Relationship of household diversity dietary score with, caloric, nutriment adequacy levels and socio-demographic factors, a case of urban poor household members of charity, Constantine, Algeria

This study was conducted to assess dietary diversity of Algerian urban poor household with household dietary diversity (HDDS) and to study its relationship with nutrient adequacy level and sociodemographic factors. A survey was followed by three 24 h dietary recalls during one year, with household’s members of charity in Constantine. Qualitative method has been realised using household dietary diversity score and quantitative method was done by calculating ratio of caloric and nutritional intakes to household needs. The study showed that the mean HDDS was 6.8±0.7 food groups. Animal proteins percentage, caloric adequacy level and adequacy levels of fourteen vitamins and minerals had positive correlation with HDDS. Household with less than six persons and those with children less than five years had a higher HDDS (p<0.035 and p<0.0001, respectively).


INTRODUCTION
Quality diet encompasses adequate coverage of basic macro and micro nutrient needs and diet variety. Households living in urban areas are more prone to food insecurity, because they source the vast majority of their food through the market. Any decline in household income or increase in food prices can have catastrophic consequences on them. Available studies on food diversity were limited as far as Algeria is concerned.
In an urban setting, where inhabitants are usually disconnected from the direct production and distribution of food, inability to access food resources can be the most immediate and critical manifestation of the many dimensions of poverty (Battersby and Watson, 2019). Algeria is the largest country on the Mediterranean and the second largest in Africa. The urban population in Algeria is estimated to account for more than 70% of the entire population (United Nations, 2016). Rate of poverty in Algeria was 10.5% in 2012 (World Bank report, 2013).
E-mail: hassanil13@yahoo.fr. Almost 75% of Algeria's poor live in urban areas, doing informal jobs or depend on subsistence agriculture (World Bank, 2016). The survey MICS4-2012 (Multiple Indicator Cluster Survey) showed that 1% of urban Algerian population suffered from multiple deprivations. The intensity of poverty 1 in urban areas was 34.8% (CNES, 2016). By mid-1992, most of the food subsidies were eliminated except for semolina, flour for bread, milk, and the other goods like energy and public services were adjusted since 1990 (Louali, 2016). In 2011, sugar and oil were exempted from customs duties and value added tax. Since Households, especially those in urban areas, buy purchased food (Girma et al., 2015), poverty lines are commonly drawn because of how much money is required to meet basic food needs. There is therefore a clear correlation between income and food security. Urban households source the vast majority of their food through the market.

Abbreviations
Studies have shown that an increase in dietary diversity is associated with socioeconomic status and household energy availability; lack of dietary diversity is a severe problem among poor populations in the developing world. The household dietary diversity score (HDDS) important in urban areas, where one of the important threats to food security is poor diets. HDDS was developed to measure household food access, one of the levels of food security. Previous research has shown dietary diversity is related to food security (Vellema et al., 2016). Thus, food may be available but not accessible to certain households during a given period of time, if they cannot acquire sufficient quantity or diversity of food through these mechanisms (Fanzo, 2017). The present study was conducted with general aim of assessment of dietary access of lowincome urban households, members of charity, from Constantine, Algerian, Eastern city and third largest city with HDDS. Secondary objective was to study the correlation of HDDS with nutritional adequacy levels and socio-economic factors.

Sampling
The study was carried out with all members of the oldest charitable association of Constantine, urban commune in the northeast of Algeria, capital of the East of Algeria, and the third most populated city in the country. The association exists since 1984. It helps poor households, after verifying their situations.

Study and subjects
A cross-sectional study, repeated 3 times in one year was conducted among the beneficiaries of aid from Constantine.
Households included in the study are declared in need by a specialized organism, after establishing files administrative proving their needs. Households with dependent children without fixed incomes, members of charity, were included in the study.
Semi-structured questionnaire was used for collecting socioeconomic details and information regarding household dietary practices. Questionnaire was followed by three 24 h recalls during one year. The study was conducted with household heads, all of whom were responsible for family food's preparation. They were asked to recall all foods eaten and beverages taken by the whole household on the previous day. Inclusion criteria were that the participants were able to answer the questions and recall their diet from the previous day. Household heads who prepare the food for family were included in the study.

Socioeconomic characteristics
A questionnaire was used to capture information on socioeconomic characteristics of household, such as sex, age, education level, sources of income, habitation, connecting to electricity, gas, water, cleaning up, and possessions. The questionnaire was administered through face-to-face interview with respondents.

h recalls
In order to avoid seasonal differences, three 24h recalls dietary were conducted to determine household food consumption. Head of households were asking to recall food consumed for the past 24h, bearing in mind all foods consumed within their homes. The details included descriptions of all foods and beverages consumed, including cooking techniques. The quantities of food consumed were estimated using household measures. The amount of nutriment consumed was then estimated and calculated using CIQUAL (2015) Table. Mean nutriment intake was estimated from the three recalls for each household.

Caloric and nutriment adequacy levels
Caloric or specific nutriment adequacy levels are the ratio of household intakes to their needs. Intakes were assessed based on caloric or specific nutrient content of each food consumed in each household. Needs are the sum of caloric recommended dietary allowance or specific nutriment recommended dietary allowance by sex, age and physical activity of household members.

Household dietary diversity score
Food access was assessed using HDDS (FAO, 2013). Information on dietary diversification was extracted from 24h recalls, where the consumed food items were grouped into twelve food groups: cereals, roots and tubers, vegetables, fruits, legumes, nuts and seeds, meat, fish and other sea food, eggs, milk/milk products, oil/fats, sweets (sugar, honey, sugar cane), condiments and beverages. Total dietary diversity score was calculated by summing the number of food groups consumed at home. The households were divided into four classes, according to eating habits of Algerians: Low dietary diversity: When households consume, less than six food groups; Medium dietary diversity: When households consume, six to seven food groups; High dietary diversity: When households consume, eight or more food groups;

Analysis
Data was analysed using Epi-Info 3.5.1 (2008). Averages are expressed on mean (±standard deviation). Analyse of variance (ANOVA) was used to compare several means. The correlation between quantitative values was tested with Pearson test. t and z tests were used to compare two averages with N>30 and N<=30, respectively. All values were considered significant at p<0.05.

RESULTS AND DISCUSSION
There are few studies about diet and dietary quality among the Algerian, this study was conducted to enrich the literature with data from this country. Its objective is the evaluation of poor household dietary diversity, and its relationship with caloric, nutriments adequacy levels and socio-demographic factors.

Household characteristics
One hundred and fifty-three households were involved in the study.  (Table 2). The whole household heads in this study were widowed, divorced or alone women, the vulnerability of female household heads derived from the fact that many women are unemployed or employed in low paying jobs such as domestic work. Furthermore, most of them were also surviving from a single income, which was not adequate for household food sustenance. In both developed and developing countries, the emergence and strengthening of the phenomenon of single-parent families present remarkable analogies, similarity of causes (family breakdown, loss of responsibility for men, gender inequalities), socio-economic markers (restricted access to education, training and skilled jobs) and consequences (marginalisation, exclusion and feminisation of poverty) (Bessis, 1996).

Household dietary diversity score
Mean HDDS was 6.8 [±0.7] food groups. Households consumed minimum 2 food groups, and maximum 11. HDDS was mainly between 6 and 8 food groups ( Figure  1). The HDDS doesn't vary through the three recalls (Table 3). About 7.2% households consumed up to five food groups (low dietary diversity), 86.9% consumed six to seven food groups (medium dietary diversity) and 5.9% of households consumed eight or more food groups (high dietary diversity) in their diet. From the twelve food groups, most of the households have consumed less than eight food groups (94.1%). The proportion of households with adequate dietary diversity in this study was therefore 5.9% (Figure 2). According to Petry et al. (2015), the dietary diversity score in all African countries is similar at minimum 6 points. This result was different from the ones found in Cambodia, 4.7 (±1.6) (McDonald et al., 2015); in Ethiopia, 5.2 (±1.9) (Geremew et al., 2019); in South Africa, 5.08 (South Africa, 2015). Also, in Tchad, 5.1 in sedentary households (Bechir et al., 2011); in Sahrawi refugees camp (Tindouf, Algeria), 3.8 (± 1.4) (SandsmarkMorseth et al., 2017); and in Indonesia: 9.1 (Trias et al., 2017). These findings are compared to those in Myanmar, 6 .2 (± 1.4) (Victoria, 2014), in Nioro 7 (± 1), in Nara 6 (± 2) (Welthungerhilfe, 2013) and in South Africa 6.63 (Grobler, 2015). Due to rarity of studies on urban household dietary diversity score based on 12 food groups, this study was compared to the studies above, which most of them were conducted with rural households. Differences in study area, study period, cultural beliefs and dietary patterns may be explained differently between studies.In this study, HDDS varied less by seasons, the same result was found in Hirvonen et al. (2016).

Description of food groups consumed
The details of the proportion of households who consumed various types of HDDS food groups, over one year (mean of three 24 h recalls distributed in one year) are presented in Figure 3. Cereals were the most commonly consumed food group by 100% of the households, followed by sugar (98.7%), milk (98.3%), vegetables (94.1%), oil and fats (89.5%), potatoes    (71.9%), pulses (35.9%), condiments and coffee (20.3%), eggs (13.5%), meat and poultry (10.2%), fruits (7.4%) and fish (1.1%). The most commonly consumed food groups were cereals, sugar, milk and oil. The exemption from customs duties and value added tax for raw materials of sugar and oil and the subvention of milk and flour for bread have made it possible, to maintain these products at an affordable cost for all the Algerian population. On the other hand, these food groups are the staple of Algerian kitchen: cereals are consumed daily in all meals; sugar and milk are also consumed daily at breakfast and at afternoon snack. While vegetables (onion and garlic specially) with oil are presumably used at the beginning of sauce in Algerian cooking methods. Potatoes and pulses were also widely consumed; condiments are used in meals preparations while coffee is a stimulant drink and the most consumed in northern Algeria.
However, fish, meat, egg and fruits were the least consumed food groups. These food groups except eggs were the most expensive in markets. These results are consistent with the study of Padilla (2008), the Southern Mediterranean countries diets, which is mainly vegetarian as only a small proportion of calories is of animal origin; cereals are the basic ingredient and pulses the main protein source. Cereals are staple food in North Africa. The proportion of energy intake from cereals, roots and tubers is about 55% (Tanyeri-Abur, 2015).

Relationship of HDDS with socio-demographic factors
The study showed that HDDS did not vary between education level and marital status of household head. The HDDS was higher at the household with less than six persons and at those with under five years' children (Table 4). Several studies have shown that education was positively correlated to high dietary diversity (McCordic, 2016), schooling impact positively on dietary diversity. However, in this study, the level of education did not show any association with household dietary diversity; the same result was found in (Grobler, 2015;Mayanja et al., 2015).The HDDS was higher at the household with less than six persons compared to those with more than 6; this result is similar to Odusina Olaniyi (2014), which found that households with larger sizes (p <0.1) had the propensity of being food insecure and Grobler (2015) which showed that an increase in household size will decrease household dietary diversity.

Relationship of HDDS with number of meals per day
Households consuming four or more meals a day had a higher HDDS compared to those taking three or less meals a day (6.99±6.63 versus 6.63±0.65 p <0.0001). The HDDS was positively correlated with the number of

Relationship of HDDS with caloric and nutriments adequacy levels
On twenty-two nutriments studied, HDDS was positively correlated with adequacy level intakes of energy and fourteen nutriments: lipids, vitamin B1, B5, B6, B9, B12, A, D, E, C, iron, zinc, selenium and calcium. HDDS increase with increasing of percentage of adequacy level of these nutriments. In the other hand, HDDS was positively correlated with animal proteins percentage and negatively correlate with plant proteins percentage (Table  5). Adequacy levels of nearly 70% of the nutriments studied, caloric and percentage of protein from animal sources were positively correlated with HDDS, as the percent of nutritional adequacy increases, the HDDS increases; which means that more households consume a balanced diet, more their HDDS increases.
According to Swindale and Bilinsky (2006) a more diversified household diet is associated with caloric and protein adequacy, percentage of protein from animal sources, as well as household income. Poor households frequently depend on a diet high in starchy foods with limited protein and other nutrients (South Africa, 2015). The increase in animal protein share in diet increases the HDDS, while the increase in those of plant origin lowers the HDDS. This may be explained by the fact that HDDS contains four-food groups vector of animal protein: eggs, meal, milk and dairy, fish and seafood; while vegetable, mostly pulses group, represents proteins. Our results are consistent with Kennedy et al. (2010), HDDS increases with increase of food groups that provide micro-and macronutrients. Increased dietary diversity correlates with increased intake of micronutrients such as calcium, vitamin A, and iron with greater consumption of animal source foods, fruits and vegetables, and dairy.

Conclusion
Nutritional mistakes made during the growth period can be permanent and irreversible, hence the importance of nutritional studies on populations at food insecurity risk. The diet of Algerian poor urban household from Constantine was quite monotonous; they have less varied diet. They have poor quality diet, based on plant sources and very few animal sources; they tend to buy less meat, dairy products, fruits and fish. Cereals, milk, sugar, vegetables, oil and potatoes are their food staple. Except vegetables and potatoes, all of them are subsidized foods. Milk and pulses are their main source of protein; while meat, eggs and fish are scarcely consumed, which may reveal a possible under nutrition.
The study findings confirm that less varied diets are found in poor households from Algeria and more diversified household diet is associated with caloric and nutriments adequacy level, with a percentage of protein from animal sources. The score was easy to use, but requires some exactness in definition of universal class HDDS. Greater household access to varied food, nutrition education and development of household's chance to increase incomes should be considered. The nutrition education may include topics that can help increase maternal knowledge and improve practices related to feeding. On the other hand, the amount of fresh diet and animal food sources could increase with distribution vouchers.