Provincial Dietary Intake Study (PDIS): Energy and Macronutrient Intakes of Children in a Representative/Random Sample of 1–<10-Year-Old Children in Two Economically Active and Urbanized Provinces in South Africa

The double burden of malnutrition is still prevalent in South Africa, hence the importance of a dietary survey to identify risks of under- and over-nutrition. A multistage stratified cluster random sampling design was applied in two economically active provinces, Gauteng (GTG) (N = 733) and Western Cape (WC) (N = 593). Field workers completed questionnaires, and a 24 h recall with children taking part aged 1–<10-years (N = 1326). Important findings were that 71% and 74%, respectively, of 3–<6-year-olds and 6–<10-year-olds had an energy intake below the estimated energy requirement (EER), while 66% 1–<3-year-olds had intakes above the EER. The percentage of children with a total fat intake below recommended levels decreased as age increased ((51%, 40% and 5%) respectively, for the three age groups). Similarly, the percentage of those who had a total fat intake above the recommendation increased with increasing age (4%, 11% and 26%, respectively, for the three age groups). Saturated fat intake above 10%E was highest in the youngest and oldest children (33% and 32%, respectively). The percentage of children with a free sugars intake above 10%E was 47%, 48% and 52% respectively, and 98–99% had a fibre intake that was less than recommended. Overall, the diet was not healthy, with the main food items being very refined, and the diet being high in salty snacks and sugary items, and low in fruit, vegetables and legumes.

Does your household ever run out of money to buy food? 1a Has it happened in the past 30 days? 1b Has it happened 5 or more days in the past 30 days? 2.
Do you ever rely on a limited number of foods to feed your children because you are running out of money to buy food for a meal? 2a Has it happened in the past 30 days? 2b Has it happened 5 or more days in the past 30 days? 3.
Do you ever cut the size of meals or skip any because there is not enough food in the house? 3a Has it happened in the past 30 days? 3b. Has it happened 5 or more days in the past 30 days? 4.
Do you ever eat less than you should because there is not enough money for food? 4a. Has it happened in the past 30 days? 4b. Has it happened 5 or more days in the past 30 days?

5.
Do your children ever eat less than you feel they should because there is not enough money for food?
5a. Has it happened in the past 30 days? 5b. Has it happened 5 or more days in the past 30 days?

6.
Do your children ever say they are hungry because there is not enough food in the house?
6a. Has it happened in the past 30 days? 6b. 5 or more days in the past 30 days? 7.
Do you ever cut the size of your children's meals or do they ever skip meals because there is not enough money to buy food? 7a. Has it happened in the past 30 days? 7b. Has it happened 5 or more days in the past 30 days?

8.
Do any of your children ever go to bed hungry because there is not enough money to buy food?  The National Cancer Institute (NCI) method [1] that was developed to distinguish within-person from between-person variation, account for extreme intakes, including zero intake, and allow for adjustment for covariates and association analyses. The NCI method is used to adjust the measurement of the observed single 24-hour dietary intake data using data from the PDIS study, to establish usual intake, and thereby improve the validity of the results.
Two additional 24hour dietary recalls were completed on a subsample of 148 (2nd recall) and 146 (3rd recall) children in the sample. The last five EAs in each province, mainly for logistical reasons, were visited three times a week apart for this purpose. Parents of children also needed to indicate whether the 24-hour recall was less, same or more than the child's usual intake. The data obtained from the three 24-hour recalls of the subsample were used to adjust the observed distributions of the single 24-hour recall completed by the larger sample for the effects of random within-person variation.
Using the NCI method, the available 3-day 24-hour recalls for the subgroup were used to estimate within-person variance and remove it from the first 24-hour recall. The Balanced Repeated Replication (BRR) method [2] was used to do variance estimation with a Fay coefficient of 0.3. Two pseudo primary sampling units (PSU) were created per stratum by randomly selecting half of the PSU (or EA) in each stratum into one pseudo-PSU, and the rest in a second pseudo-PSU [2,3]]. Therefore 6 original strata were maintained with 12 pseudo-PSUs, two per stratum. Consequently, 8 BRR weights were created, taking the original sampling weights as well as the age and gender of each child in consideration When estimating usual intakes, covariates adjusted for in this study included province, type of residential area (urban formal, urban informal or rural), gender of the child and whether the intake of the 24-hour recall was less, the same or more than usual. The three age groups, namely 1 -<3-years, 3 -<6-years and 6 -<10-years were treated as subgroup options within the macros. The NCI method calculations should be interpreted at population level, and usual intakes for individuals within the group are not produced. The website accessed is: https://prevention.cancer.gov/research-groups/biometry/measurement-error-impact/softwaremeasurement-error, and the software selected are for estimating usual intake distribution, specifically for single regularly-consumed nutrients, and the percentage of energy intake from selected macronutrients. The