A previous study showed low iron status in 12-month-old Icelandic infants associated most strongly with cow's milk intake and growth. Infant dietary recommendations were revised in 2003. This study investigated nutrition and iron status in a new infant cohort.
Adequate iron status in infancy is important because of its effects on health and development [
Iron status in infancy has been negatively associated with consumption of cow’s milk [
In 2003 revised infant dietary recommendations were published in Iceland, where iron-fortified formula was recommended in the weaning period from six months of age [
The objective of this study was to investigate a prospective cohort of 12-month olds, where data collection occurred after implementation of the revised infant dietary recommendations, to evaluate iron status and its association with diet and growth. Moreover, we aim to assess the impact of the revised recommendations on infants’ food and nutrient intake and iron status.
A random sample of 250 Icelandic infants born in 2005, from January to December, was collected by Statistics Iceland. The criteria for participation were the same as in the former prospective study on infant nutrition 1995–1997 [
Dietary data from 0–4 months of age was collected by dietary-history, including questions on breastfeeding, infant formula-feeding, other food items, and supplements. Food records were filled out monthly by the parents. Weighed food records were kept for three consecutive days (72 hours) at 9 and 12 months. All food was weighed on accurate electronic scales (PHILIPS HR 2385, Hungary; ±1 g accuracy). The breastfed infants were weighed with the same clothes before and after each breastfeeding session (Tanita model 1583, Japan or Sega model 336 7021099, Germany; both with ±10 g accuracy) to estimate the amount of breast-milk consumed as formerly described [
At 12 months of age blood samples were obtained for analyzing iron status. Hb, MCV, and SF were analyzed on the Coulter Counter STKS at Landspitali-The National University Hospital of Iceland. The definition for IDA includes all three indicators below cutoff points, Hb < 105 g/L, MCV < 74 fl, and SF < 12
Statistical analyses were performed with SPSS software (version 17; SPSS INC, Chicago, IL). Descriptive statistics were used for the participants’ characteristics and food and nutrient intake, presented as mean and standard deviation (SD) for normally distributed variables and median with interquartile range (IQR) for skewed variables. For comparison between two groups, independent
To identify predictors of the infants’ iron status indices (Hb, SF, and MCV) multivariate regression models were constructed. Variables which differed between iron-depleted infants and non-iron-depleted infants were used in the model for SF. Because of skewed distribution, SF was logarithmically transformed. The model for each iron status index also included their correlating variables (Pearson’s correlation coefficient), and for interrelated variables those with the weakest correlation to each iron status index were excluded from the model. Birth weight was adjusted for in the models. The sample size of 200, expecting 30% dropout, gives a power of 90% to detect the effect of 100 g weight gain on SF. To detect difference in iron status between the former (1995–1997) and current (2005–2007) cohorts the power was >90%. The level of significance in the study was
A total of 141 infants (73 boys and 68 girls) participated in the study, 72% of the 196 eligible participants. The characteristics of the subjects and their parents are presented in Table
Infant and parental characteristics of the subjects.
Boys | Girls | |
Mean (SD) | Mean (SD) | |
No. of subjects [ | 73 (51.8) | 68 (48.2) |
Infant characteristics | ||
Firstborn [ | 20 (30.3) | 24 (35.8) |
Birth weight (g) | 3807 (360) | 3750 (369) |
Birth length (cm) | 52.2 (1.52) | 51.7 (1.77) |
Birth head circumference (cm) | 36.2 (1.22) | 35.4 (1.14)* |
Weight at 6 mo (kg) | 8.4 (0.7) | 7.9 (0.75)* |
Length at 6 mo (cm) | 70.1 (1.7) | 67.9 (1.7)* |
Weight at 12 mo (kg) | 10.37 (1.0) | 9.76 (0.99)* |
Length at 12 mo (cm) | 78.0 (2.25) | 76.17 (2.60)* |
Hb at 12 mo (g/L) | 120.96 (8.19) | 120.28 (8.28) |
MCV at 12 mo (fl) | 76.71 (3.41) | 77.69 (3.09) |
log SF at 12 mo ( | 3.31 (0.63) | 3.57 (0.46)* |
Parental and socioeconomic characteristics | ||
Mother's BMI (kg/m2) | 24.3 (6.09) | 24.6 (5.38) |
Father's BMI (kg/m2) | 26.4 (4.14) | 25.1 (5.60) |
Mother's age (years) | 32.0 (5.25) | 30.3 (4.58) |
Father's age (years) | 34.1 (6.13) | 33.4 (5.22) |
Mother’s ≥12 y of schooling [ | 56 (82.4) | 57 (86.4) |
Father’s ≥12 y of schooling [ | 51 (77.3) | 51 (78.5) |
Smoker in the household [ | 14 (21.5) | 13 (20) |
Reside in capital and surroundings [ | 60 (62.5) | 60 (60) |
Marital status (married/partnership) [ | 63 (96.9) | 62 (98.4) |
*Significantly different from boys.
†Median and interquartile range (IQR).
In the present study 97.8% of mothers initiated breastfeeding and 38.1% of 5-month olds and 7.2% of 6-month olds were exclusively breastfed and 20.0% were still partially breastfed at 12 months of age. Average food intake was similar among boys and girls, except for milk consumption (Table
Intake of selected foods among boys and girls at 6, 9, and 12 months of age (
Boys | Girls | |||
% | Mean (SD) g/day | % | Mean (SD) g/day | |
Partial breastfeeding | 77.9 | 71.6 | ||
Exclusive breastfeeding | 11.1 | 3 | ||
Iron fortified formula | 21.9 | 19.8 | ||
Whole milk | 3.3 | 1.7 | ||
Iron-fortified formula | 59.7 | 44.2 (243.7)* | 68.3 | 48.8 (173.0)* |
Whole milk | 33.9 | 0 (14.7)* | 36.7 | 0 (27.5)* |
Breast milk† | 50 | 21.7 (287.6)* | 50 | 6.5 (297.5)* |
Fruit & vegetables‡ | 100 | 99.0 (61.5) | 98.3 | 89.5 (48.2) |
Porridge | 77.4 | 31.6 (33.1) | 85 | 30.2 (26.0) |
Dairy products§ | 50 | 2.9 (56.5)* | 55 | 3.8 (58.3)* |
Meat | 79 | 8.2 (21)* | 70 | 10.0 (26.9)* |
Iron-fortified formula | 53.4 | 56.2 (222.6)* | 75 | 212.9 (193.3)|| |
Whole milk | 56.9 | 11.8 (67.7)* | 51.9 | 6.7 (51.3)* |
Breast milk† | 26.5 | 0 (25.3)* | 13.4 | 0 (0)1, 5 |
Fruits & vegetables‡ | 100 | 83.7 (51.3) | 94.2 | 71.65 (46.2) |
Porridge | 82.8 | 78.6 (79.9) | 92.3 | 93.0 (83.7) |
Dairy products§ | 72.4 | 78.5 (81.1) | 90.4 | 90.3 (69.7) |
Meat | 100 | 39.4 (39.5) | 94.2 | 30.0 (26.3) |
*Median and interquartile range (IQR).
†Mean values include nonbreastfed infants.
‡The food group includes infant purees and fresh fruits and vegetables. Percentage represents infants receiving either fruits or vegetables.
§Dairy products are milk products and cheese, excluding drinking milk.
Intake of selected nutrients among boys and girls at 9 and 12 months of age as an average intake over a 3-day period (
Mean (SD) | Mean (SD) | ||
RDI (6–11 mo) | |||
Energy (kJ/kg) | 344.7 (82.4) | 349.4 (83.6) | 355 |
Protein (g/kg) | 2.63 (0.89) | 2.46 (0.87) | 1.1 |
Vitamin C (mg/d) | 67.2 (32.3) | 71.6 (28.3) | 20 |
Vitamin D ( | 9.7 (6.2) | 9.4 (9.9)* | 10 |
Vitamin A (RJ/d) | 854.9 (678.1)* | 947.6 (584.6) | 300 |
Zinc (mg/d) | 3.32 (1.78) | 3.23 (1.42) | 5 |
Iron (mg/d) | 6.28 (3.19) | 6.27 (2.73) | 8 |
Calcium (mg/d) | 510.8 (242.2) | 488.2 (212.7) | 540 |
RDI (12–23 mo) | |||
Energy (kJ/kg) | 351.1 (74.7) | 353.8 (80.4) | 355 |
Protein (g/kg) | 3.04 (0.99)* | 3.03 (0.95) | 1.0 |
Vitamin C (mg/d) | 58.1 (25.0) | 60.2 (32.1) | 25 |
Vitamin D ( | 9.1 (6.8) | 8.3 (5.2) | 10 |
Vitamin A (RJ/d) | 785.3 (999.0)* | 768.8 (492.1)† | 300 |
Zinc (mg/d) | 3.91 (2.01) | 2.86 (2.28)* | 5 |
Iron (mg/d) | 6.82 (3.97) | 5.77 (1.97)* | 8 |
Calcium (mg/d) | 565.2 (238.6) | 576.0 (186.3) | 600 |
* Median and interquartile range (IQR).
† Significantly different from boys (Mann Whitney
There were no children with IDA. ID affected 1.4% (
Iron-depleted infants had lower birth weight, 3402 (255.8) g versus 3805 (361.7) g (
SF level associated negatively with growth variables in infancy, but positively with birth weight. For growth from 0 to 12 months of age proportional weight gain (which takes birth weight into account) and absolute length gain correlated most negatively with SF level. For growth from 0 to 6 months weight gain correlated most negatively with SF, but growth from 6 to 12 months did not correlate significantly with the variable. After adjusting for birth weight multiple regression analysis showed that log SF level in boys decreased 0.0201
Multiple regression analysis of weight growth and food factors influencing change in iron status indices.
Dependent variable At 12 months | Independent variable | Boys | Girls | ||||
95% Conf interval of B | 95% Conf interval of B | ||||||
B | Lower Bound | Upper Bound | B | Lower Bound | Upper Bound | ||
Log SF | Length gain, cm (0–12 months) | −0.0823 | −0.150 | −0.014 | — | — | — |
Weight gain, 100 g (0–6 months) | −0.0201 | −0.040 | −0.001 | −0.0122 | −0.027 | 0.003 | |
Weight gain, 100 g (0–12 months) | −0.0174 | −0.032 | −0.003 | 0.00454 | −0.016 | 0.007 | |
Formula g/per day at 12 months | 0.00077 | −0.000062 | 0.0016 | 0.00077 | 0.00019 | 0.0014 | |
Porridge g/per day at 9 months | — | — | — | 0.0124 | 0.002 | 0.023 | |
Bread g/per day at 9 months | — | — | — | −0.0113 | −0.017 | −0.005 | |
Hb | Meat intake g/per day at 9 months | 0.127 | 0.043 | 0.211 | — | — | — |
Breastfeeding duration (mo) | −0.919 | −1.473 | −0.365 | — | — | — | |
Formula g/per day at 12 months | 0.0145 | 0.004 | 0.025 | — | — | — | |
Fruit intake g/per day at 12 months | — | — | — | 0.0384 | 0.009 | 0.068 | |
Dairy intake g/per day at 12 months | — | — | — | 0.0319 | 0.003 | 0.061 | |
Iron intake mg/per day at 9 months | 0.879 | 0.224 | 1.534 | — | — | — | |
MCV | Breastfeeding duration (mo) | −0.261 | −0.515 | 0.003 | — | — | — |
Iron intake mg/per day at 12 months | — | — | — | 0.297 | 0.061 | 0.534 | |
Vitamin C intake mg/per day at 9 months | — | — | — | 0.0359 | 0.008 | 0.064 |
Adjusted for birth weight.
—No association found.
Hb level had weaker associations with growth variables (positive) than SF level. In boys Hb associated most positively with dietary iron and meat consumption and most negatively with breastfeeding duration; formula had a weaker but significant association with Hb (positive). In girls fruits and dairy products were most strongly independently associated with Hb level (Table
MCV did not associate with any growth variable. In boys the only variable associated independently with MCV, of borderline significance, was breastfeeding duration (negative). In girls MCV level associated most positively with dietary iron intake and a weaker association was found with dietary vitamin C intake.
No associations were found between iron status indices and sociodemographic variables, that is, habitation, parents’ education, age, BMI, or smoking habits.
The present study describes nutrition and iron status in well-nourished infants with low prevalence of ID and IDA, high birth weight, and breastfeeding rate. Dietary intake in infancy has changed towards the infant dietary recommendations, revised in 2003, and iron status has improved when compared to results from an earlier prospective study of a comparable nationwide cohort. Breastfeeding has increased but the main alteration in the diet is an iron-fortified formula which has replaced regular cow’s milk in the latter half of the first year. The altered diet seems to have deleted cow’s milk as the main variable influencing serum ferritin and iron status. The formula was developed from domestic cow’s milk to diminish the change for the infants’ population as Icelandic cow’s milk has some characteristics different from major brands [
In comparison with the former study on infant nutrition (1995–1997), iron status of 12-month-old Icelandic infants has improved significantly (Figure
Box plots of iron status indices in 12-month olds, from previous (1995–1997) and present (2005–2007) studies. The box plots show the median, quartiles, maximum, and minimum values. Independent
In the present study the negative association between iron status and regular cow’s milk seen in the 1995–1997 infant study had disappeared as the consumption had decreased and been replaced by the iron-fortified formula (Figure
Difference in milk consumption between previous, (1995–1997) and present (2005–2007) studies, showed as mean and SD.
Meat consumption was most strongly independently associated with Hb in the present study’s multiple regression analysis. This association is consistent with the 1995–1997 study, and meat consumption did not differ between the two studies. Meat is an important source of iron and a known enhancer of good iron status [
Another factor negatively associated with iron status, which was consistent in both studies, was growth rate. In the present study, iron-depleted children had markedly faster weight and length gain than non-iron-depleted ones. This difference was also seen in the 1995–1997 study between iron-deficient children and non-iron-deficient ones [
Iron stores were worse among boys; their SF levels were significantly lower than in girls and 9.9% of boys were below the cut-off values versus 1.5% of girls, which can not be attributed simply to differences in growth rate. Dietary factors and growth had more impact on iron status in boys than in girls (Table
Iron status among 12-month-old Icelanders has improved enormously since the previous infant study. The largest alteration in the infants’ diet between the two studies is the replacement of regular cow’s milk by iron-fortified formula. The study demonstrates altered associations between iron status indices and food when the diet is changed. The lower intake of cow’s milk deleted former association with ferritin. The association with duration of breastfeeding, although weak changed from being positive in the former cohort study to being negative in the present study. The effect of breastfeeding may mainly be derived from the other food used, that is, milk or formula, but not the breastfeeding as such. The findings also show the effect and importance of seeking solutions for public health problems with better dietary advice.
The authors declare no conflict of interests.
This study was supported by the Icelandic Research Fund (050424031) and the Icelandic Research Fund for graduate students (080740008). The authors are most grateful to the participating children and their parents and to the two Master's students, Gudrun L. Gudmundsdottir and Ragnheidur Gudjonsdottir, for their assistance in collecting and analyzing data. They would also like to thank the staff at the Landspitali-The National University Hospital laboratories in Reykjavik, Iceland, for blood analysis.