Sociodemographic predictors of non-communicable diseases risk-related knowledge and behaviours: a cross-sectional study of in-school adolescents in a southern Nigerian State

Introduction the adolescence period is a significant phase in development of non-communicable diseases. Public health interventions that reduce risky behaviors among adolescents are beneficial across the life course. This study assessed the level of non-communicable diseases (NCDs´) risk-related knowledge, the prevalence of NCDs´ risk behavior, and the sociodemographic predictors of NCDs´ risk-related knowledge and behaviors among in-school adolescents in a Southern Nigerian State. Methods a cross-sectional study design was employed to assess the NCDs´ risk-related knowledge and behaviors among a random multistage sample of 607 students age between 10 and 19 years. Data were collected using an interviewer-administered semi-structured questionnaire adapted from the WHO STEPS questionnaire. Descriptive and inferential analyses of data collected were carried out using the IBM SPSS version 22 software. Results the mean age of the students was 14.7 (SD=1.52) years, 57.2% (n=347) of which were females, and 42.8% (n=260) were males. The proportion of students with good overall NCDs risk-related knowledge was 22.7% (n=138). Age, place of residence, family's socioeconomic status, and mother's level of education were significant sociodemographic predictors of good overall NCD risk-related knowledge. Among the students, 66.2% (n=402) self-report inadequate physical activity, 65.7% (n=399) self-report consumption of unhealthy diets, 29.2% (n=177) self-report current alcohol use, and 3.3% (n=20) self-report they were current cigarette smokers. Conclusion a significant proportion of the surveyed students had poor overall NCDs risk-related knowledge and engaged in NCDs risk behaviors. The relevant stakeholders concern with prevention of NCDs in government and non-governmental organizations should target adolescents in NCD control strategies in the study setting.


Introduction
Non-communicable diseases (NCDs) are a group of heterogeneous non-infectious, and not transmissible chronic conditions [1].They are costly and highly prevalent worldwide, making their prevention a public health priority [2,3].The global burden of NCDs is substantial as it significantly contributes to poor health and inequalities in health [4].The costs of NCDs in human, social and economic terms´ are largely preventable and avoidable [3,4].Unfortunately, about seventy percent of global deaths annually are NCDs-related, with a large proportion occurring prematurely [1,5].The enormous morbidity and mortality accrued to NCDs occur disproportionately in low-and middle-income countries (LMICs) [1,3,5].Over sixty percent of NCD-related deaths occur in LMICs [1].The reasons for the glaring inequity in NCDs-related burden distribution are multiple.They include poverty, weak health systems, rapidly growing populations, ignorance of NCDs-related risk factors and increasing adoption of unhealthy lifestyle choices [6][7][8].Non-communicable diseases are sometimes described as lifestyle diseases because many risk factors are behavioral and modifiable.Behaviors like physical inactivity, tobacco smoking, harmful use of alcohol, and consumption of an unhealthy diet rich in trans-fat and dietary salt but low in fresh fruits and vegetables are important modifiable ones [7,9].These behavioural risk factors are particularly patterned during adolescence, which is a significant phase in the life cycle of NCDs [10].Being a chronic disease, the latency period of NCDs is long, and the effects of engaging in NCD risk-related behaviors may not be immediately apparent.Thus, the adoption of healthy behavior during the adolescence period is a critical NCDs-preventive strategy [10].Public health interventions that reduce risky behaviors such as tobacco smoking, harmful use of alcohol, consumption of unhealthy food, and physical inactivity among adolescents are beneficial across the life course and will consequently reduce the burden of NCDs when they eventually become adults [10].In Nigeria, over one-fifth of the population are adolescents (persons aged between 10 and 19 years) [11].This large population of adolescents in Nigeria calls for action by the relevant stakeholders in the health and education sectors to implement school-based NCDs-preventive interventions.This study was conducted to assess the level of NCDs´ risk-related knowledge, the prevalence of NCDs´ risk behaviors, and the sociodemographic predictors of NCDs´ risk-related knowledge and behaviors among in-school adolescents in Delta State, Nigeria.

Methods
Study setting, design, and participants: the study setting was at nine secondary schools randomly selected across the three senatorial districts of Delta State in the oil-rich Niger Delta region of Nigeria.Delta State is a culturally diverse State with a total of 758 secondary schools with a total population of 262,242 students distributed across the three senatorial districts [12].The study design was a school-based cross-sectional survey carried out between December 2021 and May 2022 among a random multistage sample of apparently healthy students aged between 10 and 19 years enrolled in secondary schools in Delta State, Nigeria.The Health Research Ethics Committee (HREC), Delta State University Teaching Hospital provided ethical clearance for the study (HREC/PAN/2021/016/0326).The study conducted followed relevant guidelines and regulations.The authorities of the selected schools provide access permission before the commencement of the study.Participation in the survey was without coercion or inducement.The right to withdraw participation in the survey without untoward consequences was made known to the study participants and their parents/guardians.The study participants and their parents/guardians provided assent and written informed consent, respectively.
Sample size determination and sampling procedure: fisher's formula was employed to determine the minimum sample size.Because it was impossible to precise the extent of the various NCD-related risk behaviors in secondary school students in Nigeria, the prevalence was set at 50%.Based on a prevalence of 50%, an error margin of 5% and a standard normal variate at a 95% confidence level, the minimum sample size was estimated at 384 students.However, 607 students participated in the study.Sample selection was in two stages.In the first stage, a table of random numbers was used to select nine secondary schools (three per senatorial district) from a sampling frame of 758 secondary schools in Delta State [12].In the second stage, apparently healthy students between 10 and 19 years enrolled in the nine selected secondary schools, who gave assent and whose parents/guardians gave informed consent to participate in the study were proportionately allocated into sex strata by their class levels to allow for adequate representation.A table of random numbers was then used to randomly select students from the register in each class level by sex strata.Students who were not within the age bracket of 10 and 19 years were excluded from the sampling process.
Study instrument and data collection: data were collected using an interviewer-administered semistructured questionnaire adapted from the WHO STEPS questionnaire [13].The questionnaire comprised three sections which assessed the sociodemographic characteristics, NCDs riskrelated knowledge and behaviors of the students.
Outcome and independent variables: the outcome variables were NCDs risk-related knowledge and behaviors among the students.Knowledge of NCDs risk factors (behavioral and biomedical) was assessed with 11 questions on a 3-Likert scale (agree, disagree, not sure).Each correct response scored one, and every wrong response was zero.There was a maximum of 11 points on the NCDs' risk-related knowledge.Scores of 6 to 11 points were categorized as good overall knowledge, while a score of 0 to 5 points was categorized as poor overall knowledge.NCDs riskrelated behaviors were assessed by questions on cigarette smoking, alcohol consumption, diet/nutritional habits, and physical activity habits.Students who self-report that they currently smoke cigarettes were considered as smokers, while those who previously smoked or never smoked cigarettes were considered as nonsmokers.Students who self-report that they currently consume alcohol were considered as consumers of alcohol.In contrast, those who previously consumed alcohol or who have never consumed alcohol were considered as nonconsumers of alcohol.Students who self-report a daily intake of fewer than five servings of fruits/vegetables (raw or cooked) or regular high intake of saturated fat or sugary meals were categorized as consumers of unhealthy diets.Students who self-report less than 150 minutes/week of moderate to vigorous-intensive physical activity (walking, recreational exercise, cycling) were considered to have inadequate physical activity [14].The independent variables include age, sex, place of residence, family socioeconomic status and mothers´ educational attainment.Family socioeconomic status was determined using the scoring scheme for classifying socioeconomic status in Nigeria.The scoring scheme takes into cognisance both parents' educational attainment and occupation [15].
Statistical analyses: data generated was analyzed using the IBM SPSS version 22 software.Descriptive and inferential analysis of data collected was carried out.Categorical variables were summarized as frequencies and percentages (summarized data were presented in tables and figures).Bivariate and multivariate analysis using Pearson´s chi-square and binary logistic regression respectively was carried out.The statistical significance for both bivariate and multivariate analysis was set at p < 0.05.The 95% confidence intervals (CI) and p-values obtained were reported in two tail form and statistical significance determined at p-value less than 0.05.The binary logistic regression was performed to identify predictors of the outcome variables of interest (NCDs risk-related knowledge and behaviors).All variables significant during bivariate analysis using Pearson´s chi-square at a p-value < 0.2 were entered stepwise into the binary regression model to obtain the adjusted odds' ratio (AOR) of each factor on the outcome variable at 95% confidence interval.The model fitness was measured by the Hosmer-Lemeshow test.The statistical significance of the models (p-value ≥0.776) revealed that the binary logistic regression models (with independent variables included) were good fit to the data.

Results
The total number of students included in this study was 607, all of whom were included in the analyses.

Knowledge of NCDs behavioural and biomedical risk factors:
the proportion of students who had good overall NCDs risk-related knowledge was 22.7% (n=138), compared to those who had poor overall NCDs risk-related knowledge (77.3%: n=469) (Table 2).The knowledge of the behavioural and biomedical risk factors among the students was well below average across all the domains assessed (Figure 1, Figure 2).There was a statistically significant age difference (χ 2 = 7.79; df = 2; P = 0.02), urban-rural difference (χ 2 = 5.93; df = 1; P = 0.02), family SES difference (χ 2 = 21.69;df = 2; P < 0.001), and mothers' educational level difference (χ 2 = 13.98;df = 2; P = 0.001) in the level of NCDs risk-related knowledge among the students (Table 2).The students aged 17-19 years (AOR=2.06;95% CI: 1.53 -6.09) had two-fold increased odds of having good overall NCDs riskrelated knowledge compared to those aged 10-13 years.Students who were rural residents (AOR=0.53;95% CI: 0.34-0.83)had 47% decreased odds of having good overall NCDs risk-related knowledge compared to urban residents.Students whose family belonged to the upper SES (AOR=2.32;95% CI: 1.30-5.28)had two-fold increased odds of having good overall NCDs riskrelated knowledge compared to those whose family belonged to the lower SES.Students whose mothers had tertiary education (AOR=2.44;95% CI: 1.22-5.93)had two-fold increased odds of having good overall NCDs risk-related knowledge compared to those whose mothers had primary education.

Discussion
The overall NCDs´ risk-related knowledge among a significant proportion of the students was generally poor.An assessment of all the NCDs´ risk-related knowledge domains tested among them were below average.This is a worrisome development which calls for urgent intervention of all stakeholders in the education and health sectors involved in adolescent health in Nigeria.This is important because risk behaviors that promote NCDs thrive in the background of poor risk-related knowledge [16].The adolescence period is a critical transitory stage in life between childhood and adulthood.It is a period of life characterized by the development of mental capacity, and significant physical, emotional, and behavioral changes.Potentially, the patterns of NCDs risk behaviors are shaped during this period and tend to persist in adulthood [10].Therefore, the need to equip adolescents with the knowledge and skills to adopt healthy behaviors and choices before they become adults cannot be overemphasized.Overall, only one-fifth of the students had good overall NCDs´ risk-related knowledge.This observation corroborates with findings from previous studies conducted in Nigeria [17,18], and elsewhere in India [19], Sri Lanka [20], and Thailand [21] which have revealed similar poor levels of NCDs risk-related knowledge among in-school adolescents.Knowledge impacts health, and it is in turn determined by several factors that are sometimes complex and interrelated.In this study, sociodemographic factors such as age, place of residence, family's socioeconomic status and mothers' education were the predictors of good overall NCDs riskrelated knowledge among the students.Indeed, age, work and living environment, family income, and mother's educational attainment are welldescribed social determinants of health [22].The likelihood of having good overall NCDs risk-related knowledge was two-fold higher respectively among students who were older (late adolescents; 17-19 years), whose family belonged to the upper socioeconomic class, and whose mothers had tertiary education.Furthermore, students who were rural dwellers had forty-seven percent decreased likelihood of having good overall NCDs risk-related knowledge compared to their urban counterpart.It is not unlikely that the older students surveyed had amassed more knowledge over time as most of them were in higher school grades.It is also not unlikely that the students living in urban areas, whose family belonged to the upper socioeconomic stratum and whose mothers had tertiary education (more than 12 years of formal education) may have access to diverse sources of information aside from school-based health education, including the internet and other forms of electronic media, and thus acquire more knowledge.However, this study did not assess the sources of NCDs related information among the students.
In addition to the poor overall NCDs´ risk-related knowledge, a high rate of risk behaviors that favor the development of NCDs was observed among the students.About two-thirds of the students consumed unhealthy diets (majorly low in fruits and vegetables).This observation corroborates with findings from a previous study in Ghana conducted among in-school adolescents [23].Correspondingly, a meta-analysis of seventy-two countries drawn from all the World Health Organization (WHO) regions, and a United Nations Children's Fund (UNICEF)report respectively revealed high consumption of unhealthy diets (low in fruits and vegetables, and high in sugar and saturated fat) [24,25].Consumption of unhealthy diets among the students differed by family socioeconomic status.The likelihood of consuming unhealthy diets was two-fold higher among students whose family belonged to the lower socioeconomic stratum.Evidence have shown that individual choices of diet are constrained by socioeconomic factors including environmental, political, and cultural factors [4].Therefore, affordability of healthy diet is a critical factor for low incomes families in all countries.Consequently, low incomes families consume lower quantities of fruit and vegetables than high income families [4].Similarly, about two-thirds of the students were physically inactive.This was not surprising as over the past decades, sub-Saharan African adolescents have transitioned mainly from being quite physically active to being sedentary [26].Walking as a mode of transportation is often replaced with riding in motorized vehicles, and video/computer games have largely replaced engagement in outdoor sporting activities.Also, increased screen time (spent on computers, television and mobile phones) has adversely affected physical activity among adolescents.Other studies among adolescents outside the African continent have reported a preponderance of physical inactivity [27,28].Indeed, according to the World Health Organization (WHO), more than 80% of global adolescents lead sedentary lifestyles [29].In this study, students who were rural dwellers were more physically active compared to their urban counterpart.The likelihood of being physically active was two-fold higher among students who were rural dwellers.
In this study, about a third of the students admit to current alcohol use.The prevalence of current alcohol use observed among the students falls within the range previously reported among adolescents in Nigeria [30,31].Alcohol consumption, a self-harming behavior, is rife among adolescents [30][31][32][33].Despite policies against underage drinking, alcohol consumption remains the most common substance of abuse among young people in Nigeria [34,35].Alcohol consumption was significantly associated with age and sex.Students aged 14-16 years (midadolescents) and 17-19 years (late adolescents) had 2-fold higher odds of consuming alcohol compared with early adolescents (10-13 years).Furthermore, a significantly lower proportion of female students admit current alcohol use, and they had fifty-five percent decreased likelihood of alcohol use compared to their male counterpart.Previous studies had similarly reported male preponderance in alcohol consumption among adolescents in Nigeria [32,33].It is noteworthy that some studies report no sex difference in the prevalence of alcohol consumption among adolescents unlike adult drinkers with consistent male preponderance [31,36].Tobacco smoking, another harmful social habit, was reported by 3.3% of the students.The significant sociodemographic predictor for tobacco smoking in this study was sex, with female students having sixty-one percent decreased likelihood of smoking cigarettes compared to their male counterpart.The preponderance of male smokers in this study corroborates with the findings of a systematic review which revealed a higher prevalence of male adolescent smokers in Nigeria [37].The findings of our study should be interpreted considering the following strengths and limitations.The main strength of our study is that it targeted a relatively larger population of in-school adolescents.However, our study has some limitations: firstly, the self-report nature of NCDs risk behaviors by the students leaves room for reporter bias.Secondly, this study employed a cross-sectional descriptive design; thus, it is not possible to determine either causality or directionality of the variables analyzed.

Conclusion
A significant proportion of the students in this study had poor overall NCDs risk-related knowledge.Age, place of residence, family's socioeconomic status, and mother's level of education were significant sociodemographic predictors of good overall NCDs risk-related knowledge.In addition, a high rate of risk behaviors that favor the development of NCDs was observed among the students.There is need for concerted efforts by the relevant stakeholders that are concern with prevention of NCDs in government and non-governmental organizations to prioritize in-school adolescents in NCD control strategies in the study location.This calls for urgent implementation of interventions at all ecological levels that will improve adolescents' NCDs risk-related knowledge and equip them with the skills to adopt healthy lifestyles and choices.This consequently will deter their engagement in unhealthy behaviors.

Figure 2 :
Figure 2: knowledge of biomedical-related non-communicable diseases risk factors among the surveyed adolescents

Figure 3 :
Figure 3: self-reported non-communicable diseases risk behaviours among the surveyed adolescents

Table 1 :
sociodemographic characteristics of the surveyed adolescents

Table 2 :
predictors of good overall noncommunicable diseases risk-related knowledge among the surveyed adolescents

Table 3 :
predictors of consumption of healthy diet among the surveyed adolescents

Table 4 :
predictors of physical activity among the surveyed adolescents

Table 5 :
predictors of alcohol consumption among the surveyed adolescents

Table 6 :
predictors of cigarette smoking among the surveyed adolescents Figure 1: knowledge of non-communicable diseases behavioural risk factors among the surveyed adolescents Figure 2: knowledge of biomedical-related noncommunicable diseases risk factors among the surveyed adolescents Figure 3: self-reported non-communicable diseases risk behaviours among the surveyed adolescents 13.World Health Organization.WHO STEPS surveillance manual: the WHO STEPwise approach to chronic disease risk factor surveillance/non-communicable diseases and mental health, World Health Organization.Accessed May 06, 2021.14. James S. National Survey for Wales, 2019-20: adult lifestyle.Policy commons.Google Scholar 15.Ibadin MO, Akpede GO.A revised scoring scheme for the classification of socio-economic status in Nigeria.Niger J Paediatr.2021Feb 4;48(1): 26-33.Google Scholar 16.Tohi M, Bay JL, Tu'akoi S, Vickers MH.The Developmental Origins of Health and Disease: Adolescence as a Critical Lifecourse Period to Break the Transgenerational Cycle of NCDs-A Narrative Review.Int J Environ Res Public Health.2022 May 16;19(10): 6024.PubMed| Google Scholar 17. Shabi IN, Oyewusi FO.Health literacy and internet health information use among inschool adolescents in Osun State, South-West, Oyewole BK, Animasahun VJ, Chapman HJ.Tobacco use in Nigerian youth: A systematic review.PLoS One.2018 May 3;13(5): e0196362.PubMed| Google Scholar

Table 1 :
Sociodemographic characteristics of the surveyed adolescents

Table 2 :
predictors of good overall non-communicable diseases risk-related knowledge among the surveyed adolescents Variables

Table 3 :
predictors of consumption of healthy diet among the surveyed adolescents Variables

Table 4 :
predictors of physical activity among the surveyed adolescents Variables

Table 5 :
predictors of alcohol consumption among the surveyed adolescents Variables