Liver is widely eaten by preschool children in the Northern Cape province of South Africa: Implications for routine vitamin A supplementation

Abstract Previous research has demonstrated a virtual absence of vitamin A deficiency and adequacy of vitamin A intake through consumption of liver in preschool children of a community in the Northern Cape province of South Africa where sheep farming is common, and liver, an exceptionally rich source of vitamin A, is frequently eaten. Only 60–75 g of liver per month is needed to meet the vitamin A requirement of preschool children. Because this may have implications for routine vitamin A supplementation, and because liver consumption for the rest of the province is unknown, the study aim was to establish the prevalence and frequency of liver intake in a provincial‐wide survey. An unquantified liver‐specific food frequency questionnaire, covering a period of 1 month, complemented by a 1‐year recall, was administered to mothers of 2‐ to 5‐year‐old children (n = 2,864) attending primary health care facilities in all five districts and 26 subdistricts. A total of 86% of children were reported to eat liver, which was eaten in all districts by at least 80% of children. The overall median frequency of liver intake was 1.0 [25th, 75th percentiles: 0.5, 3.0] times per month and ranged from 1.0 [0.3, 2.0] to 2.0 [1.0, 4.0] for the various districts. Based on a previously reported portion size of 66 g, these results suggest vitamin A dietary adequacy in all districts and possibly also vitamin A intake exceeding the Tolerable Upper Intake Level in some children. Routine vitamin A supplementation in this province may not be necessary and should be reconsidered.


| INTRODUCTION
Vitamin A supplementation is known to reduce morbidity and mortality in vitamin A-deficient children (Beaton et al., 1993;Imdad, Mayo-Wilson, Herzer, & Bhutta, 2017), and the WHO recommends periodic high-dose vitamin A supplementation for children younger than 5 years living in areas where vitamin A deficiency is a public health problem (World Health Organization, 2011). A national vitamin A supplementation programme, which comprises a curative and a routine component, has been in operation in South Africa since 2002 (Department of Health, 2004). Routine supplementation, even though implementation may vary, targets 6-to 59-month children attending primary health care facilities every 6 months. South Africa is a diverse country with varied eating habits, and routine vitamin A supplementation may not be appropriate for all areas. In the Northern Cape province, for example, virtual absence of vitamin A deficiency (<5% with serum retinol <0.70 μmol/L) was reported in preschool children from a low socio-economic community in the Hantam area, a municipal subdistrict of the Namakwa district, where liver is frequently eaten (van Stuijvenberg, Schoeman, Lombard, & Dhansay, 2012). This was despite high levels of stunting and underweight in the area (van Stuijvenberg et al., 2012) and a national prevalence of vitamin A deficiency of 43% (Shisana et al., 2013).
Sheep farming is the main agricultural activity in the region, resulting in liver being readily available and affordable and thus frequently consumed (van Stuijvenberg et al., 2012).
Liver is an exceptionally rich source of preformed vitamin A (South African Food Database System [SAFOODS], 2016) and can significantly contribute to the vitamin A intake of young children. A dietary intake assessment that quantified vitamin A intake from sheep liver in the same community showed that liver alone provided enough vitamin A to meet the vitamin A requirement of the 2-to 5-year-old children in this community, even if eaten only once per month (Nel et al., 2014). It is worth noting that in 15% of the children, the Tolerable Upper Intake Level (UL) for vitamin A (Institute of Medicine, 2001) was exceeded through the consumption of liver alone (Nel et al., 2014). Although vitamin A is necessary for normal growth and immune function (Beaton et al., 1993;Sommer & West, 1996), vitamin A will accumulate in the liver if consumed in excess of requirements and may be harmful (Penniston & Tanumihardjo, 2006).
The findings on absence of vitamin A deficiency and vitamin A dietary adequacy through the intake of liver are limited to a particular subdistrict in the Northern Cape province. However, it is likely that liver consumption is more widespread because sheep farming occurs in most of the province. Similar results were found in a community in another district of the Northern Cape, ±500 km east of the study area (Faber et al., 2015). Liver consumption patterns for the rest of the province are, however, not known. Should liver be consumed to the same extent in all of the Northern Cape province, routine vitamin A supplementation in this province may not be necessary. A provincial-wide epidemiological survey, which aimed to establish the prevalence and frequency of liver consumption amongst preschool children in all districts, was thus undertaken.

| Study population and design
This is an epidemiological survey of 2-to 5-year-old children attending primary health care facilities in the Northern Cape province. The Northern Cape province is the largest of South Africa's nine provinces and covers approximately one third (30.5%) of the country's land area.
Yet it is the smallest in terms of population size, encompassing only 2.1% of the total population (Statistics South Africa, 2018). It is characterized by arid conditions with sheep farming being the main agricultural activity. There are five districts with a total of 25 subdistricts (26 at the time of the study). We aimed to include a convenience sample of at least 100 children per subdistrict so that each district would be adequately represented. This was also to allow for meaningful results at subdistrict level, should this information be needed at a later stage. The numbers surveyed per district and subdistrict, as well as the municipalities/towns covered, are in Appendix A.

Dieticians and nutritionists working for the Northern Cape
Department of Health were trained by members of the research team to collect the data from the child's mother or caregiver by means of a rapid interviewer-administered questionnaire. Data were collected as part of normal day-to-day activities undertaken by dieticians and nutritionists at primary health care facilities, and the questionnaire took less than 10 min to complete. Anthropometric measurements, which are part of routine measurements during clinic visits, were also obtained.

Key Messages
• Liver is a rich source of vitamin A, and only 60-75 g liver per month is needed to meet the vitamin A requirement of athe preschool child.
• Liver was found to be eaten in all five districts of the Northern Cape province of South Africa.
• The results suggest dietary adequacy of vitamin A in all districts, and possibly also vitamin A intakes exceeding the tolerable upper intake level (UL) in some children.
• Routine high-dose vitamin A supplementation may not be necessary in this province and needs to be reviewed.

| Questionnaire data
An unquantified liver-specific food frequency questionnaire, based on a quantified liver frequency questionnaire used previously (Nel et al., 2014) and also more recently (van Stuijvenberg et al., 2019) to assess liver intake in the Namakwa district (Hantam subdistrict), was employed in this survey. Information obtained on liver intake-at both household level and by the child-included a 1-month recall, complemented by a 1-year recall for those who did not eat liver during the last month. In this way, a liver intake of less than once a month, which could still make a significant contribution to vitamin A intake, was also captured. For example, liver, eaten only once in 2 or 3 months, was captured as 0.5 or 0.33 times per month.
In addition, data were collected on the age at which liver was introduced into the child's diet and the type of liver (i.e., sheep liver, beef liver, and chicken liver) mostly eaten. Background information was obtained on the mother's educational level, the mother's awareness of the vitamin A supplementation programme, vitamin A supplementation history, current and past breastfeeding, and age at which solid foods were introduced.

| Anthropometric status
Height was measured to the nearest 0.1 cm using a portable height measure; the same model from the same supplier was used in all 26 subdistricts (SECA 214 Leicester, Invicta Plastics). Weight was T A B L E 1 Background characteristics of the study population

| Statistical analysis
Data were captured in Excel and analysed using the IBM SPSS statistical software package, Version 25.0. Continuous data that were normally distributed were expressed as mean (standard deviation) and when not normally distributed as median [25th and 75th percentiles]. Frequency of liver intake, presented as the median number of times per month, was given for the province as a whole and at district level. Categorical data were expressed as percentages. The two-sided Mann-Whitney U test was used to compare the difference in liver intake between those whose mothers had less than 12 years of schooling and those whose mothers had 12 years or more of schooling. Pvalues < .05 were considered statistically significant.

| RESULTS
Data were collected on 2,864 children. Background characteristics of the study population are shown in Table 1. Biological mothers represented 73.3% of the respondents, and the rest were either the caregiver or guardian. Results for the group as a whole did not differ significantly when only the biological mother as respondent was included in the analyses (data not shown). The term "mother" will therefore be used throughout the text. The majority of the children (88%) were between 2 and 5 years (60 months) old. Stunting, underweight, and wasting were prevalent in 28%, 14.5%, and 5.4% of children, respectively. Only 25% of the mothers had 12 or more years of schooling. Almost all children (88%) had been breastfed in the past or were breastfed at the time of the survey; breastfeeding was continued to a median age of 13.7 months.
Liver consumption data for the province and per district are shown in  (Faber et al., 2015;Nel et al., 2014).
Liver intake in this survey was used as a proxy for vitamin A intake and hence vitamin A adequacy. The overall median frequency of liver consumption was once a month, ranging between once and twice per month for the different districts. We did not measure portion size in this study, but a previous quantified dietary intake study in the Namakwa district showed an average portion size of 66 g amongst 24-to 59month-old children (Nel et al., 2014). Based on the latter portion, the fre- Too much vitamin A is toxic, and its adverse effects in adults are well documented (Penniston & Tanumihardjo, 2006). However, there is emerging evidence that hypervitaminosis A may interfere with bone formation in young children, even before it reaches toxicity levels. In Zambian children, who had a high prevalence of hypervitaminosis A, bone formation markers improved when dietary preformed vitamin A was reduced (Tanumihardjo, Gannon, Kaliwile, Chileshe, & Binkley, 2019), and in South African preschool children, total liver reserves, as measured by RID, were found to be negatively associated with both height-for-age and weight-for-age, as well as with bone turnover markers (Tanumihardjo & van Stuijvenberg, unpublished data). Preschool children are at a crucial stage in their development and the potential benefits, and risks of any intervention that could interfere with growth should be carefully weighed before being implemented.
It is worth noting that the Northern Cape province has amongst the highest levels of stunting in the country (Labadarios et al., 2000).

Our findings have important implications for the routine vitamin
A supplementation programme in South Africa, in that vitamin A supplementation may not be necessary in the Northern Cape province.
The proportion of preschool children in the Northern Cape province of South Africa who eat liver per age category (n = 2,864). The smaller numbers in the age categories <24 and >60 months are because the study focussed on children 24 to 60 months also hypervitaminosis A in one of the communities (van Stuijvenberg et al., 2019). Furthermore, the present study shows that liver is eaten at a frequency sufficient to ensure adequate intake of vitamin A through the diet in all districts of this province.
The strength of this study is the large number of children surveyed (n = 2,864), and that all districts and subdistricts of the Northern Cape province were represented. In nutrition surveys previously undertaken in South Africa, the Northern Cape province was greatly undersampled.
Notably, in terms of vitamin A status (serum retinol), the number of children included in the South African National Health and Nutrition Examination Survey (Shisana et al., 2013), the National Food Consumption Survey Fortification Baseline survey (Department of Health, 2007), and the South African Vitamin A Consultative Group survey (Labadarios et al., 1995)  were available for comparison. However, the SADHS 2016 did confirm that liver intake in the rest of the country is not as common as the reported 86% for the Northern Cape province in the current study.
Liver vitamin A concentrations, as assessed by RID, is the best indicator, apart from liver biopsy, for measuring vitamin A status (Tanumihardjo et al., 2016). Unlike serum retinol, which cannot detect vitamin A excess, the RID test is able to measure total body stores of vitamin A over the vitamin A status continuum, even in the toxic range. However, because it would be logistically and financially impractical to perform the RID test at the scale of the current study, liver intake was used as a surrogate for vitamin A intake and consequently vitamin A adequacy. Liver intake was assessed by a short liver frequency questionnaire administered by local dieticians as part of their daily activities at the primary health care facilities. This allowed for vitamin A dietary adequacy to be assessed in a rapid, efficient, and inexpensive way and for the whole of the province to be covered. In our recent study, which showed hypervitaminosis A, frequency of liver intake was found to correlate significantly with total liver concentrations in the children (van Stuijvenberg et al., 2019).
The overall lack of awareness amongst mothers in the province of the vitamin A supplementation programme is a concern. Around 90% of mothers were unaware of the content, purpose, and the frequency of A limitation of our study is that portion size was not determined.
We assumed a portion size of 66 g, which is based on results of a previously published cross-sectional study in 24-to 59-month-old children from the Hantam subdistrict of the Namakwa district (Nel et al., 2014), an area typical of the Northern Cape province. However, even if the portion size in the current survey had been smaller, liver would still make a significant contribution to vitamin A intake. It is worth noting that in the recent study where hypervitaminosis A was demon-  (Labadarios et al., 1995).
Because we were limited in the length of our questionnaire, maternal education was used as a proxy for socio-economic status.
Liver intake in the children was generally higher where the mothers had less than 12 years of schooling, as opposed to where mothers had 12 or more years of schooling. This is similar to what was previously found in the Namakwa district, where liver intake was inversely related to maternal education (van Stuijvenberg et al., 2012), as well as to skilled employment and household income (Nel et al., 2014).
Sheep liver is available at both formal and informal outlets in the community and cheaper than meat, which results in the poor buying liver rather than meat. This suggests that the socio-economically more vulnerable may be less likely to be vitamin A deficient but, at the same time, more likely to develop hypervitaminosis A.

| CONCLUSION
This study showed adequacy of vitamin A through the intake of liver amongst preschool children in all five districts of the Northern Cape province of South Africa. The median frequency of liver intake per month was higher than the frequency in a recent study in the Namakwa district, where hypervitaminosis A was demonstrated. This indicates a risk for hypervitaminosis A in the other districts of this province as well.
Against this background and the fact that coverage of vitamin A supplementation in the province is low anyway, we recommend that routine vitamin A supplementation in the Northern Cape province be seriously reconsidered. Resources spent on an unneeded intervention should rather be used to address other nutritional problems, such as the high prevalence of stunting and underweight.

ACKNOWLEDGMENTS
We wish to thank the Northern Cape Department of Health for providing the infrastructure to conduct the survey and the dieticians and nutritionists for collecting the data as part of their day-to-day activities at the various clinics. The survey was funded by the South African Medical Research Council; no external funding was received.

CONFLICTS OF INTEREST
The authors declare no conflict of interest.