Study aim

The main aim of the study is to examine the seasonal variation in food security, lifestyle factors, neighbourhood environmental factors, and nutritional status of the urban poor adolescents in Kuala Lumpur, Malaysia. To achieve the aim, the specific research objectives to be investigated in the study are as follows:

  1. (a)

    To determine the food security, lifestyle factors, neighbourhood environmental factors, and nutritional status of the urban poor adolescents in Kuala Lumpur, Malaysia.

  2. (b)

    To determine the seasonal differences in food security, lifestyle factors, and nutritional status among urban poor adolescents in Kuala Lumpur, Malaysia.

  3. (c)

    To determine the effects of season on food security, lifestyle factors, neighbourhood environment, and its effects on the nutritional status of the urban poor adolescents in Kuala Lumpur, Malaysia.

Study design

This is a cohort study that stemmed from two cross-sectional studies that will be carried out during the onset of the Northeast Monsoon season and will be followed up during the Southwest Monsoon season to capture a different season. Recruitment of the subjects will begin upon obtaining ethical clearance from the Institutional Ethics Committee of UCSI University [Reference code.: IEC-2021-FAS-03)] and permission to conduct the study from Kuala Lumpur City Town Hall (DBKL) [Reference code: DBKL/JPKKB/SPZ/29–3 Jld 13 (48)].

Study setting and sampling method

This study followed a multistage stratified random sampling design (Fig. 1). First, a complete list of the People’s Housing Programme (Perumahan Awam, PA or Program Perumahan Rakyat, PPR) (n = 53), located in Kuala Lumpur is obtained from the Kuala Lumpur City Town Hall (Dewan Bandaraya Kuala Lumpur, DBKL) and further stratified into four zone areas based on the zone office management by DBKL. Subsequently, three flats from each zone area will be randomly selected (n = 12). A convenience sampling method will be applied for the recruitment of the subjects whereby the residential president of PA/PPR shared e-flyers through social media platforms with the residents for the recruitment of the adolescents via the parents. In addition, adolescents will be also approached by door-to-door recruitment methods at all the residential blocks. All adolescents from all ethnicities within the selected flats are invited to participate in the study. The purpose and procedures of the study will be explained to the adolescents and their parents/guardians. A detailed information sheet will be provided for a better understanding of the study. Parents/guardians will be requested to sign a written informed consent to allow their child’s participation, and assent will be obtained from adolescents to indicate their willingness to participate before the commencement of the study. Data collection will be conducted by enumerators who have been trained, monitored, and supervised throughout the study. Each adolescents will be given a unique identification number to ensure anonymity. The flow diagram of the study is depicted in Fig. 2.

Fig. 1
figure 1

Multistage stratified random sampling method of the study

Fig. 2
figure 2

Data collection procedure

Inclusion and exclusion criteria

There are a few inclusions and exclusion criteria set for the adolescents to be recruited and qualified to participate in this study. The criteria are as below:

Inclusion criteria

  1. (a)

    Malaysian adolescents aged between 10 to 17 years old.

  2. (b)

    Adolescents currently living in low-cost high-rise flats in Kuala Lumpur at least for the past 12 months.

  3. (c)

    Healthy adolescents with no physical and learning disabilities, as confirmed by parents.

Exclusion criteria

  1. (a)

    Adolescents who are under any special diet or diet restriction.

  2. (b)

    Adolescents who have chronic medical problems such as heart disease, diabetes, cancer, etc.

  3. (c)

    Adolescent females who are pregnant or lactating.

  4. (d)

    Adolescents who were infected with COVID-19 at the point of the study.

Sample size calculation

The sample size will be calculated for the baseline cross-sectional analysis of this study. The sample size is determined using single proportion formula [41] for estimating prevalence. Based on the proportion of adolescents who were obese in Kuala Lumpur, Malaysia (15.8%) [14], with the desired level of confidence at 95% and desired level of precision of 0.05, a minimum of 204 subjects are required for the study. In consideration of the non-response rate and attrition rate among adolescents, the sample size is increased (+ 25%) to 256 adolescents.


Climate data will be collected from Malaysian Meteorological Department (MMD) to confirm the variation in meteorological parameters. Spatial food and built environment data will be collected using geographic information system (GIS) for urban poor adolescents residing in PA/PPR in Kuala Lumpur, Malaysia. Two self-administrated Malay language questionnaires, a parental questionnaire and a adolescent questionnaire, will be developed and pre-tested. All parents and adolescents will be given instructions and explanations of the questionnaire variables. Anthropometric, physical fitness level, and haemoglobin level assessment will be conducted after the consent form and questionnaire are filled up. To ensure the accuracy of the data collected, participants will be re-interviewed and double check of questionnaires. Table 1 shows the details of the variables assessed during baseline and follow-up in this study.

Table 1 Summary of data collection and timeline

Climate data

Climate data will be assessed through the monthly mean data for meteorological parameters, including rainfall (mm), temperature (°C), and relative humidity (%) from the Malaysian Meteorological Department from the monitoring stations located in Petaling Jaya, FRIM Kepong, KLIA Sepang, Subang, and Hulu Langat. Climate data during the actual data collection period will be assessed for both Southwest and Northeast monsoon seasons. Besides, climate data of selected meteorological parameters for the past 10 years will also be extracted to identify the climate trend for both monsoon seasons and to confirm any climate variation over the years.

Spatial food and built environment

Spatial food and built environment are defined using a buffer zone area of defined geographic distance from the low-cost flats. QGIS 3.4 Madeira will be used to generate walking-distance circular buffers for a distance of 1000 m on-street network away from residential flats. These are the common buffer sizes applied in active research transport as this is the distance that can be covered by a 10–15-minute walk [42]. The exact address of each selected flats is required, which will be converted into geographic coordinates. The spatial data of all food and built establishments within the residential neighbourhood environment are geocoded to obtain latitude and longitude, where on-site data collection will be performed using a Qmini A7 High-precision (country that produce it) GIS handheld device. Types of food establishments included are supermarkets (e.g., AEON, GIANT, TESCO, large local wet markets, etc.), fast-food restaurants (e.g., KFC, Pizza Hut, McDonald’s, Texas Chicken, etc.), restaurants, convenience stores (e.g., 7-eleven, KK mart, Shell Mart, etc.), street food and specialty food stores (e.g., bakeries, fruit and vegetables, gourmet, meat and fish markets). Built establishments for physical activity included parks and public recreational facilities (e.g., soccer fields, basketball courts, community centers, pools, and playgrounds). Besides, the location of bus stops and the station will also be collected. The assumption is made that no substantial changes in the distribution and number of this food and built establishments throughout the study.

Using the geocoded data, the following spatial neighbourhood food and built environment indicators will be created: (1) density of food and built establishments within 1000 m network buffer zones; and (2) proximity to food and built establishments (closest distance, via the street network, from PA/PPR to each type of food and built establishment). The density and proximity to the food and the built establishments will be determined in QGIS software using the Buffer, Mapping, and Count Point analysis tools.

Adolescent’s questionnaire and assessment

Physical development

Pubertal Development Scale (PDS) will be used to assess the pubertal status of adolescents [43, 44]. A total of five items are assessed in PDS to rate the growth spurt, body hair growth, and skin changes using a four-point scale: “not yet started” (1), “barely started” (2), “definitely started” (3), “seems complete” (4), and “I don’t know” (0). On a similar scale, boys will be asked to rate their development of facial hair and voice change, whereas girls rated their breast development and whether they have reached menarche, present (4) and absent (1). Based on the sum of the scores, adolescents will be categorized into pre-pubertal, early pubertal, mid-pubertal, late pubertal, and post-pubertal [43, 44].

Physical activity and fitness level

Physical activity level will be assessed subjectively using the physical activity questionnaire for older children (PAQ-C) [45]. The Malay version of PAQ-C had been validated among adolescents aged 10–17 years old [46]. A total of 10 items are assessed in the PAQ-C to assess the general levels of physical activity of adolescents over the past seven days and will be scored based on a 5-point scale ranging from ‘no activity’ (1), and ‘7 times or more’ (5). The mean total physical activity score will be calculated and classified into low (1–2.33), moderate (2.34–3.66), and high (3.67–5) [47]. Additionally, screen time such as watching television, using a computer, and playing video games of an adolescent during weekends and schooling days will be assessed as well and further classified into screen time less than two hours and more than two hours [14].

In addition, a subsample of 50 adolescents will be randomized for objective measurement of physical activity level using a pedometer (YAMAX Digi-Walker SW-700) [48]. Adolescents are assigned a pedometer, elastic belt, and a 4-day pedometer diary. The pedometer will be distributed on the same day of data collection and adolescents are instructed to self-monitor their physical activity for four consecutive days with at least one weekend day [49] by wearing the pedometer at the waist from the time they wake up until they go to bed at night, excluding water-based activities and bathing time. Subjects are requested to record the day-end values for pedometer steps, other readings displayed, and non-ambulatory activities (e.g., swimming, cycling, etc.) and to reset to zero every day. A reminder will be sent to the subjects or their parents/guardians at least once throughout the data collection period.

The physical fitness test will be conducted using a modified Harvard Step Test [50]. This method was used in previous local studies [51, 52]. Subjects are requested to wear light clothes in order to carry out the test. Subjects are instructed to step up and down on a 30 cm high step box with both feet for a maximum of five minutes or until fatigue compelled him or her to stop and rest. Immediately, the heart rates at zero, one, and two minutes of rest will be measured using an automatic blood pressure monitor (OMRON HEM-7120 Automatic Blood Pressure Monitor, OMRON HEALTHCARE Co., Kyota, Japan) and recorded, as well as the total duration of exercise. Physical fitness scores will be calculated and classified into “unacceptable” (< 65), “marginally acceptable” (65–79), and “acceptable” (≥80) [50].

Dietary assessment

A two-day 24-hour dietary recall which comprised one weekday and one weekend, is used to determine the food consumption and nutrients intake of subjects. The dietary recall method is carried out through face-to-face interviews. Subjects will be asked to recall and report foods and beverages consumed. Detailed information including types, brand names of commercial products, food preparation method as well as the amount consumed with the aid of a set of standard household measurements such as spoons, teaspoons or tablespoons, cups, glasses, bowls, and plates, along with the use of the Food Atlas book [53]. Dietary intake data will be analyzed using the Nutritionist Pro™ Diet Analysis Software Version 4.0 (Axxya Systems, Stafford, TX, USA) based on the food listed in the Malaysian Composition Database Nutrient Composition of Malaysian Food. This was then further evaluated for the overall diet quality of the subjects using the Standardized Malaysian Healthy Eating Index (S-MHEI) [54].

The S-MHEI consisted of 11 components, including eight food groups (total grains, whole grains, fruits, vegetables, meat/poultry/eggs, legumes/nuts, and milk/milk products) and three nutrient groups (total fat, sodium, and sugar). The scoring of these components is calculated based on the serving size and nutrient intake recommended by the Malaysian Dietary Guidelines [55] and Malaysian Dietary Guidelines for Children and Adolescents [56]. The score of each component ranged from 0 to 10, which is calculated proportionately for the in-between whole-number responses. However, as total grains and whole grains are from the same food group, the maximum score provided to each component is five to avoid scoring overlap. The total S-MHEI score is obtained by summing up the score of each component with a score range of 0 to 100. A higher S-MHEI score indicates better diet quality. Based on the maximal dietary intake score of 100, the subjects will be categorized as ‘good diet quality’ (> 80%), ‘diet quality needs an improvement’ (51–80%) and ‘poor diet quality’ (< 51%) [54].

The frequency of meal consumption will be assessed using six items from the Eating Behaviours Questionnaire (EBQ) [57]. Adolescent is required to indicate the frequency of each meal consumption and snacking behaviour, including breakfast, morning snack, lunch, evening snack, dinner, and supper, based on an 8-point scale ranging from never (zero) to every day (7 times) in a week.

Changes in lifestyle-related behaviour during the COVID-19 pandemic

The 20 items short questionnaire developed by Kumari et al. [58] will be used to assess the changes in lifestyle-related behaviour of adolescents during the COVID-19 pandemic. Subjects will be asked in assessing the changes in dietary habits (intake, meal pattern, and snack consumption), physical activity (duration and type), and sleep pattern (duration and quality), using a 5-point Likert scale ranging from “significantly decreased” (+ 2) and “significantly increased” (− 2). It is noted that some of the items (Item 4–5, 11–16, 19) are reverse scored. Besides, items 3 and 18 are scored as “grossly similar” (0), “slightly increased/decreased” (− 1), and “significantly increased/decreased” (− 2). The total score will be calculated and a higher score indicating shifting toward a healthy lifestyle during the COVID-19 pandemic.

Anthropometric measurements

All equipments will be calibrated before the measurements are performed. The body height of subjects will be measured using a calibrated vertical stadiometer (Seca 213 portable stadiometer, SECA GmbH & Co., KG. Hamburg, Germany), whereas weight and body fat percentage will be measured using the Bioelectrical Impedance Analyzer (OMRON HBF-375 Karada Scan Body Composition Monitor, OMRON HEALTHCARE Co., Kyota, Japan). Both height and weight measurements are recorded to the nearest 0.1 cm and 0.1 kg, respectively. The nutritional status of adolescents will be determined based on Z-scores for height-for-age (HAZ) and BMI-for-age (BAZ) of WHO growth reference 2007 for adolescents between 5 and 19 years old using WHO AnthroPlus version 1.0.3 software (WHO, Geneva, Switzerland). The classification as followed, stunting (HAZ < -2SD), normal height (HAZ > -2SD), thinness (BAZ < -2SD), normal weight (−2SD ≤ BAZ ≤ +1SD), overweight (+1SD < BAZ ≤ +2SD), and obesity (BAZ > +2SD) [59]. The body fat percentage will be classified using percentile scores for sex and age based on the findings by McCarthy et al. [60] into underfat (<2nd percentile), normal (2nd – 85th percentile), overfat (>85th – 95th percentile), and obese (>95th percentile).

Waist circumference will be measured at the midway between the lowest rib and the superior border of the iliac crest as recommended by WHO [61] using a non-elastic SECA 201 Ergonomic circumference measuring tape (DECA GmbH & Co., Hamburg, Germany) to the nearest 0.1 cm. A cut-off point of the 90th percentile of waist circumference percentile classification for Malaysian children and adolescents [62] will be used to identify abdominal obesity. For adolescents aged 17 years and above, WHO/IASO/IOTF (2000) for Asian cut-off [63] will be used instead. Abdominal obesity was defined as ≥90 cm for men and ≥ 80 cm for women. Mean average values will be used for data analysis.

Anaemia status assessment

A haemoglobin level test will be performed to identify the anaemia status of each adolescent by using the HemoCue haemoglobinometer (HemoCue® Hb 201+ System, Angelholm, Sweden), point-of-care testing on capillary blood samples. Anaemia status and level of severity are determined according to the age and gender-specific haemoglobin cut-off point as defined by WHO [64]: (i) < 115 g/L for 5 to 11 years of age; (ii) < 120 g/L for 12 to 14 years of age and non-pregnant women (15 years of age and above); (iii) < 130 g/L for men (15 years of age and above).

Parental questionnaires

Sociodemographic background

Sociodemographic data will be collected for both adolescents and parents using a self-administrated questionnaire. Adolescent information includes age, sex, date of birth, ethnicity, household size, number of school-going children, duration of living in PA/PPR, medical illness or disabilities, medication, and supplement use, whereas parental/guardian information includes age, educational attainment, marital status, occupation, and monthly household income.

Household food security status

Food security status is measured using the United States Department of Agriculture (USDA) 18-item Household Food Security Survey Module (HFSS) which captures all severity levels of household food security and children’s condition in the household [65]. This section of the items will be filled by the parents or the guardians as parents or guardians will be responsible for purchasing or preparing the food for the household and the adolescent. A total of 18 items will be asked and answers of “yes”, “often true”, “sometimes true,” “almost every month,” and “some months but not every month” are coded as affirmative (1 point). The sum of affirmative responses to the 18 questions is referred to as the household’s raw score. With a maximum score of 18, subjects will be categorized into high food security (0), marginal food security (1–2), low food security (3–7), and very low food security (8–18). Subjects will be further dichotomized into two groups, the food secure group (high or marginal food security) and the food insecure group (low or very low food security) [65].

Perceived food environment

Perceived Nutrition Environment Measures Survey (NEMS-P) by Green and Glanz [66] will be used to assess the parents’ perceived food environment. The NEMS-P core components that measured community (seven items), consumer nutrition environment (24 items), and home food environment (seven items) will be examined. Composite scores will be calculated for each domain.

Community nutrition environments assess the store and restaurant accessibility, including mode of travel to the store, distance travelled to store/restaurant, as well as store motivation (importance of store proximity to home and other places where time is spent) and restaurant motivation (importance of convenience).

Consumer nutrition environments can be further categorized into store consumer and restaurant consumer nutrition environment. Store consumer nutrition environments include 17 items on store availability of healthy food choices, store motivation, price of fruits and vegetables, placement or promotion of unhealthy/healthy items, and nutrition information. Subjects will be asked to indicate their level of agreement on the statements using a 5-point scale ranging from “strongly disagree” (1) through “strongly agree” (5). Store motivation such as the importance of selection, quality, and price of food is assessed using a 4-point scale ranging from “not at all important” (1) through “very important” (4). Pricing for fruits and vegetables is assessed using response choices of very expensive (1) to very inexpensive (4). For the restaurant consumer nutrition environment, a total of seven items will be assessed on the availability of healthy options, promotions, and the cost of healthy options at restaurants using a 5-point scale ranging from “strongly disagree” (1) through “strongly agree” (5).

The home food environment includes items on the availability and accessibility of healthy and unhealthy food in the home. Subjects will be asked to indicate the availability of fruits/vegetables, healthy food, and unhealthy food based on the given list. Besides, the accessibility of healthy and unhealthy food will be assessed using a 4-point scale of “never or rarely” (1) to “almost always” (4).

Perceived built environment

Physical Activity Neighbourhood Environmental Survey (PANES) by Sallis et al. [67] will be used to assess the parents’ perceived environmental support for physical activity. A total of 17 questions will be asked in measuring the attributes of neighbourhood built and social environments hypothesized or known to be related to physical activity using a 4-point Likert scale ranging from “strongly disagree” (1) through “strongly agree” (4). The items can be categorized as residential density (item 1), land use mix (items 2 and 17), transit access (item 3), pedestrian infrastructure (items 4 and 13), bicycling infrastructure (items 5 and 14), recreational facilities (item 6), street connectivity (item 12), crime safety (items 7 and item 16), traffic safety (items 8 and 15), pedestrian safety (item 9), aesthetic (item 10). The mean total score will be calculated and the higher the score, the higher the perceived neighbourhood environment support for physical activity.

Data management and analysis

Data analysis will be performed using IBM SPSS 25.0 (SPSS Inc., Chicago, IL, USA). A test of normality will be performed using a skewness value (between − 2 and + 2) for the continuous variables. The univariate analysis will be used to report descriptive data including mean score, median, and standard deviation for continuous variables and percentage and frequencies for categorical variables. Independent-samples t-test and One-way ANOVA will be used to compare differences between groups for continuous variables and chi-square analysis will be conducted to examine associations for categorical variables. Correlation test will be used to see if correlation exist between variables. A paired-samples t-test will be conducted to determine the statistical differences between the two monsoon seasons. Logistic regression will be performed separately for seasons to determine the relationship between nutritional status and selected variables (Food security, lifestyle factors and neighbourhood environment factors). The Generalized Linear Mixed Model will be used to determine the seasonal effects on food security, lifestyle factors, neighbourhood environmental factors, and its effects on nutritional status. The significance level for all analyses is set at p < 0.05.

The status and timeline of the study

The first adolescent was enrolled on the 1st of November 2021. The recruitment process and baseline data collection were completed on the 30th of April 2022. At the time of submission, all de-identified subjects data are kept and manage but have not yet been processed for analysis.


By admin

Leave a Reply

Your email address will not be published. Required fields are marked *