The main findings of the study are the differences between urban and rural areas in birth weight as well as in the subsequent growth, attained weight and length and growth velocity. For birth weight the differences between boys and girls were expected as was also the associations with the gestational age proxy. The latter is the variable with the strongest correlation to birth weight and is in turn related to subsequent attained weight and length.
The area variable in itself, urban vs. rural, is of no importance when other variables, with large differences between the areas, are introduced in Model C. Some of the added variables are not statistically significantly associated to birth weight or growth but obviously form an intricate pattern that "replaces" the area variable. This finding is the same in the analysis of birth weight and in the analysis of growth. Another common finding is that there are associations between growth and household assets and education, particularly for weight growth.
Growth velocity for weight differs between the areas for both child sexes. The length growth velocity is lower for rural girls. It shall be noted that all regression models have quite low values for the determination coefficient (R2) and that the largest part is contributed by the area and sex variables meaning that rather small fractions of the variation in birth weight and growth are explained by the associations with Gap, area and child sex differences and the social and economic variables.
The result from the present study is in accordance with results from previous studies in other countries [6, 7, 21]. Differences in growth of infants between urban and rural areas have been described in Peru in 1980. Height for age and weight for age of rural infants did not catch up to urban infants . Newer studies in China show that urban infants grow faster than rural infants [6, 7].
Socioeconomic conditions, nutrition of mothers during pregnancy, antenatal care, and increased maternal weight gain during pregnancy have been seen to be associated to the birth weight of the child [22–25]. Economic advantages, better education can lead to better nutrition for mothers and faster fetal weight gain. A Vietnamese study in 1996 found that 94% of rural farming women had insufficient food intake, compared to 40% for non- farming women . This situation has improved, but there can still be considerable differences in food intake between farming and non-farming women in Vietnam. The prevalence of anemia in women was higher in a rural area than in an urban in India . In Vietnam, no results on the prevalence of anemia in urban areas are available but a study in 2005 reported that in a rural area the prevalence among pregnant women was as high as 43.2% .
The rural mothers of the children in the present study attended antenatal care (ANC) later, had fewer visits and much less of specific medical services than in the urban mothers . Differences in antenatal care could be one factor behind the differences found in birth weight and infant growth. Specifically poor adherence to the guidelines for medical services can mean that conditions disadvantageous for growth are not detected.
Several conditions and factors have been shown as associated to poor growth of infants with nutrition as the most important [1, 29]. The nutritional status of under five children is proposed as a sensitive indicator of economic condition . Some studies therefore explain differences in child growth between rural and urban areas with differences in family income and general living conditions. Fewer children in the urban families might lead to better nutrition of each child [6, 7]. Parent's education has been demonstrated to be one of the main contributing factors for under five malnutrition in Bangladesh .
In Vietnam, the total fertility rate in the rural areas was higher than in the urban area  but the income per capita in urban areas was higher than in the rural [15, 16] Maternal education was also higher in the urban area than in the rural. Both economy and education might contribute to a better nutritional situation for infants in urban areas. The present study shows drastic differences in the educational and economic situation between the urban and rural mothers and households. There is also a tendency to smaller households in the urban area.
The differences in weight gain between rural and urban infants found in this study are established at an early age. One important factor may be differences in breastfeeding patterns, especially the duration of exclusive breastfeeding. The absolute differences in growth of infants between urban and rural areas increased with increasing ages. Use of different types of supplement food for infants in the two sites could explain this.
Infants in the urban area are likely to have easier access to child health care than rural infants. Some barriers to access child health care in rural areas in Vietnam, like distance and long travel times, do exist. Financial, sociocultural, language, ethnicity are other possible barriers together with lack of knowledge, awareness and inequalities in quality of health care .
The differences of length growth between the two sites were comparatively smaller at low ages, but increased in absolute terms during infancy. This result is in agreement with results of studies from China where urban children were taller than rural children at all ages from one to 12 months of age [6, 7]. One study found that the difference of growth in length of children between rural and urban areas is statistically significant only after six months and especially after 2 years of age .
Different standards for child growth have been published by various institutions and international organizations. Recently, the World Health Organization (WHO) launched growth standards in 2006. These were constructed to show child growth under ideal conditions . A study in Vietnam that assessed the growth of children by using the new WHO child growth standards as reference showed that deficient growth of infant is widespread in Vietnam . Another study in an urban area of Hanoi found that the growth of Vietnamese infants was also lagging behind the earlier used National Centre for Health Statistic reference population . The present results put urban boys and girls above the WHO standards and the rural children below for weight. For length again the rural curves are below the standard. This can be seen as an indication that genetic factors could not explain deviations in weight growth at a population level in Vietnamese infants. A detailed analysis of the relation between the present results and the WHO standards is beyond the scope of this paper but further analysis seems urgent, not at least to explore when early signs and warning of subsequent overweight can be detected.
Compared to results of a study in urban Hanoi in the 1990's, the birth weight and growth of infants in the present study are higher for both sites , indicating that the birth weight and growth of infants in both rural and urban areas of Hanoi have improved. There is, however, still a gap between the rural area and the urban area suggesting differences in child health care and nutrition.
One limitation of the study is the short follow-up time. One year is not enough to study if differences tend to decrease or increase as the children get older. The ambition for further research shall be to continue follow-up to at least 5 years to see if the rural children catch up with urban children or if the gaps are further widened. Also the exploration of overweight tendencies will require longer follow-up. Certain unavoidable differences between the study designs, data collection and administrative procedures might be seen as limitations. For example the two cadres of interviewers have different employment conditions. But the good training and the quality control have probably minimized this problem. The situation that there are unequal sample sizes in the two areas is not optimal for comparison.
The research was conducted in two sites within the capital of Vietnam. These areas are generally considered to have rather good socioeconomic conditions compared to the rest of country. Even so, the birth weights and growth of infants are higher in the urban area than in the rural area. This suggests that differences are likely to occur also in other, comparatively poorer, settings in Vietnam.