Macronutrient composition of street food in Central Asia: Bishkek, Kyrgyzstan

Abstract Background Urban areas in central Asia are currently undergoing nutrition transition. Street food is very popular, but the specific foods available and their nutritional composition are unknown. The aim was to describe the availability and macronutrient composition of street foods in Bishkek, Kyrgyzstan. Results Trained interviewers collected data on street food vending sites’ characteristics and food availability (n = 596). Samples of the most commonly available foods and drinks were collected (n = 80 homemade; n = 40 industrial). Macronutrients were quantified through chemical analysis. Fruit, beverages, and food other than fruit were available in 4.0%, 61.7%, and 81.0% of the vending sites, respectively. Among those selling food other than fruit, 56.5% sold only homemade (e.g., bread, main dishes, snacks, pastries, sandwiches, and cakes), 23.3% both homemade and industrial and 20.2% only industrial foods (e.g., bread, snacks, pastries, cakes, and cookies). Homemade foods presented the highest energy/serving (median kcal/serving: 357 versus 145, p < .001). A high content in saturated and trans‐fatty acids was observed in some homemade traditional dishes and snacks, reaching, respectively, 30.2 g/serving and 2.9 g/serving (in homemade manty, a traditional dish). Tea and soft drinks were available in over 50% of the vending sites selling beverages. Conclusion The high availability of street food in Bishkek highlights its importance for this urban population. Traditional snacks, dishes, and beverages coexist with more westernized products. The variability in energy, macronutrients, and lipid profile of homemade and industrial products reflects heterogeneous culinary practices and ingredients. Policies promoting the availability of healthy foods and ingredients should be implemented.

In contexts of low resources and working long distances from home, such as in urban areas of central Asia, street food plays an important role on the diet of many consumers: it is cheap and easily accessible (Liu, Zhang, & Zhang, 2014;Steyn et al., 2014;Winarno & Allain, 1991) and frequently replaces home meals (Steyn et al., 2014), (Bhowmik, 2005). Only a few studies have focused on the nutritional value of street food (Abrahale, Sousa, Albuquerque, Padrao, & Lunet, 2019). Although there is a great heterogeneity regarding its nutritional composition, depending on the ingredients used, preparation and food processing methods (Draper, 1996;Namugumya & Muyanja, 2012;Steyn et al., 2014), it has been estimated that street food largely contributes to the total intakes of fat, particularly TFA, salt and sugar (Steyn et al., 2014).
There is currently a lack of representative data about dietary habits of urban populations in central Asia, and specifically, in Kyrgyzstan (Rippin et al., 2018). The objective of this study was to characterize the street food environment in Bishkek, Kyrgyzstan, focusing on the vending places and their food availability, and to describe the macronutrient composition of the most commonly available homemade and industrial foods and beverages.

| Setting and study design
This study was implemented within the scope of the FEEDcities project, supported by the World Health Organization (WHO)-Europe, which used a stepwise standardized methodology to characterize the street food environment in countries in Central Asia and Eastern Europe (World Health Organization, 2019). The definition of street food adopted in this study was the one proposed by the Food and Agriculture Organization and the WHO, as "ready-to-eat foods and beverages prepared and/or sold by vendors or hawkers especially in the streets and other similar places" (Food & Agriculture Organization, 1989;World Health Organization, 1996).
This cross-sectional evaluation was specifically designed to assess the availability and nutritional composition of the street food in Bishkek, capital city of Kyrgyzstan and was conducted between June and July 2016. Kyrgyzstan is a lower-middle-income country, characterized by a slight economic growth in recent years, of which one of the main supports is the agriculture sector. The country has about six million people (Central Inteligency Agency, 2019; World Bank, 2019). Bishkek is the main urban area and largest city in the country, with 996 000 inhabitants, approximately 17% of the national population (Central Inteligency Agency, 2019).
In a site visit conducted for the planning of data collection, members of the research team observed that most street food vendors in Bishkek were concentrated within or in the proximity of public markets. Thus, from a list of 19 public markets identified by local authorities, a fixed proportion (approximately 50%) was randomly chosen, resulting in 10 markets. The study areas were delimited by 500-meter diameter buffers around each selected market (with the centroid in its geographic midpoint), covering the market and its surroundings.
Eligible vending sites were defined as the business establishments selling ready-to-eat food, including beverages and/or snacks, from any venue other than permanent storefront businesses or establishments with four permanent walls not selling directly to the street, operating in the predefined perimeter. This included mobile vendors, as well as sellers with semi-static or stationary vending units. The exclusion criteria were the following: (a) food and ingredients. Policies promoting the availability of healthy foods and ingredients should be implemented.

K E Y W O R D S
Central Asia, Food Processing, Kyrgyzstan, Nutritional Value, Ready-Prepared Foods, Street Food establishments with four permanent walls; (b) permanent storefront businesses; (c) street vendors selling exclusively nonfood products or raw foods not ready-to-eat; (d) food stalls and carts which were part of permanent stores or licensed establishments.

| Data collection: vending sites, vendors, and food availability
The markets were assessed on consecutive days, during 3 weeks, ensuring the representation of both weekdays and weekends.
Ten interviewers, operating in pairs, canvassed the study areas in search of street food vending sites, starting by the assessment of the whole market and then moving to the surroundings. After registering the Global Positioning System (GPS) coordinates of each vending site, they collected the following information, through di- Afterward, the interviewers systematically invited vendors from one in every two vending sites (based on an initially estimated mean number of 100 eligible vending sites per market) to participate in the study. They explained the study objectives and procedures and asked for an express consent to participate in the study. When the vendor agreed, they carried out computer-assisted personal interview, enquiring about vending site ownership, characteristics (access to drinking water, to toilet facilities and electricity), food vending activity (operating periods-during the week; during the year and under which weather conditions) and food availability, including products sold and serving sizes. A shortened version, ensuring a brief description of each of these categories, was conducted with the food vendors operating mobile vending sites, given their limited time to answer the questionnaire.
A total of 1,216 eligible street food vending sites were identified and 604 vendors were invited to participate, of whom 596 (98.7%) accepted. No statistically significant differences were found between participant and nonparticipant vendors regarding mobility and the physical setup of the stationary vending sites.
Foods available was grouped according to their nature, into fruit (product in natura, either fresh or dry), beverages (any alcoholic and nonalcoholic beverage), or food other than fruit. Food other than fruit was further classified as homemade (foods of domestic manufacture cooked and/or prepared at home or on the street, even if using industrial ingredients) or industrial (food products produced by food industry and sold as is without further preparation and/or cooking, encompassing both processed and unprocessed foods (Monteiro et al., 2018)). Homemade food was also grouped in "cooked," "prepared but uncooked," or "uncooked and unprepared". Beverages were further classified into water, soft drinks, fruit juice-based drinks, fresh fruit juice, energy drinks, milk, milk-based drinks, coffee, tea, traditional beverages, or alcoholic beverages.

| Food sample collection and nutritional composition assessment
Samples of the most commonly available foods and beverages (hereafter referred to as "food," although beverages were included) of unknown composition (identified previously through the data on food availability) were collected for nutritional composition assessment. The most frequent homemade foods (n = 20) were piroshky, samsa, bread, kurut, kompot, bun, belyashi, sandwich, manty, keksi, chebureki, hamburger, boiled corn, prepared salads (namely carrot salad), ashlyamfu, cake, hot-dog, lagman, sausage roll, and porridge.
The most frequent industrial foods (n = 10) were sunflower seeds in natura, cookies, chewing gum, bun, chocolate, chips, croutons, lollypop, wafer, and candies, but given that the nutritional composition of sunflower seeds in natura, chewing gums, lollypop, and candies is not expected to vary significantly, the following most available industrial foods were selected instead: chalap, keksi, maksym (local beverage), and corn snacks. Four samples of each of these items were collected from different vending sites, in a total of 120, corresponding to 80 homemade and 40 industrial.
The selection of the vending sites where the food samples were collected was based on the list of eligible vending sites previously assessed. In each day, the GPS coordinates of 10 vending sites in a predefined market were randomly selected. One sample of homemade and one sample of industrial food, corresponding to one serving, were bought whenever possible at these vending sites. If not possible, a systematic selection procedure was followed, in which field researchers started moving north from that point and change direction clockwise (first east, then south, then west, then north again) whenever the limits of the study area or a physical barrier (such as a wall or a canal) were met, until reaching vending sites where these foods were available (World Health Organization, 2019). The 120 food samples were collected in 10 consecutive days.
After collection, samples were homogenized, weighted, and  (Cruz et al., 2013). Total carbohydrates plus fiber were estimated by difference (Horwitz, 2000). The energy value of samples was estimated after proximate analysis of food components (moisture, protein, total fat, and ash), performed in accordance with standard methods, as recommended by the Association of Official Analytical Chemists (AOAC) (Horwitz, 2000). The analytical results were the average of two determinations per food sample. A third determination was conducted only when the first two were not in agreement and, in this case, the average of the two concordant results was calculated. All the analytical results were expressed by serving size, in grams (g).

| Statistical Analyses
Absolute and relative frequencies (categorical variables) were used to describe the vending sites and food availability. Pearson's chisquare test was used to compare food availability and vending sites' characteristics. Markets were defined as the sampling units. The statistical analyses were conducted adjusting for the clustering at the sampling unit level.
Regarding the nutritional composition assessment, given that more than one sample was obtained for each food, in different vending sites, mean serving sizes per food, in grams (g), were calculated as the mean weight of the individual samples collected. The presented content of each nutrient in selected foods was obtained as a mean of the content in the individual samples. Per-serving levels of each nutrient were expressed in (g)/ serving. Results were presented for each food, as the mean and range of energy (Kcal) and macronutrients (g) per serving. The nonparametric Mann-Whitney U test was used for comparing the energy content of homemade and industrial foods. A critical level of significance (p) <0.05 was considered statistically significant.
Statistical analyses were performed using the software STATA® version 11.0 (StataCorp., College Station, Texas, USA).

| Food availability
Overall, no statistically significant differences were found regarding the availability of beverages between mobile and stationary vending sites (Table 1) nor between formal and informal stationary vending sites, except for coffee which was more frequently sold in formal physical setups (Table S1). Tea was the most frequently available beverage, found in 57.2% of the vending sites selling beverages, mostly mobile, followed by soft drinks (55.9%) and water (43.1%), mostly in stationary vending sites (Table 1).

| Nutritional composition
Industrial foods were more energy dense than homemade foods  (Table 2).
Particularly, the highest content in n-6 fatty acids was found in homemade lagman (11.2 g/serving), belyashi (9.1 g/serving), and hotdog (9.1 g/serving). N-3 fatty acids were only present in reduced amounts in most foods, reaching 0.5 g/serving in homemade manty.
The highest content in TFA was found in industrial wafers (3.8 g/ serving), but also in homemade traditional dishes such as manty (2.9 g/serving) ( Table 3).

| D ISCUSS I ON
The availability of street food in Bishkek is wide and diverse, including a variety of traditional and westernized foods and beverages.
A high variability in energy content, macronutrient, and lipid profile values was found among the most commonly available foods in this city, which may reflect heterogeneous ingredients and culinary practices.
The most frequently available street foods other than fruit were homemade and cooked, including a variety of traditional snacks, pastries and main dishes and a great share of more westernized foods (e.g., hamburger, hot-dog, and sausage roll). Homemade and industrial beverages, the latter including traditional and westernized options such as soft drinks, were also frequently found. This substantial offer of industrialized goods, resulting from greater access to new industrial foods and ingredients, formerly absent from the diet of this population is in line with recent observations from FAO, that Kyrgyzstan is transitioning from the stage 3 "Receding Famine," to stage 4 "Degenerative Disease" of the Nutrition transition (Food & Agriculture Organization, 2017). This stage is associated with a shift in lifestyles, namely an acculturation of a "western diet," rich in sweets, fats, and animal-source foods and low in fruit, vegetables, legumes, and whole grains (Kelly, 2016;Popkin & Gordon-Larsen, 2004).
Of the 30 street foods whose nutritional composition was analyzed in this study, the homemade foods had larger servings sizes and also energy content per serving, resembling the results found as lagman or ashlyamfu), and in fast-foods sandwiches, such as hamburger and hot-dog. Only a few of these foods may include vegetables or plant sources of protein, thus one can infer that most of this protein content is of animal origin. Similar findings have been reported in Mozambique, where stewed dishes containing meat or fish were the main sources of this nutrient (Sousa et al., 2018).
Altogether such findings reinforce that street food in urban Bishkek may have a central role in feeding citizens throughout the day and that it may highly contribute to the daily protein and energy intake of the population (Steyn et al., 2014).
The main contributors to the high energy content found in the analyzed street foods were carbohydrates and fat, as expected (Steyn et al., 2014). The foods with higher content in carbohydrates per serving were bread, main dishes, and fast food sandwiches such as hamburger or hot dog. A considerable amount of carbohydrates was also present in sweet snacks, both industrial (e.g., bun, sweet pastries, and wafers) and homemade (e.g., bun and cake). This could be due to a high sugar content, usually present in these foods (Becquey & Martin-Prevel, 2010;Steyn et al., 2014). (1 ingredients used, namely oils and fats, and in the culinary practices, as already described in the literature (Draper, 1996;Namugumya & Muyanja, 2012;Steyn et al., 2014). As such, it seems to be of utmost relevance to study the culinary practices and nutritional literacy of street food vendors and further promote healthier cooking techniques. The study of customers' motivations for buying street food and their nutritional literacy would be an additional benefit to attain the improvement of the nutritional quality of street foods, but also their acceptance by the population.
Also, even though the vending sites selling exclusively fruit were excluded from our analysis, the fruit and vegetables' availability in the study sample was low. A study conducted in the country has shown that the daily consumption of fruits and vegetables was positively associated with the number of establishments selling them (Goryakin, Rocco, Suhrcke, Roberts, & McKee, 2015). Other studies have found that increasing the availability of fruits and vegetables in street food vending sites, even when other products are available, was successful in increasing its consumption (Tester, Yen, & Laraia, 2010 This exploratory description of the street food environment in Bishkek highlighted its importance for this urban population in central Asia, given its wide expression throughout the city. The stepwise systematic approach for data collection and analysis aimed to provide an unbiased and comprehensive characterization, overcoming additional gaps identified in previous literature (Abrahale et al., 2019;Steyn et al., 2014). Although some results may not be generalized Mean SFA (min-max) (g/serving) Mean MUFA (min-max) (g/serving) Mean PUFA (min-max) (g/serving) Mean n-6 (min-max) (g/serving) Mean n-3 (min-max) (g/serving) Mean TFA (min-max) (g/serving)

TA B L E 3 (Continued)
to other communities, given culture-specific characteristics of the local food environment, the methodology has a wide potential to be adapted to different settings (World Health Organization, 2019), allowing comparison of results. The assessment of the nutritional composition of the most commonly available street foods was based on reliable chemical methods, also overcoming limitations of previous studies (Abrahale et al., 2019;Steyn et al., 2014). Nevertheless, these findings reflect availability rather than consumption and future insight on street food consumption in this setting would strength these assumptions.

| CON CLUS ION
Energy-dense traditional and westernized dishes and snacks, rich in saturated and trans fat were highly available in Bishkek, while healthy food options, such as fruit and vegetables were unfrequently available. Policies to prevent NCDs in Kyrgyzstan should consider the promotion of healthier street foods and the use of healthier ingredients and fats. The involvement of the food industry and street food vendors is essential to guarantee this change in food availability and preserve local traditions and gastronomy.

| DISCL AIMER
João Breda is staff member of the World Health Organization Regional Office for Europe. The author is responsible for the views expressed in this publication, which do not necessarily represent the decisions or stated policy of WHO.