age margins. Our analysis mainly shows that (i) effects of age for dietary energy and macronutrients were most pronounced in the first year of life particularly for fat intake; (ii) BF infants had lower intakes of protein and carbohydrate, but higher intakes of fat than non-BF infants, resulting in only small differences in energy intake; and (iii) time trends were found for protein, fat, and carbohydrates in infants and were contrary to trends for fat and carbohydrates in young children. In both, trends in intake are approaching dietary recommendations.

Study Background

The DONALD study is a longitudinal, observational study in paediatric age groups at a local level in Germany (⁸). Among the strengths of the study are the frequently repeated and detailed measurements of nutrition and growth. Because of the elaborate assessment procedures, a convenient, nonrepresentative study sample of relatively high socioeconomic status has been recruited. However, breastfeeding rates and weaning habits in the DONALD study were very similar to the nationwide SuSe study in 1997/1998 (¹³).

Effects of Age

In Germany, weaning food is mostly consumed in the form of commercial products. Baby food in jars, in particular, have a low-fat content. This would explain why the proportion of fat in total energy intake was lower in the second half of the first year of life than in all other age groups (^14,15).

Our results show the effects of age on protein intake in the first 6 months of life when protein intake increased due to the transition from exclusive milk feeding to a more diverse weaning diet. Older infants and young children exhibited protein intakes 2 to 3 times higher than the German references (5-6 E%) (¹⁶), similar to studies in other European countries and the US (^4,5,17).

The transition from infant food to the family diet usually starts around the age of 10 months and is almost complete between 18 and 24 months of age. In this study, large changes in nutrient proportions were found in the age range from 12 to 18 months, probably due to the food-based changes. The resulting macronutrient patterns in young children's diet were similar to the diet of older children and adolescents in Germany and other European countries (^14,18).

Effects of Milk Feeding

Exclusive breastfeeding for 6 months, followed by partial breastfeeding and adequate weaning food, is recommended worldwide (^19,20). Only a few observational studies have examined nutrient intake in BF infants throughout the weaning period (^17,21). The breastfeeding mothers in our study applied the test weighing procedures for breast milk consumption, in addition to the normal 3-day weighed dietary records of weaning food. Our results confirm those from other studies (^17,21) and indicate that BF infants have a lower energy intake in the first half-year of life and have different macronutrient patterns than non-BF infants until the second half-year of life. The differences in the age groups older than 3 months probably result mainly from a later introduction of weaning food for BF than non-BF infants. This was also found in the SuSe study (13).

Time Trends

In the 15 years reported here from the DONALD study, a steady increase in breastfeeding rates was observed, but recommendations were not yet achieved. Breastfeeding rates in our sample are similar to those in cross-sectional studies in some other European countries, such as Italy, and to repeated surveys in the US, but all of them are much lower than in Scandinavia (^13,22,23). Regarding food consumption, we found a decreasing trend for weaning food groups (eg, potatoes/pasta/rice, cereals/bread, fruit, meat) at the age of 6 months. This points to a delayed introduction of weaning food over time, concomitant with an increase in rates of full breastfeeding. In addition, the introduction of laws prohibiting the advertisement of weaning foods for infants younger than 4 months of age, put into force in the European Community in the mid-1990s (²⁴), could have encouraged the postponed introduction of weaning food found in our study. The British Infant Feeding Survey in 2000 also found that mothers were introducing solids later than in 1995 (²⁵).

Increasing breastfeeding rates and decreasing consumption of weaning food are the main reasons for the macronutrient time trends, particularly the increase in fat intake observed in these infants. In contrast to infancy, fat intake decreased in young children, similar to older children and adolescents of the DONALD study (²⁶). Reference values were not yet reached in infants but were achieved in young children in 2003, the end of the period examined here. The suitable fat intake in early childhood, however, is still a matter of debate. Observational and intervention studies show that prudent diets even under 30 E% from fat are safe with regard to nutrient adequacy, growth and cognitive development in older infants and young children (^27,28).

The only time trends that we found in protein intake were decreasing trends in infancy due to the increase of breastfeeding and a decrease in meat intake. The decrease in relatively high protein intakes in infancy found in our sample would lead to a decreased osmolar load on kidney function (²⁹) and may reduce the risk for obesity later in life (^2,30,31). Although the carbohydrate intake increased in young children, the intake of added sugars remained almost stable. This indicates that more complex and nutritionally valuable carbohydrates have been consumed over time.

Despite significant trends in fat and carbohydrate intake in young children, we found only a few trends for food group intake. This finding points to the complex interactions of food groups and nutrients in common adult-type diets (²⁶).

CONCLUSION

In conclusion, the macronutrient patterns that we found in the early childhood diet were clearly affected by age and reflected the transition from a high-fat milk-based diet to a low-fat weaning diet in the infant age groups and to a higher-fat family diet that was almost established at the age of 18 months. Time trends in intake in the last 15 years were almost in favour of the nutritional and preventive dietary quality. Time trends for macronutrients in infancy were influenced by increasing breast-feeding rates concomitant with a delayed introduction of weaning food and were often contrary to the trends observed in young children. Despite significant long-term changes in macronutrient proportions, energy intake was hardly affected suggesting an undisturbed internal regulation of energy metabolism in the early childhood age groups examined here.

Acknowledgements:

The DONALD study is supported by the Ministry of Innovation, Science, Research and Technology of North Rhine-Westphalia, Germany.