Protein Consumption in Communities Affected by Stunting in Daerah Istimewa Yogyakarta, Indonesia

Inadequate protein consumption is characterized by a low intake of energy and proteinous foods, leading to stunted growth and negative public health outcomes. Therefore, this study aims to determine the factors influencing consumer behavior in purchasing animal and vegetable proteinous foods in Daerah Istimewa Yogyakarta (DIY), as well as to analyze the relationship between protein consumption and these influential factors. A quantitative design was used along with a survey approach, involving both primary and secondary data. The study participants were selected using a non-probability sampling method. Data were collected through observations, interviews, literature studies


INTRODUCTION
Protein, an essential nutrient, intricately orchestrates numerous vital life processes.Comprising a spectrum of amino acids, some of which are indispensable as they elude endogenous synthesis, its acquisition via dietary sources is imperative.The corpus of research underscores its pivotal role in cellular proliferation, renovation, and rejuvenation within the human framework (Maulidiana & Sutjiati, 2021).The indispensability of a robust intake manifests in the Dwi Aulia Puspitaningrum 1 , Khoirul Hikmah 2 , Muhammad Maulana Azimatun Nur 3 , Aditya Ananta Putra 4 , Ravid Ikhsan T.J 5 , Mukminan Arief H 6 , and Vista Nur Defiana 7  optimal panorama of growth and maturation.
Moreover, protein underpins the genesis of nascent tissues while also stewarding the perpetuity of extant substrates.Inadequacies herein have been correlated with malnutrition, an affliction that casts a long shadow over human resource viability (Umaroh & Vinantia, 2018) .A parallel strand of significance is interwoven with protein's contribution to the architectural scaffold of bones, thereby intimately influencing skeletal advancement.This nutrient further exerts a hand in modulating the exudation and efficacy of the osteotropic hormone IGF-I, thereby imprinting upon the genetic latitude for attaining zenith bone mass (Sari et al., 2016).In Indonesia, insufficient intake of protein coupled with shifts in dietary patterns has emerged as a prominent catalyst for malnutrition.Despite persistent endeavors spanning the last two decades, a distressing statistic persists: a substantial one-third of children below the age of five continue to grapple with malnourishment, presenting as stunting, wasting, or even exhibiting overweight conditions.
Concurrently, the remaining two-thirds teeter on the precipice of malnourishment and veiled undernourishment due to suboptimal dietary consumption.Evidently, a staggering cohort of Indonesian children and adolescents bear the brunt of stunting as a consequence of malnutrition's grip (Suryana & Azis, 2023) Of paramount concern, the insufficiency of protein uptake during the critical growth phase of toddlers can precipitate developmental lags and the onset of stunting.A cogent antecedent inquiry unveiled a direct correlation between the adequacy of protein provisioning and the incidence of stunted growth (Amalia et al., 2022).Stunting, a condition colloquially denoted as reduced stature among toddlers, encompasses a state of growth faltering characterized by chronic malnutrition, culminating in an age-incongruent diminution in physical stature.This predicament holds pivotal import, as it intertwines intricately with the human resource quotient.Beyond malnutrition, its etiology extends to encompass recurrent infections within the formative period of initial 1,000 days of existence .et al., 2020) .Furthermore, consumer behavior is often affected by environmental factors, such as culture, social class, family, and situation, as well as individual differences in motivation, knowledge, attitudes, personality, lifestyle, and demographics (Kodish et al., 2015) Consumption of protein in a household can be influenced by the psychological process of information processing, learning, and attitude, as well as behavior change.Income has also been reported to be a crucial factor in measuring the ability of the household to purchase essential elements related to improving child nutrition (Vaivada et al., 2020) .Families with high are more likely to consume a variety of protein-rich foods (Annisa et al., 2022).
Research on stunting, as examined through the adequacy of both animal and plant protein intake, is not well known so far, particularly when approached through an integrated lens that considers the causal factors and potential solutions to the problem.Previous research has primarily focused on food access and household food security (Sitompul et al., 2023) food access and its relationship with the development of small businesses , and household protein consumption in rural settings ((Umaroh & Vinantia, 2018) .To the best of our knowledge, no prior studies have investigated households impacted by stunting in Yogyakarta.The novelty of this research is to understand the relationship of the protein access on the stunted children in Daerah Istimewa Yogyakarta (DIY).To achieve balanced nutrition and prevent stunting, it is essential to promote local food production in DIY.Despite the significant effort to reduce the incidence of stunted growth, with the prevalence falling from 21% percent in 2019 to 17.3 percent in 2021, access to adequate nutrition for pregnant women and toddlers, especially those from lower-income households, remains a challenge.Therefore, this study aims to determine the factors influencing consumer behavior in purchasing animal and plant protein foods in DIY, as well as to analyze the relationship between protein consumption and these influential factors.The results are expected to help in providing alternative solutions to reduce the prevalence of stunting in DIY.

METHODS
The  (Huriah et al., 2019) .This study employed nine variables, consisting of seven independent variables and two dummy independent variables.Therefore, the minimum required sample size is 9 x 10 = 90 respondents.For Multivariate data analysis Include regression analysis, the sample size should be 10 times greater than the number of variables (Roscoe, 1975) .In this research, a sample size of 150 respondents was employed, considering the study's span across five districts and cities.Each district and city contributed 30 respondents, selected through a random sampling method using a simple random sampling technique involving a lottery system by Spiner Api.Some areas were selected due to their stunting rate exceeding 10%, as reported in the SSGBI, showing the need to provide alternative solutions to reduce the rates.This study employed descriptive quantitative methods, which were in line with concrete, objective, measurable, rational, and systematic scientific principles, using numbers and statistics for analysis.The descriptive method was used to explain, predict, and control a symptom by systematically describing and creating an accurate picture of the facts, nature, and relationships between the phenomena studied.The procedure carried out included field surveys and data collection via questionnaires.Laboratory tests were carried out for data processing at the Statistics laboratory and Department of Agribusiness, Faculty of Agriculture, UPN Veteran Yogyakarta.The field survey was performed through direct observation of the respondent's condition and all the data obtained were recorded.Data collection was carried out using triangulation techniques, which involved semi-structural interviews with observed respondents, forum group discussions (FGD), direct observation in the field, and statistical laboratory tests.

Multiple Linear Regression
This study was focused on protein consumption among communities and families affected by stunting, rather than their staple food intake.This is due to the consideration that the primary cause of stunting is a deficiency in protein, which serves as a crucial building block for the body.Setyani's research in 2021 found that the lowest protein intake in the stunted group was 68.7%, whereas in the non-stunted group, it was 90.1%.There are three articles utilizing a cross-sectional research design and seven articles using a case-control design, all of which report prevalence rates exceeding 20% (high).Nine articles indicate a significant relationship between protein intake and the occurrence of stunting, with a minimum protein intake sufficiency level of 90%.Therefore, this research focuses specifically on protein consumption within families affected by stunting and employs a Multiple Linear Regression approach for analysis.
Multiple regression analysis was carried out to analyze the factors influencing consumer behavior in purchasing animal and vegetable protein foods in DIY, as shown below:

Factor Analysis
The relationship between animal and Plant protein food consumption and stunting conditions in DIY was analyzed using factor analysis, as shown below: Fm = ℓm1

Consumption of Animal Protein
The data presented in Table 1 illustrates that within Sleman Regency, the average respondent exhibited a predilection for fish consumption, while in Gunungkidul Regency, Kulonprogo Regency, and Yogyakarta  to vegetable protein intake.Based on the relatively high cost of animals, it was not affordable for some people, leading to a dependence on vegetable protein products (Swarinastiti et al., 2018).

Analysis of Factors Affecting Protein Consumption in Households with DIY Stunting Potential
Table 3 shows the results of the regression analysis of protein consumption in the Special Region of Yogyakarta (DIY).The analysis revealed that several factors significantly influenced protein consumption in the province.These factors include household income (0.000), women's income (0.000), number of family members (0.000), number of children under five (0.000), age of women (0.007), and distance of the house to the nearest shop/minimarket (0.001).The significance of t values for all six variables was less than the alpha level of 0.05, indicating that they had a significant impact.These findings suggested that efforts to increase the intake of this nutrient as well as prevent stunting in the region must consider these significant factors, especially household income and women's income The regression analysis results showed that the selected variables significantly influenced protein consumption, as indicated by the significance value of the F test, namely 0.000<alpha 0.05.Furthermore, the nine variables had a 64.5% effect on the intake of this nutrient, which was indicated by the Adjusted R 2 value of 0.645.The remaining 35.5% was influenced by other factors that were not included in this model.The results showed household income, women's income, women's age, number of family members, dummy toddlers, and distance from home to the nearest shop/shop/mini market, significantly affected protein consumption in families and homes with stunting potential.Supariasa & Purwaningsih (2019) reported that one of the main factors causing stunting was household income.A low income could affect the ability of the family to purchase food needs.Insufficient consumption of nutrition in toddlers had been shown to be one of the causes of stunted growth.Women's age also had a significant influence on the potential for stunting through protein consumption.(Sani et al., 2020).stated that women played a major role in the occurrence of this condition.
Furthermore, the occurrence of pregnancy at an age of <20 years or >35 years posed the risk for complications.These pregnancies could lead to intrauterine growth restriction, premature birth, infant death, and poor child growth.The number of toddlers in a family had also been reported to be a factor affecting protein consumption and potential stunting among children.Families with many children, especially those with less economic conditions, could not provide adequate attention and food to all the toddlers (Siswati et al., 2020).
In households within the Yogyakarta Special Region (DIY) that are susceptible to stunting, there exists a notable divergence in the consumption of animal and vegetable proteins.This disparity is particularly evident across different districts, as illustrated in Figure 1.The data indicates that the highest protein intake occurs in Bantul Regency, succeeded by Sleman, Gunungkidul, Yogyakarta, and Kulonprogo.These variations are attributed to a spectrum of factors within the population, encompassing economic status, environmental influences, and other pertinent variables.The distinct circumstances prevailing in each district contribute to the diversity in protein consumption levels.Consequently, tailored strategies for averting stunting must be devised, accounting for the unique conditions present in every district within DIY.
The five regencies/ municipalities in DIY, particularly Kulon Progo, had low protein consumption, which could be attributed to various factors highlighted in Table 3, as well as environmental conditions, topography, and geography.Furthermore, Kulon Progo had a mountainous area and was quite remote, making it challenging for households to access food.These conditions led to limited food diversity and reliance on carbohydrates in the area.Socio-demographic and economic factors, such as the head of the household, their age and education level, number of household members, area of residence status, and household per capita income, could affect the diversity of household food intake (Dewanti, 2020).The low protein intake in families with stunting potential also significantly contributed to the presence of stunted growth in DIY (Siringoringo et al., 2020).This study demonstrated that the risk of stunting was 6.495 times higher in toddlers with low consumption, highlighting the importance of increasing protein intake, particularly among vulnerable populations in the region.The result from KMO and Bartlett's Test showed a sphericity coefficient of 126.433 with a significance of 0.000.This indicated that the correlation between the variables was significant at the 0.05 level.Furthermore, KMO Measure of Sampling Adequacy test showed a value of 0.518, indicating that the sample adequacy measure was satisfactory, as shown in Table 4.

Analysis
Diet plays a crucial role in determining the nutritional status of individuals and their families.Factors such as economic status, availability and accessibility of food, cultural beliefs and practices, and knowledge about nutrition all contribute to the dietary choices made by families, which in turn can impact their nutrition .(Lin et al., 2011) Additionally, the level of protein consumption in a family can also significantly affect their nutrition.Research has shown that protein intake is strongly correlated with family nutrition.Higher protein consumption is associated with improved nutritional status, including better growth and development in children.

Communalities Test
The concept of communality pertains to the extent to which the variance exhibited by a given variable can be elucidated by the latent or underlying factors present within a dataset.This metric is quantified within a range of 0.0 to 1.0, whereby a value of 0.0 signifies an absence of any discernible correlation between the variable in question and the latent factors.Conversely, a communality Source: Primary Data Analysis 2022 value of 1.0 signifies that the entire spectrum of variance displayed by the variables is entirely attributable to a set of shared underlying factors.
In essence, communality serves as a means to gauge the degree of influence that latent factors exert on the variability observed in a specific variable.This measure allows researchers to comprehend the extent to which a variable's behavior can be ascribed to hidden factors, offering insights into the interplay between these latent ascribed to hidden factors, offering insight into the interplay between these latent influences and the observed data patterns.By calculating and interpreting communality values, researchers can gain a deeper understanding of the relationships that contribute to the overall variance exhibited by the variables under investigation.The communality of the study variables ranged from 0.494 to that the variable could be reduced into several factors, as shown in Table 5.

Principle
Component Test The principal component analysis (PCA) was employed as a technique to condense the original set of 7 study variables into a more concise representation defined by several underlying factors.This dimensionality reduction aimed to capture the essential information within the dataset while minimizing redundancy.The criteria chosen for extracting these factors was the Latent Root Criterion, which involves identifying factors that possess eigenvalues greater than 1, indicative of their significance in explaining the data's variance.
Upon executing the PCA, the outcome was the identification of 3 distinct factors as evident from the results presented in Table 6

Factor Rotation
Factor rotation was carried out to obtain solutions that were theoretically and practically meaningful, thereby providing an explicit interpretation of the factors obtained.Furthermore, it could be used to improve interpretation by reducing the dualism in the solutions.This method was carried out using Varimax due to its ability to maximize variance.
Based on Table 7, factor 1 consisted of Distance to the Traditional Market (X6) and Distance to the Store/ Minimarket/Warung (X7), which could be referred to as the shopping location factor.Furthermore, factor 2 comprised variables of Household Income (X1), Age of the Woman (X3), and Number of Family Members (X5), and could be referred to as the family needs factor.The results also showed that factor 3 consisted of variables of Women's Income (X2), and Women's Education (X4), and was further considered the Women's factor.The distance from the house to the nearest minimarket or store (X7) shows a significant negative correlation.This phenomenon can be attributed to the Source: Primary Data Analysis 2022 fact that, for the community, proximity to the residence influences the ease of food access based on location.However, it exerts a negative influence because even though the distance is not a hindrance, the purchasing power of the family regarding food items, especially protein-rich foods available at that location, is not their preferred choice.The small significance value of 0.001 (which is less than 0.05) indicates a clear difference in protein consumption (Y) among the level II regions being studied.In simpler words, the data suggests that the differences in protein consumption across these regions are not random but are due to real disparities between them.This finding highlights the importance of investigating the factors behind these regional variations in protein consumption and emphasizes the need to adopt region-specific strategies when addressing dietary habits and nutrition-related matters.Previous study reported that the factors contributing to stunting prevalence revealed a correlation between the household's economic status and the incidence of stunting.Poor economic conditions impacted purchasing ability, access to nutritious food, and adequate healthcare services.The presence of numerous family members, including children aged 1-3 years, played a significant role in the occurrence of stunting.This is due to the larger family size affecting the availability and accessibility of quality food, resulting in The coefficient of the i-th variable X1 = Household income (Rp/year) X2 = Women's income (Rp/year) X3 = Female age (years) X4 = Women's education (years) X5 = Number of household members (people) DI = Dummy information DI = 0, information obtained from print media DI = 1, information obtained from electronic media D2 = Number of children under five in the family (Dummy) D 2 = 0 , No children under five in the family D 2 = 1, Any Childrem under five in the family X6 = Distance from house to the nearest market X7 = Distance from house to the nearest shop/minimarket/store µ = Error rate

Table 1 .
Animal Protein Consumption Data in Each Regency/City in the Special Region of Yogyakarta (rupiah/month/household)

Table 2 .
Plant Protein Consumption Data in Each Regency/City in the Special Region of Yogyakarta (rupiah/month/household)

Table 3 .
Regression Analysis of Protein Consumption in DIY Source: Primary Data Analysis 2022

Table 6 .
Total Variance Analysis

Table 8 .
Analysis of Covariance Protein Consumption in DIY

Table 9 .
Inter-District Covariance Analysis Test Results in DIY