Elsevier

Nutrition

Volume 30, Issue 1, January 2014, Pages 55-60
Nutrition

Applied nutritional investigation
Which dietary diversity indicator is best to assess micronutrient adequacy in children 1 to 9 y?

https://doi.org/10.1016/j.nut.2013.06.002Get rights and content

Abstract

Objectives

The aim of this study was to determine the best dietary diversity indicator to measure dietary diversity and micronutrient adequacy in children.

Methods

A national representative cross-sectional survey of children ages 1 to 9 y (N = 2,200) was undertaken in all ethnic groups in South Africa. A 24-h recall was done with the mother or caregiver of each child. A dietary diversity score (DDS), the number of food groups consumed at least once in a period of 24 h, was calculated for each child in accordance with 6-, 9-, 13-, and 21-food group (G) indicators and compared with a mean adequacy ratio (MAR). The nutrient adequacy ratio (NAR) was calculated for 11 micronutrients by comparing the distributions of estimated intakes with the Estimated Average Requirements for that micronutrient. The MAR was the average of all NARs. Correlations were done between MAR and DDS and sensitivity and specificity calculated for each group indicator.

Results

Pearson’s correlations between food group indicators and MAR indicate that r values were all highly significant (P < 0.0001). There were no consistent or large differences found between the different group indicators although G13 and G21 appeared to be marginally better. Sensitivity and specificity values in the current study lay between DDS of 3 and 5, suggesting one of these as the best indication of (low) micronutrient adequacy.

Conclusions

Overall results seem to indicate that any of the four G indicators can be used in dietary assessment studies on children, with G13 and G21 being marginally better. A cut-off DDS of 4 and 5, respectively, appear best.

Introduction

The dietary diversity score (DDS), as measured by a quantitative number of food groups, has become a widely used method of determining variety in the diet, and by proxy, nutrient adequacy [1], [2], [3], [4], [5]. A low DDS also has been associated with low weight and stunted growth [6], [7], cardiovascular risk [8], [9], dyslipidemia [10], and higher probability of metabolic syndrome [11]. Numerous classification systems have evolved in determining dietary diversity adequacy with the number of food group indicators ranging from 6 to 21 groups (Table 1). Although the outcomes from using various food group indicators have been tested in adults [12], to our knowledge, this has not been the case in children.

One study [12] evaluated four different indicators in an attempt to establish the best indicator of micronutrient adequacy in adult women. All four food group indicators (G6, G9, G13, G21) tested were positively correlated with the mean probability of adequacy of micronutrients, even when controlling for energy intake [12]. However, their predictive strength differed among the five sites tested. In South Africa, one study evaluated the effectiveness of using a G9 diversity indicator in preschool children [7]; the resulting DDS of 4 was shown to be the best cut-off of mean adequacy ratio (MAR) of 11 micronutrients because it provided the best sensitivity and specificity [7].

However, it is unknown whether other group diversity indicators would work as well (or better) than the G9 used in children. Hence, the primary objective of this study was to test the use of different food group indicators, namely G6, G9, G13, and G21, on the micronutrient adequacy of children, in order to identify the best indicator to use in future dietary studies. To our knowledge, this is the first study to do this in children. There might be multiple benefits since the results from a single unquantified 24-h recall can be used to calculate DDS and the indicator that attains the best sensitivity and specificity with regard to the identification of micronutrient deficiencies can be used to identify children at risk when undertaking dietary analyses in future studies.

Section snippets

Source of nutrient data

The dietary database used in the current study was that of the National Food Consumption Survey (NFCS) [13], which took place in 1999. The NFCS population comprised children ages 1 to 8.9 y (12–108 mo) in South Africa and was a nationally representative sample (N = 2200, weighted for provincial representativeness). A detailed description of this process is given elsewhere [14], [15]. The NFCS collected data by means of a 24-h recall, and a dietary frequency; however, only the 24-h results were

Results

In this sample, 8.6% (95% confidence interval [CI], 7.4%–9.9%) of children had a weight for age < −2 SD of the median of the National Centre for Health Statistics (NCHS); a 19.4% (95% CI,17.5%–21.2%) had a height of age < −2 SD NCHS, and 3.3% (95% CI, 2.5%–4.1%) weight for height < −2 SD NCHS standards indicating that overall stunting was highly prevalent in this population. Prevalence for all three nutritional status parameters were significantly higher in rural than in urban areas, indicating

Discussion

In this study we set out to determine which food group indicator of dietary diversity would provide the best sensitivity and specificity to categorize children ages 1 to 9 y according to micronutrient adequacy of their diet. The results could guide future research on dietary studies in children. Among the children in this nationally representative study, 19.4% were stunted (chronic undernutrition), 8.6% were underweight (acute undernutrition), and 3.3% were wasted (severe acute undernutrition).

Conclusion

This study showed that all the food group indicators tested showed good correlations with MAR. When deciding on a DDS value to use as a cutoff point for micronutrient adequacy, the following can be used as a guideline: for G6, a DDS of 3 would provide the best compromise of both sensitivity and specificity; for indicators G9 and G13, a DDS of 4 is recommended; for G21 a DDS of 5. Overall, however, all the group indicators proved to be satisfactory in predicting MAR, with G13 and G21 being

Acknowledgments

We thank the Department of Health, UNICEF, Micronutrient Initiative, and USAID Micronutrient Program for funding. We also thank the directors of the National Food Consumption Survey and the Department of Health for their participation in the planning and execution of the study.

References (21)

There are more references available in the full text version of this article.

Cited by (66)

View all citing articles on Scopus

NPS was the main writer and was involved in all aspects of the study. JN conducted all statistical analyses and interpretation thereof. DL was the principal investigator of the original study. EM was a co-principal investigator of the original study and contributed to the writing. HSK contributed to the writing.

View full text