Vitamin A-fortified cooking oil reduces vitamin A deficiency in infants, young children and women: results from a programme evaluation in Indonesia

Abstract

Objective: To assess oil consumption, vitamin A intake and retinol status before and a year after the fortification of unbranded palm oil with retinyl palmitate.

Design: Pre-post evaluation between two surveys.

Setting: Twenty-four villages in West Java.

Subjects: Poor households were randomly sampled. Serum retinol (adjusted for subclinical infection) was analysed in cross-sectional samples of lactating mothers (baseline n 324/endline n 349), their infants aged 6-11 months (n 318/n 335) and children aged 12-59 months (n 469/477), and cohorts of children aged 5-9 years (n 186) and women aged 15-29 years (n 171), alongside food and oil consumption from dietary recall.

Results: Fortified oil improved vitamin A intakes, contributing on average 26 %, 40 %, 38 %, 29 % and 35 % of the daily Recommended Nutrient Intake for children aged 12-23 months, 24-59 months, 5-9 years, lactating and non-lactating women, respectively. Serum retinol was 2-19 % higher at endline than baseline (P<0·001 in infants aged 6-11 months, children aged 5-9 years, lactating and non-lactating women; non-significant in children aged 12-23 months; P=0·057 in children aged 24-59 months). Retinol in breast milk averaged 20·5 μg/dl at baseline and 32·5 μg/dl at endline (P<0·01). Deficiency prevalence (serum retinol <20 μg/dl) was 6·5-18 % across groups at baseline, and 0·6-6 % at endline (P≤0·011). In multivariate regressions adjusting for socio-economic differences, vitamin A intake from fortified oil predicted improved retinol status for children aged 6-59 months (P=0·003) and 5-9 years (P=0·03).

Conclusions: Although this evaluation without a comparison group cannot prove causality, retinyl contents in oil, Recommended Nutrient Intake contributions and relationships between vitamin intake and serum retinol provide strong plausibility of oil fortification impacting vitamin A status in Indonesian women and children.

Keywords: Children; Food fortification; Impact evaluation; Mothers; Vitamin A.

Fig. 1

Vitamin A content of cooking oil through the oil distribution chain in a pilot fortification of unbranded palm oil with vitamin A in West Java, Indonesia: (a) mean vitamin A content of oil samples, analysed immediately after collection, from one factory, two district-level distributors, twenty-four small stalls and sixty-four households*; (b) mean vitamin A content of samples from households in Ciamis () and Tasikmalaya () by month and year of collection and analysis†. *Numbers in parentheses denote number of oil samples collected and immediately analysed. Samples with vitamin A content below the iCheck assay’s lower detection limit of 3 mg retinol equivalents/kg (i.e. 10 IU/g) were included with value of 1 mg/kg (3·3 IU/g) in data analyses. A retinyl-palmitate level below the detection limit was not found in any of the factory or distributor samples, but it was found in three of the seventy-five household samples and in one of the fifty-four small-stall samples analysed immediately. †Over November 2011 to May 2012, the number of samples analysed per district per month ranged from 5 to 6 in Tasikmalaya and from 5 to 8 in Ciamis (mode 5), for a total of n 75. Over July–October 2011, a period for which the original data were lost and only aggregate results retained, similar numbers of samples were analysed immediately every month

Fig. 2

Serum (CRP/AGP-adjusted) retinol concentration and breast milk retinol concentrations at baseline (– – – –) and endline (——) in a pilot fortification of unbranded palm oil with vitamin A in West Java, Indonesia: (a) infants aged 6–11 months; (b) children aged 12–23 months; (c) children aged 24–59 months; (d) children aged 5–9 years; (e) lactating mothers; (f) non-lactating women aged 15–29 years; and (g) breast milk. CRP, C-reactive protein; AGP, α-glycoprotein

Fig. 3

Prevalence of vitamin A deficiency at baseline () and endline () of a pilot fortification of unbranded palm oil with vitamin A in West Java, Indonesia. Vitamin A deficiency, defined as serum retinol <20 μg/dl after adjustment for subclinical inflammation based on serum CRP (>5 mg/l) and AGP (>1 g/l), decreased significantly in all groups at endline (P<0·001 for all except non-lactating women aged 15–29 years, for whom P=0·011). CRP, C-reactive protein; AGP, α-glycoprotein