Food beliefs and practices among the Kalenjin pregnant women in rural Uasin Gishu County, Kenya

Background: Understanding food beliefs and practices is critical to the development of dietary recommendations, nutritional programmes, and educational messages. This study aimed to understand the pregnancy food beliefs and practices and the underlying reasons for these among the contemporary rural Kalenjin communities of Uasin Gishu County, Kenya. Methods: Through semi-structured interviews, data was collected from 154 pregnant and post-natal Kalenjin women about restricted and recommended foods, and why they are restricted or recommended during pregnancy. Respondents were purposively selected (based on diversity) from those attending Maternal and Child Health (MCH) care in 23 rural public health facilities. Key informant interviews (n = 9) with traditional Birth Attendants (TBA) who were also herbalists, community health workers, and nursing officers in charge of MCH were also conducted. Quantitative data was analysed using SPSS software. Data from respondents who gave consent to be tape recorded (n = 42) was transcribed and qualitatively analysed using MAXQDA software. Results: The restriction of animal organs specifically the tongue, heart, udder and male reproductive organs, meat and eggs, and the recommendation of traditional green vegetables and milk was reported by more than 60% of the respondents. Recommendation of fruits, traditional herbs, ugali (a dish made of maize flour, millet flour, or Sorghum flour, sometimes mixed with cassava flour), porridge and liver, and restriction of avocadoes and oily food were reported by more than 20% of the respondents. The reasons for observing these dietary precautions were mainly fears of: big foetuses, less blood, lack of strength during birth, miscarriages or stillbirths, and maternal deaths as well as child’s colic and poor skin conditions after birth. Conclusion: Pregnancy food beliefs were widely known and practised mainly to protect the health of the mother and child, and ensuring successful pregnancy outcome. Given the deep-rooted nature of the beliefs, it is advisable that when nutritious foods are restricted, nutritional interventions should rather search for alternative sources of nutrition which are available and considered to be appropriate for pregnancy. On the other hand, nutritional advice that does not address these health concerns and assumptions that underlie successful pregnancy and delivery is unlikely to be effective.

The research findings of this study were interpreted and understood through the symbolic interactionist perspective as expounded by Messer [30]. Symbolic interaction theory (SIT) is a frame of reference for how people act toward things based on the meaning those things have for them, and these meanings are derived from social interaction and modified through interpretation [31]. Messer (1989) classified food according to a number of symbolic dimensions which are culturally constructed in the process of social interaction and this influences human food selection. Having no biological base, these social constructs of nutrition are based on what people believe to be true about food and not just on what is objectively true. So, the symbolic meaning of food overrides those actual biological facts regarding diet and health. These may be the dimensions most prominent in the food proscriptions and prescriptions of nutritionally “vulnerable groups” such as infants, children, and pregnant and lactating women. Messer [30] found that while people shared the same general structure and rules for classification, they did not necessarily judge all items equivalently, given the differences in individual experience. Also, individuals differed in the extent to which they had acquired information and applied it to their own diets and health. Understanding these food symbolic dimensions therefore can greatly aid in interpretation of food habits that might be beneficial or harmful to a particular population or subpopulation. Messer [30] classified food into eight cultural symbolic dimensions including: hot-cold, health, age, gender, illness, rituals and economic status. These symbolic classifications influence food selection, food preferences, and dietary intakes. In this study, we adopted these symbolic classifications as guiding principles in the analysis of the research findings. However, hot-cold and economic status factors were excluded because they were not established in this study. On the other hand, symbolic classfication of food as “traditional” or “cultural” and “dirty food” emerged in the study and these were included in the model. This study was conducted in Uasin Gishu County in Kenya. Uasin Gishu County is one of the 47 counties of Kenya and is geographically located in the western part of the country. It has its headquarters in Eldoret town. The county is divided into six sub-counties and it covers a total area of 3,345.2 km2 with a total estimated population of 1,023,656, comprising 50% male and 50% female [32]. The predominant settlement pattern is rural (64.1%). There are 171 health facilities in the county, of which 90 are public [32]. Most of the facilities are concentrated within Eldoret Municipality. The climatic conditions and soil type in this region are generally favourable for a wide range of livestock and crop production. The Kalenjin are the predominant ethnic population in Uasin Gishu County. This ethnic group is composed of smaller subtribes: the Kipsigis, Nandi, Tugen, Keiyo, Marakwet, Pokot also called the Suk, Sabaot and the Terik. The ethnic group has its own mother tongue. However the sub-tribes speak their own dialect. The Nandi have the highest settlement in the county, followed by the Keiyo [13]. The Kalenjin had been semi-nomadic pastoralists of long standing. They had been raising cattle, sheep and goats and cultivating sorghum and pearl millet since at least the last millennium B.C. The Kalenjin have a common staple diet: Kimyet (ugali) (a paste of cooked maize or millet flour sometimes mixed with sorghum flour), native vegetables and mursik (sour milk mixed with charcoal dust, sometimes mixed with cow’s blood), supplemented with roast meat (usually beef or goat). Fish was also part of the traditional diet though largely limited to residents bordering the lake region community. Childbirth among the Kalenjin was exclusively women’s concerns from which men were excluded. After delivery, women were considered unclean as a result they were secluded for a certain period of time during which, their movement was restricted to a given point within the household and were fed on special diet until the cleansing rite was performed. Semi-structured and key informant interviews were used for data collection in 23 rural health facilities selected from the six sub-counties of Uasin Gishu County between April and August 2015. Uasin Gishu County was purposively selected because this study is part of a broader project in the County on maternal and child nutritional health. The main factor considered in selecting the health facilities is that they must be in the rural area and have their population mainly composed of the Kalenjin clients, and be spatially distributed from each other to diversify responses. Pregnant and post-natal Kalenjin women who came for routine antenatal and child welfare check-ups in the Maternal and Child Health (MCH) section in the rural health facilities were included in the study, based on availability and willingness to participate [33]. A combination of closed and open questions were used to investigate women’s subjective experience of pregnancy and nutrition [34]. The women were individually interviewed face to face in a private room at the health facility to give the opportunity to observe their facial expressions and body language, particularly important for correct interpretation of the answers [35]. Each woman was interviewed once by the first author and two female research assistants. The research assistants were well trained to ensure consistency and to ensure inter-researcher reliability. The interview guide was originally written in English but was translated into Swahili and into the local Kalenjin language where necessary. The local language interviews were conducted by the research assistants because they could speak and understand the language. The researchers obtained written informed consent before commencing with data collection. Respondents were asked about the following main issues: food that is restricted or recommended during pregnancy, and the underlying reasons, opinions on the beliefs and who gave the advice. Respondents’ social demographic characteristics were also captured. Each interview lasted for 25–90 min. None of the respondents pulled out of the interview before completion. The data collection exercise was continued until saturation was reached [36] at a sample size of 154 responses. Out of the 154 respondents, only 42 gave consent to be audio taped; these 42 respondents provide the direct quotations in the results section. Most respondents feared to be audio-recorded, probably because many audio recorded witnesses to the 2007 post-election violence died mysteriously. As a result, for the other 112 respondents, extensive note taking was preferred to create a more relaxed interview setting and encourage full disclosure [36]. Demographic characteristics of these respondents are presented in Table 1 below. Demographic Characteristics of Respondents Out of the 154 respondents, 62 were antenatal clients (ANC) and 82 post-natal clients (PNC) who had children younger than one year old. Various sub-ethnic groups of the Kalenjin were represented: the Nandi were the majority (71%) followed by the Keiyo (18%), the Marakwet (5%), and the Kipsigis (3%). The Nandi have the highest population in Uasin Gishu County followed by the Keiyo and the Marakwet [37]. There were two Tugen respondents, while Terik and Pokot were represented by one respondent each. The age of the respondents ranged from 18 to 42 years, with the majority of the respondents (61%) ranging between 20 and 29 years old. More than half (58%) of the respondents work in the domestic informal sector as either subsistence farmers or housewives. Most of them (88%) had not received more than secondary education. Most women were married (71%), while 18% were unmarried. Of the women who were unmarried, 50% were either pupils or students. 28% of the respondents were gravida 1. In-depth key informant interviews were conducted with Traditional Birth Attendants (TBAs) who were also traditional pregnancy herbalists (n = 6), community health workers (n = 2), and the nursing officer in charge of MCH (n = 1). The nursing officer and the community health workers were selected from one of the largest facilities because they are likely to encounter a wide range of pregnancy nutritional experiences and challenges given their large catchment area. The TBAs were identified by snowball and convenient sampling through respondents who gave birth at home and who took herbal remedies during pregnancy. The TBAs were interviewed at their practise, either in their homes or market centres. Information gathered from key informants was employed to explore meanings and enrich the responses obtained from respondent interviewees. Key informants were interviewed about the following main issues: perceived restricted and recommended foods; in-depth information on underlying reasons for these recommendations; and their food advice to pregnant women. Each interview took 85–100 min. The key informants, particularly the TBAs, accompanied the researcher to the gardens to identify the food crops that were mentioned in the study, giving them their local names. The identified crops and animals were then photographed and collected for proper scientific identification and classification in consultation with the plant taxonomists at the University of Nairobi herbarium and the National Museums of Kenya Forty-two interviewees and most key informants (except the MCH nursing officers and community health workers) consented to being tape-recorded. The Swahili tapes were transcribed in English by the first author while the Kalenjin tapes were transcribed by a Kalenjin speaking transcriber. For the other respondents, extensive note taking took place. The transcripts and notes were coded and categorised in themes using the software for handling qualitative data MAXQDA (version 12.1.3) with each participant identified by a pseudonym. The pseudonyms are used in the narratives presented in the results section. The codes were then analysed for patterns, pre-set themes, emerging themes and categories including food that is recommended during pregnancy, food that is restricted during pregnancy and the underlying reasons, opinions on the beliefs and who gave the advice. The SPSS program (version 23) was used to establish the frequencies of descriptive statistics. The research study was commenced after being approved by the National Commission for Science, Technology and Innovation (NACOSTI), Kenya. At the county level, research clearance was issued by the Uasin Gishu County Commissioner of Health, the County Administration Commissioner and the County Commissioner of Education. Request to access individual health facilities was granted by officers in-charge of the facility prior to the actual data collection day. The study was explained to each respondent and those consenting to participate were then requested to sign a written consent form. The respondents could withdraw from participation or answering further questions whenever they wished. The respondents’ identity was kept anonymous and their responses confidential.

N/A

• riang’a, roselyter monchari
• broerse, jacqueline e.w.
• nangulu, anne kisaka