Undernutrition among Bedouin Arab children: a follow-up of the Bedouin Infant Feeding Study

Abstract

After 10 y of urban settlement, 680 Bedouin Arab children, who had had anthropometric assessment from birth (1981-1982) through early childhood, were reassessed in 1991-1992 to compare the rates of stunting in early and later childhood as well as to describe the factors influencing current height-for-age. Stunting had dropped from 32.7% at 18 mo to 7.2% at 10 y in the 1981 birth cohort and dropped from 17.5% at 9 mo to 8.2% at 9 y in the 1982 birth cohort. Based on a multiple-linear-regression analysis, height in early childhood and maternal height were statistically significantly and positively associated with current mean height-for-age in both cohorts. In the 1982 cohort socioeconomic status in early childhood was positively and current family size was negatively and significantly associated with current mean height-for-age. Thus, conditions that were present in early childhood had the largest influence on current height. In 1992, 10% and 6% of the infant siblings of the 1981 and 1982 cohorts, respectively, were stunted compared with 17% and 1% of the siblings aged 1-2 y of the respective cohorts. Therefore, the high rates of early childhood stunting in 1981-1982 appeared to be a birth cohort-specific phenomenon.

PIP: During 1991-1992 in Israel, researchers re-examined the anthropometric assessment of Bedouin children 9-10 years old, living in the Negev, whose anthropometric assessment was initially examined when they were 6-18 months old, to compare the rates of stunting in early and later childhood and to determine the factors affecting current height-for-age. For the 1981 birth cohort, stunting fell from 32.7% at 18 months to 7.2% at 10 years. The corresponding figures for the 1982 birth cohort were 17.5% at 9 months and 8.2% at 9 years. Thus, despite the social change and urban settlement the children experienced in the 10 year span, they still had some catch-up growth. Stunting among the siblings less than 2 years old of the 1991-1992 cohorts was much less common than stunting of the index children in early childhood in 1981 (17% vs. 44%), but not so in 1982 (14% vs. 11%). In 1992, the prevalence of stunting stood at 10% and 6% of siblings less than one year old of the 1981 and 1982 cohorts, respectively, while it was 17% and 1% for siblings 1-2 years old of the 1981 and 1982 cohorts, respectively. Maternal height and index child's height-for-age in early childhood were positively associated with current height-for-age in the 1981 cohort (p = 0.0001). They explained 33% of the variation in height-for-age. In the 1981 cohort, neither socioeconomic status (SES) in early childhood nor family size were associated with current height-for-age. Yet, in the 1982 cohort, both SES in early childhood and family size were associated with current height-for-age, SES positively (p = 0.04) and family size negatively (p = 0.03). Maternal height and height-for-age in early childhood were positively associated with current height-for-age (p = 0.002 and 0.0001, respectively). These 4 variables accounted for 34% of the variation in height-for-age in the 1982 cohort. These findings indicate that factors or conditions present in early childhood had the greatest impact on current height-for-age.