Poor sleep quality and its association patterns with eating disorders, body composition and physical activity among university students in Douala-Cameroon Authors: Jerson Mekoulou Ndongo, William Mbang Mbian, Ahmadou, Tresor Rabellais Mbitcha Fongang, Emmanuel Guilou Njimongna Njoya, Elysée Claude Bika Lele, Brinda Yabet Bekima, Abraham Sylvin Mefiré Rengou, William Richard Guessogo, Loick Pradel Kojom Foko, Péguy Brice Assomo Ndemba, Abdou Temfemo, Samuel Honoré Mandengue, Clarisse Noel Ayina Ayina Published: 2026 ABSTRACT AbstractIntroductionPoor sleep (PS) impacts university students’ life including risk of eating disorders (EDs), changes in anthropometric parameters and body composition. However, physical activity (PA) protects against PS and its consequences. This study aimed to determine the prevalence of PS and its association patterns with EDs, body composition, and PA.MethodologyA cross-sectional study was conducted among students at the University of Douala, Cameroon. PS, the risk of EDs, and PA were assessed using the PSQI, the SCOFF, and the IPAQ questionnaires, respectively. Body mass index (BMI), waist circumference (WC), and body composition were measured.ResultsA total of 522 students (female dominance, 61.7%) was included. The prevalence of PS was 73.7% and 35.6% for risk of EDs with females’ dominance (40.0%; P = 0.007). In univariate analysis PS in overall population was associated with high level of PA (aOR:1.77, P = 0.01), risk of EDs (aOR:1.55, P = 0.04), BMI (aOR: 1.23, P = 0.001), fat mass (aOR: 1.08, P = 0.001) and body water (aOR: 0.92, P = 0.001). In male, PS was associated with moderate (aOR: 1.91, P = 0.03) and high PA (aOR: 2.94, P = 0.001), risk of EDs (aOR: 1.69, P = 0.047). In female, PS was associated with BMI (aOR: 1.40, P = 0.001), WC (aOR: 1.09, P = 0.03), fat mass (aOR: 1.25, P = 0.001), and body water (aOR: 0.32, P = 0.0006). In multivariate analysis, PS in males was linked with high PA (aOR: 0.95, p = 0.03), while moderate (aOR: 5.98, p = 0.02) and high level (aOR: 5.19, p = 0.02) in females.ConclusionPS is highly prevalent among university students and is significantly associated with the risk of EDs, anthropometric parameters, body composition, and PA.Keywords: Poor sleep, Eating disorders, Physical activity, University students, Cameroon Introduction Sleep is a crucial moment that affects the quality of life and well-being of individuals. Sleep is a crucial period that affects the quality of life and well-being of individuals. It helps to maintain physiological health and plays an essential role in restoring bodily functions (Buysse, 2014). Good sleep is an important process for physical, psychological, and emotional well-being in adolescents and young people. It plays a key role in emotion, attention, cognition, and school performances (Diekelmann et al., 2009). It is related to energy repair, protection, and recovery, and is an essential factor for maintaining homeostasis and the well-being of overall health (Chattu et al., 2018, Rasch and Born, 2013). However, students' lives are influenced by the use of electronic devices (television, computers, games), technology, and the internet, regardless of whether these technologies are used for study, entertainment, or communication (Mireku et al., 2019). The excessive use of these technologies and the internet, leading to addiction at night, is one of the main causes of poor sleep (PS) (Khan et al., 2022, Leow et al., 2023, Owens and Weiss, 2017) associated to stress of university studies associated with the stress of university studies (Nakie et al., 2024). Previous studies have highlighted the critical effects of poor sleep on numerous physiological and cognitive processes, including memory, learning, behavior, mood, and academic performance (Almalki et al., 2025, Owens and Weiss, 2017, Simpong et al., 2025). PS is a public health issue affecting about half of the global population (Chattu et al., 2018). PS is more prevalent in young adults, especially university students, than in the general population (Jiang et al., 2015, Samaranayake et al., 2014) even among African students (Nakie et al., 2024). It has also been revealed that PS is strongly related to eating disorders (EDs) (Mutti et al., 2023) and related to a variety of non-communicable diseases with increased mortality risk (Chattu et al., 2018, Kanki et al., 2023, Lee and Cho, 2022). EDs are considered as a psychiatric conditions characterized by abnormal eating and weight control-related behaviors, very often accompanied by concerns about body weight and shape. According to the last version of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5), the major feeding and EDs are anorexia nervosa, bulimia nervosa, binge ED, avoidant/restrictive food intake disorder, pica, and rumination disorder (Erzegovesi and Bellodi, 2016). The associated complications with EDs can be serious, leading to certain cardiometabolic diseases. The complications associated with EDs can be serious, including certain cardiometabolic diseases (Özyürek Arpa et al., 2025), highlighting the importance of early identification to prevent long-term health consequences (Voderholzer et al., 2020). EDs among university students have become a major public health challenge (Galmiche et al., 2019, Tavolacci et al., 2015). Some authors have shown that PS is related to EDs among university students. Some authors have shown that PS is related to EDs among university students (Ramón-Arbués et al., 2022) and is also known to exert significant effects on body mass index, weight, and body composition, predisposing individuals to health problems (Khushaim et al., 2025). Eating and sleeping are two powerful biological and psychological needs. These behaviors are also essential to an individual’s physical and psychological functioning (Stickgold and Walker, 2007), and when disturbed, are particularly influential in the development and course of adolescent psychopathology (Lunsford-Avery et al., 2024). PS and EDs are major health concerns for students, two problems that can affect mental and physical health. Therefore, regular practice of physical activity (PA) is known as a non-pharmacological intervention to fight against PS (Brand et al., 2017) and collaterally to EDs (Galasso et al., 2020). In Africa, no study has simultaneously investigated sleep quality and its associations with EDs, body composition, and PA. Therefore, the present study aims to determine the prevalence of PS and its associations with EDs, body composition, and PA among university students. Methods Study design A cross-sectional study was conducted between October 2024 and March 2025 among students of the University of Douala in Cameroon. Ethical considerations and participants The study was reviewed and approved by the Institutional Ethics Committee of the University of Douala under the reference number: CEI-UD/3602/04/2023/T and was conducted in conformity with the recommendations of the Declaration of Helsinki revised in 2013. Students provided informed consent before participating. They were informed about the anonymous analysis of information and assured that personal information would not be disclosed. Eligibility criteria were full-time students’ willingness to participate voluntarily after being informed about the aim of the research and providing written informed consent. Individuals who met the following criteria were excluded from the research: pregnant or lactating students, those diagnosed with a psychological disorder or a sleep disorder. The participants were asked whether they met the above criteria for making this distinction, and those who responded “no” were included in the research. Among the 600 participants, 78 students (28 men, 50 women) returned uncompleted questionnaires and refused to continue participating in the study, and 522 students who fully completed the survey were definitively included in the study. Sample size The minimum sample size was calculated using the following formula: [Z²p(1-p)]/C², with a 5% margin of error and a 95% confidence level, based on the prevalence of PS of 63.31% reported by Nakie et al. (2024) among university students in Africa. An estimated minimum sample size of 357 was obtained. Variable measurements Sociodemographic information’s anthropometric measurements Participants responded to the questions concerning their sociodemographic characteristics (such as gender, age, and academic level). Anthropometric parameters and body composition Height was measured using a non-stretchy tape measure, without shoes and with the head in the Frankfort plane position. Waist circumference (WC) was measured from the middle point of the region between the lower rib bone and the crystal iliac bone. Body weight and body composition were measured using a portable scale (Tanita BC−525, Tanita Corporation, Tokyo, Japan) with lightweight clothing and no shoes. BMI and the waist to height ration (WhtR) were calculated and classified according to the World Health Organization recommendations. Quality of sleep The Pittsburgh Sleep Quality Index (PSQI) was used to assess sleep quality (Buysse et al., 1989). This questionnaire assesses a participant's sleep quality over one month. PSQI consists of 19 items on seven components of sleep quality: sleep onset latency, sleep duration, efficiency, quality, disturbances, medication, and daytime dysfunction. Each component of PSQI is scored between 0–3, and the sum of the seven components yields a global score of sleep quality with a total score ranging from 0 to 21 points; a high score (PSQI score > 5) is an indication of PS. The PSQI internal consistency was assessed using Cronbach’s alpha coefficient, which was 0.88, indicating good reliability. Eating disorders The French version of the Sick, Control, One stone, Fat, Food (SCOFF) questionnaire was used to assess the risk of EDs. This questionnaire is a validated tool for screening EDs in many populations (Morgan et al., 2000). This questionnaire is considered a simple and effective screening tool for EDs, particularly used in student populations (Garcia et al., 2010). The SCOFF questionnaire has been validated in several populations and in different languages (Hill et al., 2010). It consists of 5 dichotomous questions (Yes/No); answering positively to 2 or more questions indicates a probability of having an ED with an internal consistency of the Cronbach’s alpha coefficient, which was 0.95. Level of physical activity The short version of the International Physical Activity Questionnaire (IPAQ) was administered to assess the level of PA. This tool has good validity and reliability in quantifying PA This tool demonstrates good validity and reliability for quantifying PA (Craig et al., 2003). This questionnaire assesses PA in three domains (PA at school, travel to and from places and recreational activity). The first six items require participants to report the frequency and duration of low, moderate, and vigorous PA performed during the previous week. The seventh item quantifies the total sedentary time during the weekdays. For analysis, the metabolic equivalent task (MET) weeks/minutes was calculated from the raw data obtained through six items according to the guidelines of the IPAQ scoring procedure the MET values were calculated from the raw data obtained from the six items according to the IPAQ scoring procedure guidelines (Ipaq Research Committee, 2005). Based on total MET values, participants were further categorized as low, moderate, and highly active according to the guidelines of this tool. The internal consistency of the IPAQ was assessed using Cronbach’s alpha coefficient, which was 0.91. Statistical analysis Data was first entered into Microsoft Excel and analyzed using Statistical Package for Social Sciences (SPSS version 20 software). Data were first entered into Microsoft Excel and analyzed using the Statistical Package for Social Sciences (SPSS, version 20). The Kolmogorov-Smirnov test verified the normality of the data. In the descriptive stage of the analysis, mean and SD of quantitative variables were calculated, as well as the absolute and relative frequencies of categorical variables and lower and upper limits of the 95% confidence interval (95%CI) for a proportion confidence interval of PS and EDs were determined (Newcombe, 1998). The Pearson’s chi-square test was used to compare unpaired variables, and the Student’s t-test on unpaired series was performed to compare quantitative variables according to the gender and quality of sleep. Pearson’s chi-square (χ2) test was conducted to verify the difference between proportions. Logistic regression models were performed to identify factors associated with poor sleep, EDs, anthropometric parameters, body composition and level of PA. Logistic regression models were applied to identify factors associated with poor sleep, EDs, anthropometric parameters, body composition, and level of PA. The association was quantified by computing odds ratios adjusted for age, with 95% confidence intervals (95%CI). The statistical significance threshold was set for any value of P < 0.05. Results A total of 522 students were enrolled in the study, with the majority being female (61.7%). The majority had normal ponderal status (80.6%), and half (44.8%) were from level 2 in the university academic. The majority had a normal weight status (80.6%), and the largest proportion (44.8%) were enrolled in the second year of university. The proportion of participants with a high level of PA was significantly higher in males than in females (57.8% vs 40.3%, p = 0.0004). A statistically significant association (p = 0.047) was found between sleep quality and weight status. Specifically, PS was more frequently observed than good sleep quality in participants with overweight (16.3%) and obesity (1.3%) (Table 1). Table 1Characteristics of participants.ParametersCriteria’sTotal(n= 522)Males(n= 197)Females(n= 325)p-valuePoor sleep(PQSI>5)Good sleep(PSQI<5)p-valuePonderalstatusUnderweight11 (5.5)5 (4.9)6 (6.1)0.3349 (5.9)2 (4.2)0.047*Normal162 (80.6)87 (85.3)75 (75.7)117 (76.5)45 (93.8)Obese2 (0.9)1 (1.0)1 (1.0)2 (1.3)0 (0)Overweight26 (12.9)9 (8.8)17 (17.2)25 (16.3)1 (2.1)Educational level1136 (26.1)52 (26.4)84 (25.8)0.05630 (21.9)106 (27.5)0.2572234 (44.8)89 (45.2)145 (44.6)60 (43.8)174 (45.2)3145 (27.8)50 (25.4)95 (29.2)46 (33.6)99 (25.7)47 (1.3)6 (3)1 (0.3)1 (0.7)6 (1.6)Level of PhysicalactivityLow128 (24.5)35 (17.8)93 (28.6)0.0004*84(21.8)44(32.1)0.050Moderate149 (28.5)48 (24.4)101 (31.1)112(29.1)37(27.0)High245 (47)114 (57.8)131 (40.3)189(49.1)56(40.9)Pearson’s independence chi square test*Statistically significant at p-value < 0.05 Characteristics of participants. Pearson’s independence chi square test *Statistically significant at p-value < 0.05 As resumed in Table 2 anthropometric measurements, body composition, and sleep components categorized by sex show that males were older than females (p < 0.0001), with high values of weight, height, body water percent and body mass (p < 0.01). Furthermore, females presented higher mean values of waist to height ratio (p = 0.006), fat mass percent (p < 0.0001), and values of sleep quality PSQI component (p = 0.0001) compared to males. Regarding sleep quality, students with PS had significantly higher BMI than those with good sleep quality (p = 0.001). Similarly, students with PS had higher fat mass than their counterparts with good sleep quality (p = 0.001). Table 2Anthropometric, body composition, measurements and sleep components categorized by gender.ParametersCriteria’sTotal(n = 522)Girls(n = 325)Boys(n = 197)p-valuePoor sleep(PQSI>5)Good sleep(PSQI<5)p-valueMoy ± DSMoy±SDMoy±SDMoy±SDMoy±SDAge (Years)Age (Years)21 ± 220 ± 221 ± 2<0.0001*21 ± 221 ± 20.35Anthro-pometricWeight (Kg)63.2 ± 9.460 ± 9.266.2 ± 8.7<0.0001*64.1 ± 10.060.3 ± 6.50.016*Height (m)1.69 ± 0.091.64 ± 0.071.74 ± 0.07<0.0001BMI (Kg/m2)22.2 ± 3.0322.4 ± 3.121.9 ± 2.90.2022.5 ± 3.220.9 ± 2.10.001*WC (m)75.0 ± 7.074.0 ± 7.576.0 ± 6.00.0975.0 ± 7.073.0 ± 6.00.11Waist/Height0.44 ± 0.050.45 ± 0.050.44 ± 0.040.006*0.45 ± 0.050.43 ± 0.040.06Body compositionFat mass (%)17.8 ± 7.323.1 ± 5.112.6 ± 5.0<0.0001*18.7 ± 7.214.9 ± 6.80.001*H2O (%)58.5 ± 6.553.4 ± 4.163.5 ± 4.1<0.0001*57.7 ± 6.461.1 ± 6.30.001*Muscle mass (%)43 ± 7.236.5 ± 3.349.2 ± 3.5<0.0001*45.2 ± 3.145.3 ± 7.10.976PSQISubjectivity1.34 ± 0.741.31 ± 0.741.38 ± 0.740.33Latency1.12 ± 1.71.15 ± 0.971.07 ± 0.900.32Duration1.35 ± 11.29 ± 0.931.44 ± 1.00.06Efficiency0.62 ± 0.90.62 ± 0.950.62 ± 0.960.98Disturbances1.32 ± 0.591.40 ± 0.601.20 ± 0.550.0001*Medication0.21 ± 0.650.21 ± 0.660.20 ± 0.60.85Daytimedysfunction1.56 ± 1.021.55 ± 1.031.58 ± 0.990.78Score PQSI7.51 ± 3.107.53 ± 3.207.48 ± 2.890.87PSQI: Pittsburgh sleep quality index, BMI: Body mass index, WC: waist circumference, Data are summarized as mean ± standard deviation (SD), Unpaired samples t Student test, *Statistically significant at p-value < 0.05 Anthropometric, body composition, measurements and sleep components categorized by gender. PSQI: Pittsburgh sleep quality index, BMI: Body mass index, WC: waist circumference, Data are summarized as mean ± standard deviation (SD), Unpaired samples t Student test, *Statistically significant at p-value < 0.05 As illustrated in Fig. 1, the prevalence of PS was 73.7%(95%CI:69.8–77.3), and 35.6% (95%CI:31.6–39.8) for the risk of EDs with high prevalence in females (40.0% 95%CI:34.8–45.4; P = 0.007). Among students with poor quality of sleep, the prevalence of EDs was 38.2%(95%CI:33.5–43.2), with a high rate in females (43.3%, 95%CI:34.8–45.4; P = 0.008). Table 3 presents EDs categorized by sleep quality and sex. The prevalence of PS quality was significantly higher in participants with risk of EDs compared to those not at risk (61.8% vs 38.2%, p = 0.041) regardless of gender. Among girls, the risk of EDs was higher in those with PS (31.0% vs 43.3 %, p = 0.046). By stratifying the analysis with regard to gender, no significant association was found in males (p = 0.421), but an inverted pattern was found only in females (61.8%, p = 0.041). The mean values of five PSQI components (i.e., global PSQI score, subjective sleep quality, sleep efficiency, sleep disturbances, and daytime dysfunction) were significantly higher in participants at risk of EDs. In supplementary Table S1, sleep disturbance was more common in students with the risk of EDs (41.9%, P < 0.0001) in the overall study population. The same link was reported in females (47.7%, P = 0.0002) and males (30.4%, P = 0.01). Furthermore, in the overall population, participants with risk of EDs, daytime dysfunction ranging 3–4 was more prevalent in them (40.3%, P = 0.01). Even in females (43.8%, P = 0.005). Prevalence of poor quality of sleep and eating disorders according to gender. Table 3Eating disorders categorized by sleep quality and gender.ParametersGenderCategoriesNGood SleepPoor sleepp-valueOverallRisk of ED27739 (28.5)238 (61.8)0.041a*No Risk of ED24598 (71.5)147 (38.2)Eating disorders (ED)MalesRisk of ED5612 (23.5)44 (29.9)0.421aNo Risk of ED14239 (76.5)103 (70.1)FemalesRisk of ED13027 (31.0)103 (43.3)0.046a*No Risk of ED19560 (69.0)135 (56.7)Eating disordersRisk(Moy ± SD)No Risk(Moy ± DS)p-valuePSQIPQSI score8.24 ± 3.197.11 ± 3.0<0.0001b*Sleep subjectivity1.43 ± 0.761.28 ± 0.730.030b*Sleep onset latency1.22 ± 0.711.07 ± 0.930.081bSleep duration1.32 ± 0.921.36 ± 0.980.644bSleep efficiency0.75 ± 1.010.55 ± 0.880.021b*Disturbances1.52 ± 0.601.22 ± 0.56<0.0001b*Medication0.23 ± 0.660.19 ± 0.640.585bDaytime dysfunction1.79 ± 0.961.44 ± 1.030.0002b*Data are summarized as frequency (percentage) and mean ± standard deviationaPearson’s independence chi square testbIndependent samples Student's t-test*Statistically significant at p-value < 0.05 Eating disorders categorized by sleep quality and gender. Data are summarized as frequency (percentage) and mean ± standard deviation aPearson’s independence chi square test bIndependent samples Student's t-test *Statistically significant at p-value < 0.05 Based on univariate logistic regression analysis in Table 4, in the overall study population, the odds of PS were higher in participants with a high level of PA (aOR: 1.77, 95% CI: 1.10–2.83, p = 0.01) and in those at risk of EDs (aOR: 1.55, 95% CI: 1.01–2.37, p = 0.04). A one-unit increase in BMI and fat mass was associated with a 23% (aOR: 1.23, 95% CI: 1.08–1.40, p = 0.001) and 8% (aOR: 1.08, 95% CI: 1.03–1.14, p = 0.001) higher risk of PS (Table 5). In males, the risk of PS was higher in participants who were at risk of EDs (aOR: 1.68, 95% CI: 1.0–2.86, p = 0.047), and in those with moderate (aOR: 1.91, 95% CI: 1.04–3.51, p = 0.03) and high (aOR: 2.94, 95% CI: 1.61–5.41, p = 0.001) levels of PA, compared to those with a low PA level. In females, the likelihood of PS was increased with increasing values of BMI (aOR: 1.40, 95% CI: 1.14–1.72, p = 0.001) and fat mass (aOR: 1.09, 95% CI: 1.01–1.18, p = 0.03). However, multivariate analysis revealed that, in males', a high level of PA was protective against PS (aOR: 0.95, 95% CI: 0.01–0.81, p = 0.03), while in females', moderate (aOR: 5.98, 95% CI: 1.19–30.01, p = 0.02) and high (aOR: 5.19, 95% CI: 1.25–21.46, p = 0.02) levels of PA constituted risk factors of PS. Table 4Associated factors of poor sleep according to gender.Univariate analysisMultivariate analysisParametersCriteria’sOverallMalesFemalesOverallMalesFemalesaOR(95%CI)p-valueaOR(95%CI)p-valueaOR(95%CI)p-valueaOR(95%CI)p-valueaOR(95%CI)p-valueaOR(95%CI)p-valueGenderFemales11Males1.08(0.72–1.61)0.72--2.87(0.45–18.57)0.266--Level ofPhysicalActivityLow11111Moderate1.59(0.94–2.75)0.081.91(1.04–3.51)0.03*0.84(0.29–2.44)0.751.63(0.61–3.23)0.340.13(0.01–1.21)0.075.98(1.19–30.01)0.02*High1.77(1.10–2.83)0.01*2.94(1.61–5.41)0.0005*0.64(0.25–1.61)0.341.34(0.56–3.21)0.530.95(0.01–0.81)0.03*5.19(1.25–21.46)0.02*Eating disordersNo risk111111Risk1.55(1.01–2.37)0.04*1.69(1.01–2.86)0.047*1.35(0.65–2.83)0.421.46(0.66–3.23)0.341.10(0.34–3.35)0.872.11(0.56–7.89)0.26Anthropo-metricBMI(Kg.m−2)1.23(1.08–1.40)0.001*1.09(0.93–1.30)0.2881.40(1.14–1.72)0.001*1.13(0.85–1.49)0.400.93(0.57–1.50)0.760.96(0.62–1.48)0.86WC(Cm)1.04(0.99–1.09)0.111.01(0.94–1.08)0.8491.09(101–1.18)0.03*1.03(0.91–1.16)0.661.17(0.97–1.43)0.101.02(0.86–1.22)0.81WhtR1.32(0.55–3214)0.063.27(0.95–1.34)0.822387177(1.57–95277)0.048.33e06(9.87e15−7021.03)0.263.27e−9(2.27e−23−470)0.241.18e−6(7.17e−2–1948)0.37Boby compositionFat mass (%)1.08(103–1.14)0.001*1.06(0.98–1.20)0.1351.25(1.01–1.43)0.001*1.13(0.85–1.49)0.401.33(0.75–2.33)0.321.02(0.75–1.39)0.81H2O(%)0.92(0.87–0.97)0.001*0.91(0.81–1.03)0.1420.32(0.61–0.87)0.0006*0.88(0.73–1.10)0.290.80(0.50–1.30)0.360.90(0.67–1.18-)0.89Muscle (%)1.00(0.99–1.01)0.970.88(0.76–1.02)0.091.01(0.98–1.02)0.691.01(0.98–1.03)0.611.25(0.63–2.48)0.520.88(0.72–1.08)0.21WhtR: Waist/ Heigh ratio, WC: waist circumference, PA: physical activities, BMI: Body mass index, aOR: adjusted Odds ratio on age, CI; confidence interval, *Statistically significant at p-value < 0.05. Associated factors of poor sleep according to gender. WhtR: Waist/ Heigh ratio, WC: waist circumference, PA: physical activities, BMI: Body mass index, aOR: adjusted Odds ratio on age, CI; confidence interval, *Statistically significant at p-value < 0.05. Discussion Sleep quality, eating behavior and practice of PA are essential for academic performance, emotional regulation, and prevention of cardiometabolic health among students for better health in adulthood. Sleep quality, eating behavior, and physical activity are essential for academic performance, emotional regulation, and the prevention of cardiometabolic diseases among students, contributing to better health in adulthood. This study firstly aimed at determining the prevalence of PS among university students in Cameroon. The prevalence of PS noticed in the present study was 73.7%(95%CI:69.8–77.3). This prevalence was significantly higher than that of different studies conducted among university students at a different period. This prevalence was notably higher than that reported in several studies conducted among university students across different periods and regions. In a systematic review and meta-analysis of Leow et al.(2023), a pooled PS prevalence of 57% was reported among university students. In the same vein, the current prevalence is also higher than the 59.2% reported in a systematic review and meta-analysis conducted among Korean university students (Hwang and Shin, 2020). In the African context, the current prevalence noticed of this study is elevated compared to 63.31%; in a recent systematic review and meta-analysis study on sleep quality and associated factors among university students in Africa (Nakie et al., 2024). Also, in a study conducted in Ghana among student aged 20–29 years, 48.5% of participants exhibited PS (Simpong et al., 2025). Besides, beyond the university population, it should be noted that the finding prevalence of the current study is also superior to that of a global systematic and meta-analysis study on the general population with a prevalence of 57.3% (Limongi et al., 2023). This high prevalence of PS could initially be explained by racial or ethnic factors, which adversely affect the quality of an individual's sleep. PS is more prevalent among young African adults compared to those from other continents. Previous studies have established that Black individuals are more likely to suffer from sleep disorders, thereby increasing the likelihood of PS in Africa (Johnson et al., 2019, Kingsbury et al., 2013, Owens, 2014). Also, young students often spend a large part of their evenings on social media for escapism. While others use electronic devices such as computers, tablets, and smartphones to read educational materials (Das-Friebel et al., 2020). As a result, most young adults may suffer from sleep deprivation and excessive daytime sleepiness while their bodies attempt to synchronize their naturally shifted circadian rhythm with their daily social schedules and academic activities (Uccella et al., 2023). Finally, this finding may also be attributed to academic stress and increasingly rigorous academic demands (Nakie et al., 2024). The results of the present study demonstrated a higher prevalence of EDs risk among students with poor sleep. This finding is consistent with the literature on the neuroendocrine mechanisms of appetite regulation. From a physiological standpoint, appetite-regulating hormones, notably leptin and ghrelin, which also regulate orexins, play an essential role in maintaining physiological homeostasis between the sleep-wake cycle and eating behavior (Suna and Ayaz, 2022). Sleep restriction has been associated with reduced leptin levels and elevated ghrelin, accompanied by increased hunger and appetite (Broussard et al., 2016). Indeed, elevated orexin levels in a state of hunger promote greater alertness and increased food intake, representing a potential mechanism linking PS to a heightened risk of EDs (Suna and Ayaz, 2022). Students with PS display significantly higher BMI and body fat percentage than their peers with good sleep quality. These observations are consistent with those of a growing body of scientific evidence establishing a bidirectional link between sleep quality and body composition. A meta-analysis highlighted a significant positive association between PS and higher BMI (Ahn et al., 2026). Short sleep duration contributes to weight gain, while a high BMI in turn impairs sleep quality and duration (Ahn et al., 2026). The deterioration of sleep leads to profound hormonal dysregulation, primarily involving leptin and ghrelin. This dysregulation of these two hormones plays a key role in the increase in BMI among individuals with PS (Mosavat et al., 2021). High body fat mass may itself be involved in the secretion of inflammatory cytokines that disrupt sleep-wake rhythms (Kawasaki et al., 2023). This further reinforces the vicious cycle between sleep disturbance and excessive adiposity. Evidence indicates that PS is associated with lower hydration levels upon waking. The body's ability to regulate hydration is closely linked to the circadian rhythm; the hormone vasopressin, which helps the kidneys manage fluid balance, is released in greater quantities during the final hours of the sleep cycle, and any sleep disturbance may therefore interfere with this process (Rosinger et al., 2019). Our findings highlight that in the overall study population, a one-unit increase in BMI and fat mass was associated with a 23% and 8% higher risk of PS quality, respectively. These findings are consistent with those of other authors who reported that individuals with sleep disturbances had significantly higher BMI and greater body fat percentage (Xu et al., 2025). Similarly, a dose-response relationship has been demonstrated, indicating that higher BMI levels are associated with increased susceptibility to sleep disturbances (Wang et al., 2024). Excess fat mass, particularly visceral fat, disrupts sleep quality through several converging physiological pathways. Visceral adipose tissue secretes pro-inflammatory cytokines (IL−1, IL−6, and TNF-α), which contribute to chronic inflammation and impair sleep regulation. TNF-α and IL−1β reach their nocturnal peak and play an important role in deep slow-wave sleep; in overweight or obese individuals, these cytokines are elevated in the morning and have been associated with sleep disorders (Liu et al., 2025). These inflammatory mechanisms, by altering sleep architecture and promoting nocturnal micro-arousals, largely account for the dose-response gradient observed between BMI, fat mass, and the risk of PS quality in the present study (Fu et al., 2025, Wang et al., 2022). The positive association between a high level of PA and PS in univariate analysis represents an apparently paradoxical finding that warrants careful interpretation. It should first be noted that PA has a positive effect on sleep, but that factors such as age, sex, level and specific characteristics of PA play crucial roles (Korkutata et al., 2025). PA stimulates circadian rhythms, particularly those governing cortisol, melatonin, and core body temperature, thereby influencing sleep quality. In particular, intense PA performed in the evening can lead to sympathetic hyperactivation and elevated cortisol levels, delaying sleep onset and fragmenting sleep (Kim et al., 2023). Based on the multivariate analysis, a high level of PA emerges as a protective factor against PS in males. This finding is consistent with established evidence on the beneficial effects of regular PA on male sleep. Regular PA confers a range of health benefits, including improved sleep and reduced sleep disorder symptoms (Korkutata et al., 2025). Sustained PA in males promotes better regulation of the circadian cortisol profile, increased nocturnal secretion of testosterone and growth hormone, and a greater proportion of deep slow-wave sleep, the most restorative sleep phase. Hypothalamic pituitary adrenal (HPA) axis responses to PA do not appear to differ between men and women. However, regular PA in males is associated with improved cortisol regulation and normalization of circadian rhythms (Nys et al., 2022). In multivariate analysis, moderate and high PA levels emerged as risk factors for PS quality in women. This result appears to contradict the generally beneficial effects of PA on sleep, but can be interpreted in light of sex-specific physiological mechanisms. The HPA axis response to exercise is modulated by female sex hormones. Women appear to respond differently to chronic stress than men, displaying greater increases in cortisol levels under conditions of chronic stress. Furthermore, the HPA axis is influenced by estrogens, which may alter stress reactivity via the regulation of cortisol receptors (Nys et al., 2022). In female students combining moderate-to-intense PA with high academic and psychosocial demands, chronic activation of the HPA axis may produce evening hypercortisolemia, which is detrimental to sleep onset and sleep architectural quality. In addition, intense PA in women can induce a state of reduced energy availability, with consequent suppression of reproductive hormones and disruption of the menstrual cycle, directly impairing sleep quality. The intensity and duration of PA stimulate the hypothalamic-pituitary-adrenal axis, leading to elevated cortisol levels, while an inadequate diet and low energy availability further suppress reproductive hormones (Mennitti et al., 2024). In univariate analysis, the risk of PS among women increased with BMI and fat mass. This finding, consistent with the existing literature, can be explained by the following mechanisms: excess fat mass in women predominantly subcutaneous rather than visceral in distribution can, nevertheless, beyond a certain threshold, generate low-grade inflammation and hormonal disturbances involving leptin and adiponectin, which impair sleep quality. A sex-stratified analysis showed that sleep disturbances were positively associated with BMI, particularly in women (Xu et al., 2025). Conclusion Our findings revealed that PS is highly prevalent among university students, affecting both sexes. PS quality was associated with the risk of EDs, with modifications in BMI and body composition. The protective effect of PA practice was observed in men, while it represented a risk factor in women. This finding has important public health implications, as it underscores the need to improve sleep quality among students and to address its consequences for cognition and the prevention of EDs and related diseases among university students in an African context. Ethical approval The present study was conducted in accordance with the ethical standards of the Institutional Ethics Committee of the University of Douala, and with the 1964 Helsinki Declaration and its later amendments in 2013. Ethical approval was obtained from the Institutional Ethics Committee of the University of Douala (Approval No. CEI-UD/3602/04/2023/T). Verbal informed consent was obtained from all participants prior to their inclusion in the study. Funding The authors received no specific funding for this work. Declaration of Competing Interest The authors have no conflicts of interest related to this work. Footnotes Supplementary data associated with this article can be found in the online version at doi:10.1016/j.ibneur.2026.06.011. Supplementary material Supplementary material Data availability All data from this study are available upon request from the corresponding or first author. References Associated Data Supplementary Materials Supplementary material Data Availability Statement All data from this study are available upon request from the corresponding or first author.