Sleep duration moderates association between screen time and emotional and behavioural problems in young children

Mental health is a fundamental component of well-being and a basic human right [1]. Childhood and adolescence are critical life stages characterized by significant brain growth and development [2]. A recent United Nations International Children's Emergency Fund (UNICEF) report highlights an alarming prevalence of mental disorders among children and adolescents aged 10–19 years in Western Europe, with a prevalence of about 16% in boys and 17% in girls, compared with a global prevalence of 13% [2]. Specifically, in Spain, the rate of psychiatric diagnoses in adolescents significantly raised from 3.9% in 2000 to 9.5% in 2021 [3]. While certain mental health disorders are more common in adolescents, behavioral and emotional difficulties, including hyperactivity, conduct issues, and emotional dysregulation, may also emerge in younger children and persist into adulthood [4‐7]. This underscores the importance of early identification and prevention strategies to mitigate potential long-term consequences [8].

Despite this, most research has focused on adolescents, with relatively few studies examining the early childhood period. The evidence suggests that preschoolers and young children are also vulnerable to psychosocial and behavioral challenges linked to lifestyle factors such as screen time and sleep disturbances [8‐10]. Several studies and systematic reviews have reported associations between excessive screen time and adverse developmental outcomes in children, including poorer mental health, sleep problems, cognitive delays, and behavioral difficulties [5, 8‐10]. Nonetheless, the extant literature is controversial, with some studies reporting significant relationships and others finding weak, inconsistent, or null associations, highlighting the need for further high-quality research in this area [11]. Consequently, international health organizations, including the World Health Organization (WHO) [12], the American Academy of Child and Adolescent Psychiatry [13], and the American Academy of Pediatrics [14], recommend limiting screen exposure in children and adolescents as part of broader preventive health strategies.

In addition to the potential direct impact of screen time on mental health, certain reviews have suggested that the COVID-19 pandemic has contributed to increases in screen time and mental health problems in children and adolescents [15, 16]. In addition, both child lifestyle factors (including diet and sleep duration) and parental factors (including socioeconomic status) appear to play important roles in moderating this relationship [7, 17‐19]. Some studies have reported that greater screen exposure is associated with shorter sleep duration, later sleep onset, and reduced sleep quality [6, 18, 20]. Moreover, children who do not meet sleep duration recommendations are particularly susceptible to the negative effects of screen time on mental health [18]. However, the literature remains mixed, and meta-analyses indicate that these associations are generally small in magnitude [11, 21].

The present study aimed to examine the relationship between screen time and behavioral and emotional difficulties in children, as assessed via the strengths and difficulties questionnaire (SDQ), with a particular focus on the role of sleep duration. We hypothesized that greater screen time and noncompliance with screen time recommendations would be associated with greater emotional and behavioral difficulties, with sleep duration potentially playing an important role in these associations. By analyzing these relationships in a large multicenter cohort of 3–6-year-old children in Spain, we aim to contribute to the growing body of evidence supporting strategies that promote healthy screen use and sleep habits in early childhood.

Of the 1509 participants who attended the CORALS baseline visit, 67 and 22 participants were excluded because of missing data on screen time and SDQ, respectively, resulting in a total population of 1420 children being analyzed (Supplementary Fig. 1). The participants had a mean age of about five years and an average screen time of 1.8 hours/day. About 10%–12% of the children scored above the SDQ cutoff thresholds for emotional and behavioral difficulties. Mothers completed questionnaires for 87% of the participants.

As shown in Table 1, children who met the screen time recommendations had significantly lower scores on the emotional and behavioral subscales of the SDQ, except for the prosocial subscale. These children presented longer sleep durations and adequate sleep patterns, greater adherence to the Mediterranean diet (MedDiet), a greater likelihood of being exclusively breastfed for the first six months, and lower BMI z scores and a higher prevalence of overweight/obesity. Both parents of children in the compliance category were younger, had a lower BMI and healthier weight status, and had higher education and socioprofessional status (Supplementary Table 1).

Table 2 shows that, in the fully adjusted models, higher screen time was associated with higher scores on the total SDQ and the emotional symptoms, conduct problems, and peer problems subscales. Associations with hyperactivity-inattention and prosocial behavior were not significant. Children adhering to screen time recommendations had significantly lower total SDQ, conduct and peer problems scores, although the effect sizes were small (Cohen’s d =  − 0.077). Secondary analyses revealed that children under five years of age who adhered to the screen time recommendation of ≤ 1 hour/day (n = 255) had lower total SDQ scores [− 0.86 (95% CI − 1.58, − 0.14), P = 0.019] than those who did not adhere to this recommendation. As illustrated in Fig. 1, greater screen time was associated with greater odds of behavioral and emotional difficulties, as indicated by the SDQ cutoff score of 16. In all these models, the mean variance inflation factor (VIF) values were less than 2.0, indicating that there was no multicollinearity.

Bild vergrößern

A longer sleep duration was associated with a lower total SDQ score [β (95% CI), –0.46 (–0.74, –0.18); P = 0.001]. Compared with participants with an inadequate sleep pattern, those with an adequate pattern had a lower total SDQ score [–0.86 (–1.49, –0.24); P = 0.007]. The moderating effect between screen time and sleep duration on total SDQ was significant (P for interaction = 0.020), and the results were stratified accordingly in Table 3, which shows that the association between higher screen time and higher total SDQ was particularly apparent in children with inadequate sleep duration.

Figure 2 shows that replacing 30 minutes from screen time to sleep was associated with lower scores on the total and subscales of the SDQ, except for the prosocial subscale.

Bild vergrößern

As sensitivity analyses, Supplementary Fig. 2 shows the associations stratified by participants’ age, sex, weight status, adherence to the MedDiet, paternal education level and socioprofessional category, as well as by whether participants were assessed during the presence or absence of COVID-19 sanitation restrictions in Spain. In children with lower MedDiet adherence or whose fathers had higher education levels, positive associations were found between screen time and total SDQ. No additional significant interactions were observed. Supplementary Fig. 3 shows the Pearson correlation matrix between key study variables, including screen time, SDQ subscales, sleep duration, and age.

Our study revealed that greater screen time is associated with greater emotional and behavioral difficulties, whereas adherence to screen time recommendations is related to fewer psychosocial difficulties in children aged 3–6 years, particularly those with inadequate sleep duration. Additionally, our results suggest that substituting screen time with sleep duration may benefit children’s emotional and behavioral well-being.

However, recent meta-analyses examining the relationship between screen time and mental health have reported minimal effects or inconclusive results [11, 21]. These inconclusive findings likely reflect differences in study designs, sample sizes, age groups, cultural contexts, definitions of screen time (e.g., total time vs. type of use), measurement methods (self-reported vs. objective), and whether moderating factors such as sleep, diet, and parental involvement were considered. Additionally, the rapidly evolving nature of digital media, including the emergence of interactive platforms, educational apps, and diverse screen content, complicates generalization and comparisons across studies conducted in different time periods. Nevertheless, our findings align with those of several studies conducted in Europe, Asia, and North America that reported negative associations between screen time and children’s mental health [29‐34].

Some mechanisms may explain the association between excessive use of screens and psychosocial issues. Physiologically, elevated screen time may disrupt neuroendocrine pathways (e.g., melatonin, cortisol, and insulin), leading to emotional dysregulation [35‐37]. One hypothesis suggests that exposure to radiofrequency electromagnetic fields from wireless devices could impact the nervous system and contribute to emotional and behavioral problems [30, 35]. Lifestyle factors such as sedentary behaviors, sleep, diet, and parental supervision have also been suggested to contribute to the relationship between screen use and emotional and behavioral issues in childhood [9, 36, 38].

Sleep has received considerable attention in studies examining the associations between screen time and mental health [20, 21]. Our results showed that not only was longer sleep duration independently associated with fewer emotional and behavioral difficulties, which is consistent with prior literature [6, 18], but also a moderating effect of sleep duration was detected. Specifically, among children with inadequate sleep duration, greater screen time was linked to greater emotional and behavioral problems, suggesting a detrimental synergistic effect of greater screen time combined with reduced sleep on children’s psychosocial well-being. This finding aligns with previous research [18, 36, 39]; however, some studies have reported only small or negligible effects of sleep [21, 40], and others have emphasized the need for further research using longitudinal or repeated measures designs to better disentangle the complex relationship between screen time and sleep in children [21]. The results of our isotemporal substitution model further highlight the importance of sleep duration. This model assumes that replacing time from one activity to another can provide valuable insights into the behavioral trade-offs affecting health outcomes [41]. Replacing 30 min of daily screen time with sleep was associated with improved emotional and behavioral outcomes, highlighting the potential benefits of even small adjustments in daily routines to support psychosocial health. Physiological mechanisms may influence the relationships among screen time, sleep, and psychosocial health. Exposure to light-emitting screens—especially before bedtime—can delay sleep onset, reduce sleep quality, and suppress melatonin secretion [42‐44]. Moreover, insufficient sleep duration has been associated with increased screen time in 17,7091 children [45], collectively suggesting a detrimental feedback loop between higher screen time and reduced sleep duration that disrupts circadian rhythms and may negatively impact mental health [46]. Nevertheless, this relationship has also been challenged in some studies involving children [40].

Our findings further indicate that greater screen time was particularly associated with greater emotional and behavioral issues in children with lower adherence to the Mediterranean diet. This suggests that inadequate adherence to nutrient-dense foods—rich in fruits, vegetables, whole grains, and healthy fats—may reduce the neuroprotective and anti-inflammatory benefits associated with better mental health [38, 47]. Interestingly, the association between screen time and total SDQ was stronger among children whose fathers had higher education levels, contrary to previous studies [38, 48]. This counterintuitive finding suggests that a higher parental educational level by itself may not protect against excessive screen exposure or its potential negative effects. This could reflect contextual challenges faced by parents, such as perceptions of educational screen benefits or difficulties in balancing work and caregiving [48]. Broader socioeconomic conditions, digital literacy, and the availability of time and resources to supervise or engage with children’s media use should also be considered relevant factors influencing these associations. In fact, co-using screens with caregivers and engaging with educational content can foster learning and positive interactions, whereas unsupervised use or exposure to inappropriate content is linked to poorer outcomes [49]. Furthermore, excessive parental screen time may contribute to reduced family engagement, parental stress, and family conflict, which in turn may impact children’s emotional and behavioral well-being [50].

Although the associations observed were statistically significant, the effect sizes reported in our study were small, indicating modest impacts at the individual level. However, in population-based research, even small effects can be meaningful, especially when exposures—such as screen time—are highly prevalent and accumulate over time. Nonetheless, we acknowledge the importance of avoiding overinterpretation of statistically significant yet potentially trivial effects [51].

This study has notable strengths, including a large sample size spanning seven Spanish cities, a standardized recruitment methodology, adjustments for multiple confounders, and several potential interactions. In addition, our results on psychosocial outcomes remained similar when the AAP or WHO compliance with screen time cutoffs was used, suggesting robustness.

However, several limitations should be considered. First, the cross-sectional design of this study does not allow the establishment of causal relationships. Second, the study was conducted in Spanish children, which could limit its generalizability. Third, several variables—including sleep duration, screen time, and behavioral outcomes—were based on parent reports, which are subject to recall and social desirability bias [52], potentially affecting the accuracy and validity of the reported data. For example, parents may overestimate sleep duration, as they may not accurately know when their child falls asleep, wakes up, or experiences night-time awakenings. Similarly, screen time may be underestimated due to concerns about excessive use. Fourth, although multiple confounders were considered, residual confounding cannot be ruled out. Fifth, while the present study focused on sleep duration, other relevant aspects of sleep—such as sleep timing and quality—were not evaluated because they were not assessed in the CORALS study. Sixth, when interpreting our findings, it is important to consider that although the WHO and AAP recommendations are widely used, they have also been criticized for relying partly on expert consensus rather than strong empirical evidence [11]. Seventh, although the COVID-19 pandemic did not have a moderating effect on our models, we cannot rule out its potential influence. Eighth, the study did not account for key contextual factors related to screen use, and the measure used was relatively coarse, as it did not capture the time-of-day children were exposed to screens, whether screens were used in bed, the types of screens used, or whether the activity was passive or interactive. Additionally, the content of screen exposure (e.g., educational vs. entertainment, violent vs. nonviolent) was not assessed, despite its potential relevance to mental health outcomes. Finally, the SDQ scores were unstandardized, and simple imputation methods were used for missing values on the covariates, which may have introduced minor bias.

In conclusion, the present study reinforces the growing body of evidence suggesting that greater screen time is associated with greater emotional and behavioral difficulties in children aged 3–6 years, particularly those with insufficient sleep duration. While our findings indicate that adherence to current screen time guidelines—less than two hours per day or under one hour for children under the age of five—may be beneficial for psychosocial health, the cross-sectional nature of the study limits causal interpretation. Additionally, replacing screen time with sleep may be associated with fewer emotional and behavioral difficulties. These findings support the importance of promoting balanced screen time and healthy sleep habits as part of broader public health initiatives to foster young children’s well-being in increasingly digital environments. Nevertheless, longitudinal and experimental studies are warranted to clarify causal pathways and better inform future evidence-based recommendations.