The mix of good nutritional attitude and poor nutritional knowledge is associated with adequate intake of vitamin A and iron from wild fruits and vegetables among rural households in Acholi subregion of Northern Uganda

Abstract Nutritional knowledge and attitude of a population greatly affect intake of essential micronutrients. Much of the understanding on the impact of nutritional knowledge and attitude on nutrient intake has been derived from studies conducted among affluent societies using commercial food products. However, information on the link between status of nutritional knowledge and attitude on intake of essential nutrients from wild fruits and vegetables among rural communities is largely lacking. This study examined the status of nutritional knowledge and attitude associated with intake of vitamin A and iron from wild fruits and vegetables among rural households in Acholi subregion of northern Uganda. A cross‐sectional study design making use of individual household questionnaire and proximate dietary recall of a calendar year consumption period involving 248 randomly selected rural households was applied. Results showed that households had good nutritional attitude but poor nutritional knowledge. Status of knowledge was dependent on nutrition training (p = 0.028; coefficient, β = 0.710) and age of the respondent (p = 0.044; β = −0.023), whereas status of attitude was predicted by age of the respondent (p = 0.014; β = 0.048). Consumption of wild fruits and vegetables contributed adequately to the dietary vitamin A and iron requirements of the households. The results have demonstrated that intake of vitamin A and iron from wild fruits and vegetables among rural households is associated with good nutritional attitude but poor nutritional knowledge. A concerted effort is necessary to improve household nutritional knowledge for better utilization of wild fruits and vegetables.


| INTRODUC TI ON
Globally, Sub-Saharan Africa has the highest proportion of food insecure people, with the prevalence of undernutrition standing at 23.8%. This level of undernutrition is more than twice the global undernutrition rate that stands at 11.3%. The situation is even worse for Uganda, for which the prevalence of undernutrition has been reported at 25% (FAO, IFAD, & WFP, 2014). In rural areas in Sub-Saharan Africa, household food security is largely dependent on own production and to a lesser extent through purchase from the market (Cooper, Dimes, Rao, Shapiro, & Twomlow, 2006). However, during lean seasons or in case of crop/livestock failure, the meager food available from own production cannot support household food needs. Thus, the use of wild fruits and vegetables becomes paramount among other food insecurity coping strategies (Okori, Obua, & Baryamureeba, 2009;Pilgrim, Cullen, Smith, & Pretty, 2007). In the context of this study, wild fruits and vegetables refer to edible plants that are not managed under routine farming systems. On the other hand, in extreme circumstances such as prolonged drought or disaster situations, households rely on food donation from the government or other agencies such as the World Food Programme (Tusiime, Renard, & Smets, 2013).
Vitamin A and iron deficiencies are important nutritional constrains in many countries in Sub-Saharan Africa (Camaschella, 2015;UNICEF, 2018;World Health Organization, 2009). In Uganda, deficiency of vitamin A and iron among vulnerable groups especially children and pregnant mothers is still very high (Uganda Bureau of Statistics (UBOS) and ICF International Inc, 2017). This suggests that households are unable to meet vitamin A and iron needs from own production and/or through food purchase. On the other hand, as is the case with several other countries in Sub-Saharan Africa, there are areas in Uganda such as the Acholi subregion where wild fruits and vegetables exist and are routinely consumed (Acipa, Kamatenesi, & Oryem-Origa, 2013;Oryema, 2014). Wild fruits and vegetables have been reported to be rich in micronutrients such as vitamin A and iron (Acipa et al., 2013;Oryema, 2014). Nevertheless, it is important to appreciate that information available on the utilization of wild fruits and vegetables for household nutrition among rural communities is largely qualitative (Bharucha & Pretty, 2010;Loki & Ndyomugyenyi, 2016a,b;Pardo-de-santayana et al., 2007). However, the actual contribution of those wild food resources in terms of the Recommended Dietary Allowance (RDA) for critical nutrients such as vitamin A and iron has largely remained unknown. Therefore, the significance of wild fruits and vegetables to nutrient needs of households that utilize them is still vaguely understood.
Empirical studies have shown that availability and access to food are essential, but they are not the only factors that ensure household food and nutrition security (Brown & Funk, 2015;Cook, Reilly, Derosa, Rohrbach, & Spruijt-metz, 2014;Ozor, Umunnakwe, & Acheampong, 2014). Nutritional knowledge and attitude have been shown to influence dietary practices and consequently the quantity and quality of nutrients derived from a given food item (Hoogenboom, Morris, Morris, & Schaefer, 2009;Sichert-hellert et al., 2011). Factors associated with nutritional knowledge and attitude such as choice of food, prior knowledge of expected nutritional benefits, and the frequency of consumption have been shown to influence household food and nutrition security (Carrillo, Varela, & Fiszman, 2012). Despite the well-known significance of appropriate nutritional knowledge and attitude on food consumption behavior, it is important to appreciate that much of the understanding has been derived from studies conducted among affluent societies using commercial food products (Hoogenboom et al., 2009;Scully, Dixon, & Wakefield, 2008;Sichert-hellert et al., 2011). However, there is very limited information on the status of nutritional knowledge and attitude associated with consumption of wild fruits and vegetables among rural households such as those in Acholi subregion of Uganda. In other words, it is largely unknown whether consumption of wild fruits and vegetables by rural households is backed by appropriate nutritional knowledge and attitude to enable the community to use wild food resources to derive critical nutrients such as vitamin A and iron. Making use of the Acholi subregion of Uganda as a study area, the objective of this study was therefore to assess the status of nutritional knowledge and attitude associated with intake of vitamin A and iron from wild fruits and vegetables among rural households.

| Study design and study area
A cross-sectional study design employing household survey was used. The study was conducted in Amuru and Gulu districts of Acholi subregion of Uganda ( Figure 1). The choice of the two districts was guided by the fact that (a) previous studies that documented consumption of wild fruits and vegetables were conducted in them and (b) communities that live in them are prone to acute food insecurity (Uganda IPC Technical Working Group, 2013). Therefore, as a coping strategy to mitigate food insecurity, communities routinely include wild fruits and vegetables in household diet (Okori et al., 2009;Oryema, Oryem-origa, & Nanna, 2013). Gulu and Amuru districts are largely inhabited by the Acholi tribal community that speak the Acholi language. Communities in the two districts practice mainly subsistence farming as a major source of livelihood although a few large-scale commercial farmers have inhabited farmlands of Amuru district in the recent past. Children (0-17 years) and youth (18-30 years) constitute the majority of the population (77%). The average household size is about 5.0 for both Gulu and Amuru districts.
The majority of the households in Amuru (91%) and Gulu (66%) are in rural areas, and at least 51% of the population in the two districts are females (UBOS, 2016).

| Study population and sampling framework
The study population comprised of rural households living in Amuru and Gulu districts ( Figure 1). Population size of rural households inhabiting the two districts is estimated to be 472, 127 based on projection from the latest national census (UBOS, 2014). Previous statistics put the proportion of households that experienced stressed food insecurity situation in Acholi subregion at 16% (Uganda IPC Technical Working Group, 2013). On the basis of this statistic and allowing for an error of 5% (p = 0.05), the sample size for the study was estimated to be 207 households according to standard formula for nutritional epidemiology previously used by Kasiulevičius, Šapoka, and Filipavičiūtė (2006). Considering an attrition rate of 20%, the sample size was adjusted upwards by the same percentage and the final sample size used was 248 households. This sample size (248) is more than what had been used in similar studies conducted before (Agea, Obua, Waiswa, Okia, & Okullo, 2010;Redzic, Barudanovic, & Pilipovic, 2010).
Following determination of the sample size, a multistage sampling framework was used to select participating households. From each of the two districts (Amuru and Gulu), two subcounties were randomly selected (Gulu district: Patiko and Bobi subcounties;  Before the actual data collection, the questionnaire was pretested and adjustments made before the final version could be produced. Questions that were difficult to interpret by the respondents and did not provide valid results were adjusted accordingly. The study made use of interviewers that had undergone prior training in individual household method (Petty & Ellis, 2015) and had experience in nutrition survey. To ensure proper data collection on consumption, the interviewers were trained on questionnaire administration and on techniques of translating local measurements of assessing food quantity to standard measures (grams). Because not all parts of fruits and vegetables are consumed, the interviewers were also trained on how to adjust the quantity measured for edible portion. Actual data collection took place between November and December 2015. The main respondents were women. This is because in a rural setting as is the case in the study area, women are responsible for collection and preparation of food in households (Hyder et al., 2005). In situations where women were unavailable (e.g., the widowed, divorced, the woman being absent for some other reasons), men were interviewed.

| Status of nutritional attitude and knowledge
Data were coded and analyzed using Statistical Package for Social Scientist (SPSS) version 20.0. Before actual analysis, negatively worded questions were reverse-coded. Descriptive statistics (frequencies and percentages) were used to analyze socio-demographic characteristics. Knowledge was scored on a 2-point scale (0 score for "wrong" or "don't know" response and 1 for correct response). Knowledge scores were then expressed as a percentage of total score (8). Attitude was scored on a 3point scale (disagree, not sure, and agree). The scores ranged from zero for the most negative (disagree) to 2 for the most positive (agree). Attitude scores were expressed as a percentage of the total points-26 (Masuku & Lan, 2014). Scores below 50% and ≤ 57% were considered as poor for knowledge (Bas, Ersun, & Kıvanc, 2006)

| Contribution of wild fruits and vegetables to vitamin A and iron requirements
To determine the contribution of wild fruits and vegetables

| Socio-demographic characteristics
Data on socio-demographic characteristics of the respondents that participated in the study are presented in

| The status of nutritional knowledge and attitude
The mean score for nutritional knowledge and attitude was 48% (3.84) and 60% (15.63), respectively ( TA B L E 1 Vitamin A and iron content of wild fruits and vegetables used in the study Data on the distribution of responses to questions testing nutritional attitude are presented in Table 5. Majority of the respondents (82%) considered wild fruits and vegetables to be safe for consumption notwithstanding the fact that the same proportion considered wild fruits and vegetables not to be nutritious. Nevertheless, only less than 10% of the respondents did not like the taste of wild fruits and vegetables. About 72% of the respondents considered wild fruits and vegetables not to be expensive although 30% of the respondents anticipated consumption of wild fruits and vegetables to be unsustainable. Interestingly, a greater proportion of the respondents (80%) was ready to consume wild fruits and vegetables as supplements for domestically produced foods. On the other hand, 56% of the respondents were willing to consume wild fruits and vegetables as a substitute for those that are domestically cultivated. Last but not least, there was no association between nutritional knowledge and attitude (p = 0.115).

| Socio-demographic predictors of nutritional knowledge and attitude
The results of binary logistic regression analysis executed to determine socio-demographic predictors of nutritional knowledge and attitude are presented in Table 6. Nutritional knowledge was determined by attendance of nutritional training (p = 0.028) and age of the respondent (p = 0.044). Attendance of nutrition training increased nutritional knowledge by 71% while an increase in age by 1 year decreased nutritional knowledge of the respondent by 2%.
Nutritional attitude was only influenced by age of the respondent (p = 0.014). An increase in age by one year improved the status of nutritional attitude by 5%.

| Contribution of wild fruits and vegetables to household intake of vitamin A and iron
The contribution of wild fruits and vegetables to the pooled household requirement for vitamin A and iron is presented in Figures 2 and   3, respectively. In the case of vitamin A, all the households met the RDA while about 70% of them derived the micronutrient more than the RDA from wild fruits and vegetables. Similarly, as was the case for vitamin A, most of the households achieved sufficient dietary intake of iron from wild fruits and vegetables. In fact, more than half of the households derived more than twice the RDA for iron from those wild food resources. TA B L E 5 Distribution of responses to questions testing specific aspects of nutritional attitude of nutritional knowledge regarding wild fruits and vegetables could also be explained by lack of information on nutritional characteristics of wild food resources available for use by the community in the study area. A concerted effort is therefore necessary to foster comprehensive documentation of nutritional characteristics of wild food resources. Such information would be of great use in community nutrition education.

| D ISCUSS I ON
The poor level of nutrition knowledge observed in this study is in contrast with findings from other nutrition studies. For instance, Anand and Puri (2013) and Das and Mukherjee (2014)  It was interesting to note that attitude of the respondents toward consumption of wild fruits and vegetables was good. This observation could be explained by the fact that wild fruits and vegetables contribute significantly to household food security in the study area (Loki & Ndyomugyenyi, 2016b;Oryema et al., 2013). In addition, availability of those wild food resources within the vicinity of the study participants (rural areas) could also explain households' good attitude toward them (Msuya, Kideghesho, & Mosha, 2010;Pilgrim et al., 2007). This observation contrasts sharply with findings from other studies conducted in Spain and Ethiopia where consumer attitude toward consumption of wild fruits and vegetables was poor (Giday, Asfaw, Elmqvist, & Woldu, 2003;Pardo-de-santayana, Tardio, & Morales, 2005). This disparity illustrates the differences in the nature of attitude that exist among communities between different geographic locations. By implication therefore, it becomes apparent with regard to wild fruits and vegetables that information on attitude obtained from a given geographic location may not be applicable in another location, thus providing further justification for the current study. Culture is an important factor that modulates consumer attitude toward a particular food type (Kruger & Gericke, 2002;Thurber et al., 2016). It is therefore plausible that cultural differences between communities in Acholi subregion of Uganda and those in Ethiopia and Spain where the respective studies were conducted could explain the disparity. A peculiar observation from this study is that nutritional attitude was only determined by age of the respondent and in a positive way. Improvement in household nutritional attitude with age could be explained by the mere-exposure effect theory (Jones & Kervin, 2010;Kumar, Onufrak, Zytnick, Kingsley, & Park, 2014;Scully et al., 2008). This is a psychological phenomenon by which people tend to develop a preference for things merely because they are familiar with them. In the context of this study, the more the number of years the respondent used wild fruits and vegetables, the more he/she got used to wild food resources and hence the better the attitude toward them.
The fact that adequate nutrient intake was associated with good nutritional attitude but poor knowledge challenges the common belief that good knowledge manifests into good attitude (Anand and Puri (2013). The observed results could be a consequence of limited nutrition education received by the households. This is evidenced by the fact that up to 63% of the households never attended any nutrition education while only 24 and 14% received nutrition education once and twice, respectively. Multiple attendance of nutrition education has been shown to improve nutrition knowledge (Eyles & Mhurchu, 2011;Pomerleau, Lock, Knai, & McKee, 2005 Typical example is indiscriminate tree-cutting for charcoal production (Akena, 2012). From economic point of view, national governments do spend a lot of resources in ameliorating vitamin A and iron deficiency through medical supplementation in areas where such micronutrients are endemically deficient. Therefore, strategic interventions are required to protect natural ecosystems within communities that are endowed with wild fruits and vegetables. Otherwise, future sustainability of wild food resources in supplying rural communities with critical nutrients such as vitamin A and iron will be compromised.
There are three limitations inherent in this study. First, substantial information on the content of vitamin A and iron for wild food resources studied was derived from literature. Considering the fact that nutritional composition of fruits and vegetables depends on geographic location (Okello, 2010;Okullu et al., 2010) and variety (Howard et al., 2000), future studies should consider using data from real-time laboratory analysis. Secondly, it is known that fruits and vegetables generally contain antinutritional factors that lower bioavailability of nutrients, and in the context of this study, iron in particular (Rathod & Valvi, 2011;Umaru, Adamu, Dahiru, & Nadro, 2007). This implies that the contribution levels reported in this study are to some extent theoretical. Future studies should consider contribution based on iron which is biologically available. Thirdly, the assumption that food distribution in households is in accordance with individual household member food needs may not hold. Thus, nutrient adequacy determined based on pooled household estimate may not reflect intrahoused nutrient adequacy.

| CON CLUS ION
This study has demonstrated that intake of vitamin A and iron from wild fruits and vegetables among rural households is associated with good nutritional attitude but poor nutritional knowledge. Despite the mix of good nutritional attitude and poor nutritional knowledge, and taking into account the limitations of the study, theoretically, households derived more than adequate amount of vitamin A and iron from the wild fruits and vegetables studied. Attitude but not knowledge seems to be a key impetus that enabled rural households to meet vitamin A and iron requirements through consumption of wild fruits and vegetables.

ACK N OWLED G EM ENTS
The authors are thankful to the district and local leaders, research assistants, and rural communities in Acholi subregion for their active participation in the study. This study was funded by Regional Universities Forum for Capacity Building in Agriculture (RUFORUM) (grant number: RU 2014 NG 13). RUFORUM had no role in the design, analysis, or writing of this article.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.