Over the past few decades, childhood obesity has emerged as one of the most pressing global public health challenges. Rapid technological advancement has fundamentally altered the daily routines of children worldwide, leading to sedentary lifestyles, disrupted sleep patterns, and a host of associated metabolic consequences. Research around the world has estimated that about 38.3 million children under the age of five are already suffering from obesity, with many becoming obese adults and perpetuating an inter-generational cycle of metabolic ill health.

There has been a clear epidemiological association between poor sleep and the development of obesity in adolescence in developed countries. It turns out that in a large study of 66,817 adolescents in China, poor sleep quality both on the short and long ends were significantly linked to overweight and obesity.In a large study of 66,817 Chinese adolescents, it was found that both a short and long sleep duration were significantly associated with overweight and obesity compared with a sleep duration of 7.0-8.9 hours per day. A study of Korean adults also found that short sleep duration was a risk factor for obesity. The metabolic burden being faced by the developing economies of Asia is reflected in the extent of overweight and obesity amongst adults in India, which is estimated at 10-30%.

Statement of the Problem

Although the relationship between sleep deprivation and childhood obesity has been observed in other countries around the world, there is not enough information available in the local context to show that there is such a relationship. Urban teenagers in Pakistan are facing a rapid nutritional transition where they are exposed to sedentary screen time, irregular sleep patterns, academic stress and consuming unhealthy food. Data on medical students showed that 75% of medical students were sleeping less than six hours during weekdays, indicating the extent of sleep defecits in the country's youth, revealed data from Islamabad Medical College. Increasing prevalence of overweight and obesity has been reported in children of all ages attending the schools of Karachi, Lahore, Hazara Division and Swat.The incidence of obesity in school going children has been reported to increase in different studies in Karachi, Lahore, Hazara Division and Swat ranging from 4.78% to 21.8% of overweight and obesity combined.

To the authors' knowledge, however, no study had investigated the association between sleep time and childhood obesity in school-going children in the largest urban center, Karachi, Pakistan. This is a concerning gap in evidence as there is a lack of understanding of the local determinants of childhood obesity, which makes it difficult for healthcare providers and policymakers to have the evidence base needed to design effective childhood obesity preventive interventions. This is exacerbated by Pakistan's nutritional epidemiology of under nutrition and over nutrition, and isolation and quantification of the independent effect of sleep on obesity risk is of particular importance.

Rationale of the Study

This study had three related driving factors: scientific, clinical, and public health.

The scientific literature has documented scientific links between sleep loss and obesity via several mechanisms: disruption of the regulatory systems of appetite-regulating hormones (leptin and ghrelin), elevated cortisol levels, insulin resistance, and altered energy expenditure but none of the Pakistani adolescents had been found in the literature directly. These associations can be modified locally by biological, cultural and socioeconomic context, which can not be predicted from Western or East Asian data.

Family physicians and primary care providers in Pakistan have a distinct advantage in screening and treating sleep related risk factors for overweight and obesity in children at their routine primary care visits. Such counselling, however, without local evidence, has no epidemiological basis to encourage behaviour change in families. This study was undertaken to provide that evidence.

A public health approach is to intervene in school to address sleep hygiene in conjunction with nutrition and physical activity to achieve a broad population impact at an affordable cost. Knowing the level of association (with odds ratio, and stratified by age, gender, grade and parental education) gives a quantitative basis for the design of such programmes. This study is valuable as it provides timely evidence for policy development in Pakistan, where the National Health Vision and school health programs are lacking in addressing adolescent sleep health.

Therefore, this study was conducted to find the association between sleep duration and obesity among 6th, 7th and 8th grade students of Gulshan-e-Iqbal Town Karachi and to assess the variation of this association (sleep and obesity) according to age, sex, grade, family history of obesity, and educational status of parents

Study Design and Setting This case-control study was conducted from 4ᵗʰ December 2018 to 4ᵗʰ April 2019 in Gulshan-e-Iqbal Town, Karachi, Pakistan. The study was carried out in three governments and three private secondary schools. Inclusion of both government and private schools was intended to ensure representation of students from diverse socioeconomic backgrounds. Sample Size and Sampling The sample size was calculated using Epi-Tools software. Based on existing literature, the expected odds ratio was 2.87 with a 20% proportion of controls (non-obese) having reduced sleep exposure. A 95% confidence level was applied for both groups. Using stratified random sampling, 252 participants — 126 cases and 126 controls — were enrolled. Inclusion and Exclusion Criteria Students enrolled in Grades 6, 7, and 8, aged 12–15 years of either sex, were eligible. Children with a BMI above the 95ᵗʰ centile were classified as cases (obese); those with a BMI below the 85ᵗʰ centile were classified as controls (non-obese). Participants not meeting these criteria or who declined to participate were excluded. Ethical Approval and Informed Consent Ethical approval was obtained from the Ethical Review Committee of the Aga Khan University Hospital and the College of Physicians and Surgeons of Pakistan prior to commencement. Verbal informed consent was obtained from all participants. Data Collection Students were approached during free periods and a structured questionnaire was self-administered in 10–15 minutes, with the Principal Investigator present throughout. The questionnaire captured demographic information (age, gender, grade) and detailed questions about sleep duration on weekdays and weekends separately. Sleep of ≥8 hours was classified as normal; <8 hours as poor. Average sleep duration was computed as: (Weekday sleep × 6 + Weekend sleep × 1) ÷ 7. Anthropometric measurements were taken by the Principal Investigator using calibrated scales at all schools to minimise inter-observer variability. BMI was plotted against age- and sex-specific growth charts for centile classification. Statistical Analysis Data were entered and analysed using SPSS version 27. Descriptive statistics summarized baseline characteristics. Multivariate logistic regression assessed the independent effect of variables significant in univariate analysis. Odds ratios with 95% CIs were calculated; p ≤0.05 was considered statistically significant. Effect modifiers were controlled through stratification and post-stratification chi-square testing.

Baseline Characteristics

A total of 252 children were enrolled: 126 obese cases and 126 non-obese controls. In the case group, the mean age was 13.01 ± 0.79 years (range: 12–15); in the control group, 13.49 ± 0.40 years (Table 1). Among cases, 80 (63.5%) were male and 46 (36.5%) female; among controls, 65 (51.6%) were male and 61 (48.4%) female. Grade-wise, 33 (26.2%), 60 (47.6%), and 33 (26.2%) case-group children were in Grades 6, 7, and 8 respectively, compared to 24 (19%), 54 (42.9%), and 48 (38.1%) in the control group.

Sleep Duration and Obesity

Poor sleep duration (<8 hours) was present in 85 (67.5%) cases versus 60 (47.6%) controls, a statistically significant difference (p = 0.01; OR = 2.27). Obese children were therefore 2.27 times more likely to experience inadequate sleep than their non-obese peers (Table 2). Figure 1 illustrates this distribution visually.

Figure 1. Poor Sleep Duration Distribution: Obese Cases vs. Non-obese Controls (OR = 2.27; p = 0.01)

Stratified Analyses

When analyzed by age strata (Figure 4; Table 3) the association between obesity and poor sleep was borderline significant for children 12–13 years (OR = 2.25; p = 0.05) and non-significant for children 14–15 years (OR = 2.28; p = 0.20); possibly due to a smaller number of children in the latter age group, and

hence less statistical power.

Figure 4. Prevalence of Poor Sleep Duration by Age Group: Cases vs. Controls

-stratified results (Figure 2; Table 4) revealed a striking difference. In females, the association was highly significant (OR = 3.21; p = 0.01), while in males it fell short of significance (OR = 1.80; p = 0.08). Girls therefore showed a considerably stronger link between obesity and poor sleep.

Figure 2. Odds Ratios for Poor Sleep Duration by Gender (Obese vs. Non-obese)

Grade-level analysis (Figure 3; Table 5) demonstrated elevated odds ratios across all three grades — Grade 6 (OR = 2.66; p = 0.07), Grade 7 (OR = 2.40; p = 0.10), and Grade 8 (OR = 2.60; p = 0.07) — though none individually reached statistical significance, likely owing to reduced subgroup sizes.

Figure 3. Prevalence of Poor Sleep Duration by Grade Level: Cases vs. Controls

Family history of obesity also was important. Among children with a positive family history, the OR was 2.06 (p = 0.01); among children without a positive family history, the OR was 3.21 (p = 0.01), indicating that sleep deprivation is a risk factor for obesity even in children who would not be at risk based on their genetic makeup (Table 6). The strongest association was seen with maternal education level, where children of highly educated mothers were most likely to have late bedtimes (OR = 4.45; p < 0.001), likely due to academic-driven bedtimes and screen time in higher socioeconomic class families (Table 7).

TABLES

Table 1. Descriptive Statistics of Age in Case and Control Groups (n = 252)

Table 2. Sleep Duration According to Obesity Status (n = 252)

Table 3. Sleep Duration According to Age Group (n = 252)

Table 4. Sleep Duration According to Gender (n = 252)

Table 5. Sleep Duration According to Grade (n = 252)

Table 6. Sleep Duration According to Family History of Obesity (n = 252)

Table 7. Sleep Duration According to Educational Status of Mother (n = 252)

Table 8. Sleep Duration According to Educational Status of Father (n = 252)

Pakistan, like many other middle-income countries, is undergoing a rapid nutritional transition characterised by shifts in dietary patterns and a marked decline in physical activity among adolescents. These changes are directly contributing to rising rates of overweight and obesity in the population, particularly in urban centres like Karachi where sedentary, screen-driven lifestyles are increasingly prevalent. Our study adds to this growing body of evidence by demonstrating a significant and independent association between poor sleep duration and obesity among school-going children in Karachi. In our study, 67.5% of obese children reported poor sleep duration, compared to 47.6% among non-obese controls. The overall OR of 2.27 indicates that obese children are more than twice as likely to experience inadequate sleep. International evidence is consistent: a study in China of 66,817 adolescents revealed an association with being short or long sleeper with increasing obesity rates compared to an optimal range of 7.0–8.9 hours. An additional study in Korea also found that short sleep was a big risk factor for obesity in adults. The results that are presented separately for women and men are especially interesting. The larger association in females (OR = 3.21; p = 0.01) than males (OR = 1.80; p = 0.08) is similar to the results of a Chinese longitudinal study which found that sleep duration had gender-specific metabolic effects. This difference is likely due to the hormonal regulation of sleep systems and the extra sociocultural challenges for adolescent girls in Pakistan, such as academic burden, domestic duties and usage of social media. The association is stronger in younger children (12–13 years; OR = 2.25; p = 0.05) further supporting the need for early preventive intervention and highlighting the start of the sleep-obesity relationship early in the adolescent years. A non-significant result in the stratum 14–15 (p = 0.20) could be due to decreased power resulting from fewer numbers in the groups or due to the true absence of action. It is also clinically relevant, because the children who did not have a family history of obesity had a higher OR (3.21) than those who had a family history (2.06), which shows that sleep deprivation is a risk factor for obesity even in less predisposed children. This is consistent with the mechanistic literature in which it is proposed that sleep deprivation dysregulates leptin and ghrelin (the hormones that control appetite), raises cortisol, alters glucose metabolism, and lowers levels of motivation for exercise — all of which can make people gain weight, regardless of their genetics. The relationship of increased maternal education with increased obesity-sleep linkage (OR = 4.45) is unintuitive but not without explanation. Children are more likely to have personal electronic devices, use social media and experience academic pressures in higher socioeconomic households, which contributes to later sleep times and shorter sleep. The protective intermediate education stratum (OR = 0.15) could be due to structured household routines that involve an awareness of and the effective implementation of bedtime. Some studies done in Lahore suggest that 11.9% of the school children were obese while the Hazara Division showed 4.78% and Swat showed 6.7% in boys and 10% in girls respectively. The higher burden of obesity in our urban sample is reflective of the higher sedentary burden associated with urban lifestyle. It is also noted that a study was conducted among the medical students of Islamabad which revealed 34% obesity level among female medical students, thus revealing that the scope of the challenge is not limited to age group. The major drawback of this study is that it was an observational case-control study, which does not allow for causation to be inferred. Actigraphic or polysomnographic measures were not used to objectively assess sleep duration. Further, there was no full socioeconomic profile performed and dietary intake and screen time and physical activity level were not well quantified. Objective measurements are indicated for longitudinal studies.

This study establishes a significant association between poor sleep duration and childhood obesity among secondary school children in Karachi, with obese children being more than twice as likely to sleep fewer than eight hours per night. The effect is especially pronounced among girls and younger adolescents, and persists independently of family history of obesity. These findings have immediate and actionable public health implications.

Primary care physicians and school health officers should integrate sleep hygiene counselling into routine adolescent health assessments. School-based programmes targeting physical activity, dietary behaviour, and structured sleep schedules represent a high-impact, low-cost preventive strategy. Parents are critical stakeholders: consistent bedtime routines, limiting evening screen exposure, and modelling healthy sleep behaviours at home can collectively reduce the incidence of childhood obesity and its long-term cardiometabolic sequelae.

This study is among the first to quantify the sleep-obesity association with stratified odds ratios in a Karachi school-going population and provides an evidence base for policy makers to develop context-specific adolescent sleep health guidelines within Pakistan’s national school health framework.

ACKNOWLEDGEMENTS

The authors thank all co-investigators for their support in data collection, statistical analysis, manuscript drafting, and final approval for publication. Gratitude is also extended to the school administrations and students who participated. This study was conducted as part of the FCPS Dissertation of Dr. Aqib Ali.

CONFLICT OF INTEREST

The authors declare no conflict of interest

1. Narzisi K, Simons J. Interventions that prevent or reduce obesity in children from birth to five years of age: A systematic review. Journal of Child Health Care. 2020 Apr 15:1367493520917863.
2. Carrillo-Larco RM, Bernabe-Ortiz A, Miranda JJ. Short sleep duration and childhood obesity: cross-sectional analysis in Peru and patterns in four developing countries. PloS One. 2014;9(11):e112433.
3. Zhang B, Hao Y, Zhou J, Jia F, Li X, Tang Y, et al. The association between sleep patterns and overweight/obesity in Chinese children: a cross-sectional study. Neuropsychiatr Dis Treat. 2015;11:2209–16.
4. Kalra S, Unnikrishnan AG. Obesity in India: The weight of the nation. J Med Nutr Nutraceut. 2012;1:37–41.
5. Fatima K, Rizvi F, Ali M, Afzal M. Sleep pattern and sleep duration of medical college students. Ann Pak Inst Med Sci. 2011;7(2):79–81.
6. Shams N, Niaz F, Motwani R, Shaikh Z, Saleem F. Obesity and hypertension in female medical students: frequency and risk factors. J Liaquat Uni Med Health Sci. 2015;14(01):26–32.
7. Balhareth A, Meertens R, Kremers S, Sleddens E. Overweight and obesity among adults in the Gulf States: A systematic literature review. Obesity Reviews. 2019;20(5):763–93.
8. Shah S, Ihsanullah M, Rasheed A. Association of lipoprotein-a with obesity in children and adolescents in district Swat. J Ayub Med Coll Abbottabad. 2009;21(3):145–8.
9. Haq I, Siddiqui T, Jan M. Prevalence of obesity in school children of Hazara division. J Ayub Med Coll Abbottabad. 2010;22(4):50–2.
10. Anwar A, Anwar F, Joiya H, Ijaz A, Rashid H, Javaid A, et al. Prevalence of obesity among school-going children of Lahore and associated factors. J Ayub Med Coll Abbottabad. 2010;22(4):27–32.
11. Güngör NK. Overweight and obesity in children and adolescents. J Clin Res Pediatr Endocrinol. 2014;6:129–143.
12. Huang L, Lyu J, Long Z, Xia Y, Chen Y, Ye X, Li S. Gender-specific longitudinal association of sleep duration with blood pressure among children: Evidence from CHNS 2004–2015. International Journal of Hypertension. 2020;2020.
13. Al-Hazzaa HM, Al-Rasheedi AA, Alsulaimani RA, Jabri L. Anthropometric, familial- and lifestyle-related characteristics of school children skipping breakfast in Jeddah, Saudi Arabia. Nutrients. 2020;12(12):