Dietary Diversity among Indian adolescents and young adults: Evidence from UDAYA study

2 Objective: This work studied minimum dietary diversity (MDD) and explored its linkages with 3 background characteristics like household consumption behavior, presence of grandparents in the 4 household, number of siblings, involvement in paid work, etc. 5 Design: For bivariate analysis and sex differentials, chi-square test was done to study the 6 association between MDD and different covariates. Logistic regression analysis was performed to 7 identify determinants of MDD. 8 Setting: Data was collected from two majorly populous and backward states of India, namely 9 Bihar and Uttar Pradesh (UP). 10 Participants: Follow-up survey of the UDAYA study (2018-19) was conducted among 11 adolescents and young adults aged 12-23 years old. 12 Results: We found the prevalence of MDD to be 59% among males and 56% among females. 13 Bihar performed better overall with higher MDD and lesser gender inequality. Wealth Index and 14 caste were observed to be significantly associated with MDD. Food Consumption Score (FCS) of 15 the household and media exposure were significantly impacted the MDD. 16 Conclusions: Improving dietary practices at a younger age eventually results in improved 17 nutritional status and overall development of an individual. It can serve as the key to prevent any 18 nutritional deficiencies and diseases linked with it at later ages. The government should focus more 19 on imparting healthy practices related to diet among both adolescents and their families. Action is 20 required to refine the current schemes present for improvement in household food consumption, 21 more so for the poorer population. Programs are needed that work on reducing gender inequalities, 22 especially in the state of UP. 23 of acceptability


Introduction 34
India's 253 million adolescent population (highest in the world) presents an unprecedented 35 opportunity as well as a challenge (1). Every fifth person in India is an adolescent (i.e., 10-19 years 36 of age). They can be divided into two categories -the younger (10-14 years) and the older (15-19 37 years), based on their behavioral attitudes and needs (2). Adolescence is a complex transitional 38 phase from childhood to adulthood, in which they undergo various rapid changes from physical 39 appearances to changes in their food habits (3,4). The adolescence stage is considered the second adolescents. It found that one-fourth of adolescents were deficient in vitamin D were deficient in 50 vitamin B12 and folate. In contrast, one-third of adolescents have vitamin A deficiency (7). 51 Micronutrient deficiency during early childhood can often transverse into adolescence with a long-52 term effect on health, cognition, education, and productivity (8). Dietary diversity prevents the 53 deficiency of micronutrients, and hence the onset of deficiency diseases and other related health 54 issues. Also, it is strongly associated with various factors such as food insecurity, socio-economic 55 Methods 82 Data Source 83 The UDAYA (Understanding the lives of adolescents and young adults) dataset has been used in 84 the study conducted by the Population Council, New Delhi. UDAYA is a longitudinal study done 85 in Uttar Pradesh and Bihar following a cohort of adolescents aged 10-19 years. These two north- 86 Indian states comprise 28% of the adolescent population in the country, given they are large, highly 87 populated, predominantly rural, high poverty states (18). 88 The study used both cross-sectional and longitudinal designs for sampling at the point of wave 1, 89 and a multi-stage systematic sampling design was employed. UDAYA was designed to provide  PSUs list was stratified using four variables, namely, region, village/ward size, the proportion of 96 the population belonging to scheduled castes and scheduled tribes, and female literacy. The 97 household sample in rural areas was selected in three stages, while in urban areas, in four stages.

98
Data collection for Wave 1 was done in 2015-16, and after three years, data was collected for wave 99 2 data in 2018-19. This paper analyses dietary intake in the past 24 hours, and this information 100 was collected only at wave 2. Hence, for the current study, a cross-sectional sample of only wave  The outcome variable of the study, i.e., minimum dietary diversity, was defined as the intake of 107 food belonging to at least 5 or more food groups in the past 24 hours, as defined by FAO  and; Other fruits. Out of these 10, UDAYA asked for information on only the first seven food 113 groups in its survey. For the last three food groups, questions were asked on the overall intake of 114 "fruits" or "vegetables" without emphasizing whether they were vitamin-A rich or not. To adjust 115 for this data limitation, a weightage of 1.5 was given instead of 1 to the two proxy food groups, 116 namely "Fruits" and "Vegetables", as they may or may not be rich in vitamin-A. This way, we got 117 the total dietary diversity score to fall in the range of 0 to 10 as it generally should, where 118 consumption of food from five or more food groups was defined as the proxy measure of minimum 119 dietary diversity for our study population. every day" and "at least once a week", whereas infrequent usage refers to "at least once a month", 131 "rarely" and "not at all". The final added score ranges from 0 to 8, which is then categorised in to 132 three terciles as Low (0-3), Medium (4-5) and High (6-8).

134
Food consumption of the household is reflected using a proxy index given some data constraints. household-level data on the diversity (quality) and frequency (quantity) of food groups consumed 138 over the previous seven days. It is then weighted according to the relative nutritional value of the 139 consumed food groups. For instance, food groups containing nutritionally dense foods, such as 140 animal products, are given greater weight than those containing less nutritionally dense foods, such as tubers. Broad food groups and associated FCS weights are: main staples-weighted at 2, 142 pulses-weighted at 3, vegetables-weighted at 1, fruit-weighted at 1, meat and fish-weighted 143 at 4, milk-weighted at 4, sugar-weighted at 0.5, and oil-weighted at 0.5. Condiments can also 144 be captured but are weighted at 0. Consumption frequencies are computed by adding the number 145 of days a food item is eaten in a week. Then they are added and rounded off to a maximum limit 146 of 7 per food group. An additive score combining the weighted consumption frequencies of each 147 food group results in a total score ranging from 0 to 112.     Table 1 represents consumption of items from different food groups in the past 24 hours among 168 younger and older adolescents by the sex of the respondents. It was found that almost 100% of the 169 adolescents were consuming grains, white root tubers, and plantains in their diet. Older adolescents 170 consumed dairy products and pulses slightly more (83% and 63% approx.) compared to younger adolescents (80% and 58% approx.). More than half of the adolescents ate fruits and vegetables 172 the previous day. Females were found to be eating fruits more compared to males, particularly 173 among older adolescents. Mainly, intake of older adolescents was found to be higher than younger 174 adolescents in most food groups. Only one-third of the adolescents ate dark green leafy vegetables.

175
Male adolescents consumed more nuts and seeds, especially in the group of younger adolescents 176 (around 29%) than females (24%). The gender differentials were quite prominent with regards to 177 the consumption of non-vegetarian foods among adolescents. Intake of eggs was almost double in 178 the case of males (18%) compared to females (9%) in older adolescents. A similar pattern was 179 observed in the consumption of meat, poultry, and fish in both younger and older adolescents.
180 Table 2 shows the prevalence of minimum dietary diversity where Bihar portrays better dietary 181 habits (MDD=61% in both males and females) than UP (MDD=58% in males and 54% in females).

182
Females faced more discrimination in UP compared to Bihar. Advancement in age leads to a   Table 3 shows output of Binary logistic regression analysis. We find out that caste, media where adolescents from the richest category households had 90% higher likelihood of achieving 208 minimum dietary diversity than those from the poorest category households (95% CI (1.2-3.1)).

209
Adolescents belonging to households with high food consumption were three times more likely to  Gender-based discrimination is still prevalent in northern, central, and eastern zones of India (22).

229
For instance, females across India consume nutrient-rich food less frequently compared to males 230 (23). Our study shows considerable differences in the non-vegetarian diet practices among male 231 and female adolescents in UP and Bihar. Females have consumed more fruits, whereas males 232 consumed more eggs and meat. This is observed among both age groups, but the gap is even wider 233 in the case of older adolescents. Likewise, a study based in rural India found gender disparity in 234 the dietary pattern of adolescents (24), and it corroborates with another study conducted among 235 adolescents in Bangladesh, where the inadequate dietary deficiency among adolescent girls and 236 boys was 55.5% and 50%, respectively (25). Contrary to that, a study from Australia did not find  found that exposure to mass media increases fruit consumption among adolescents (38).

258
The limitation of our study is that it uses a proxy measure of MDD and FCS as there was a lack of In developing countries, lack of food diversity due to constraints in access to different food groups 267 and monotonous consumption of certain cereals or food groups impedes achieving optimal 268 nutrition status. Therefore, there is a need to reemphasize the importance of dietary diversity.

269
Vegetables and fruits intake is as important as compared to that of cereals and pulses as it contain