Longitudinal Associations Between Use of Mobile Devices for Calming and Emotional Reactivity and Executive Functioning in Children Aged 3 to 5 Years

Key Points

Question  Is the use of mobile devices to calm young children’s emotions and behavior associated with long-term difficulties with their executive functioning and emotional reactivity?

Findings  In this cohort study of 422 parents and 422 children, increased use of mobile devices for calming children aged 3 to 5 years was found to be associated with decreased executive functioning and increased emotional reactivity at baseline; however, only emotional reactivity had bidirectional, longitudinal associations with device use for calming at 3 and 6 months of follow-up. The associations were found to be increased in boys and children with higher temperamental surgency.

Meaning  The findings of this study suggest that, particularly in young boys or young children with higher surgency, the frequent use of devices for calming should be avoided.

Importance  Mobile devices are often used to keep young children occupied or calm, but it is not known whether this practice influences child development.

Objective  To examine the longitudinal, bidirectional associations between the parent-reported frequency of using mobile devices to calm young children and children’s executive functioning (EF) and emotional reactivity, testing moderation by child sex and temperament.

Design, Setting, and Participants  This prospective cohort study included a community-based convenience sample of English-speaking parents of typically developing children aged 3 to 5 years. The study duration was from August 2018 to January 2020, with baseline (T1), 3-month follow-up (T2), and 6-month follow-up (T3) waves.

Exposures  Parent-reported frequency of use of mobile devices to calm children when upset (5-point Likert scale).

Main Outcomes and Measures  At each wave, the child’s EF was assessed with the Behavior Rating Inventory of Executive Function–Preschool Version Global Executive Composite and emotional reactivity with the Child Behavior Checklist Emotional Reactivity subscale. Structural equation models were built to examine cross-lagged associations of the use of devices for calming, EF, and emotional reactivity, testing for moderation by child sex or temperament (Child Behavior Questionnaire–Very Short Form surgency score, median split).

Results  Of 422 eligible parents with data at T1, 375 (88.9%) provided data at T2 and 366 (86.7%) at T3. At baseline, the mean (SD) age of the 422 children was 3.8 (0.5) years, the number of boys in the sample was 224 (53.1%), the number of individuals of non-Hispanic White race and ethnicity was 313 (74.2%), and among the parents, 254 (60.2%) had a college degree or higher. Among the boys, the use of devices to calm at T2 was associated with higher emotional reactivity at T3 (r [standardized regression coefficient] = 0.20; 95% CI, 0.10-0.30), while higher emotional reactivity at T2 had a nonsignificant association with increased device use for calming at T3 (r = 0.10; 95% CI, −0.01 to 0.21). Among children with high temperamental surgency, the use of devices to calm at T2 was associated with increased emotional reactivity at T3 (r = 0.11; 95% CI, 0.01-0.22), while higher emotional reactivity at T2 was associated with increased device use for calming at T3 (r = 0.13; 95% CI, 0.02-0.24).

Conclusions and Relevance  The findings of this study suggest that the frequent use of mobile devices for calming young children may displace their opportunities for learning emotion-regulation strategies over time; therefore, pediatric health care professionals may wish to encourage alternate calming approaches.

Early childhood is a critical window for developing higher-order emotional and cognitive processes. Research suggests that processes such as executive functioning (EF) and emotion regulation are more important for school success than crystallized intelligence, as they allow children to stay calm, focused, and flexible as they face new challenges.^1,2 Executive functioning is a multidimensional construct that encompasses inhibitory control, working memory, and attention flexibility.³ Deficits in EF are apparent in children with attention-deficit/hyperactivity disorder (ADHD) and underlie many problematic everyday behaviors, such as poor impulse control and difficulty following directions. Emotion regulation is a construct that includes both the propensity for emotional reactivity and the capacity to calm down when upset.^4,5 These skills develop rapidly from ages 2 to 5 years,⁶ concordant with frontal lobe development, and are thought to develop through transactional processes between the child’s innate characteristics and their caregiving environment.^7-9 Digital media is one aspect of young children’s environments¹⁰ that has become particularly salient in the past 2 decades and during the COVID-19 pandemic.¹¹

Prior longitudinal research on media use and EF-related outcomes has primarily examined television (TV) or global screen time estimates. For example, low-quality TV viewing has been associated with ADHD,¹² preschooler EF deficits,^13-15 and early academic problems.^16,17 Yet, mobile devices (ie, smartphones, tablets) may have a unique association with child attention and impulse inhibition, as they provide on-demand handheld access and thus can fragment daily routines.¹⁸ Popular mobile content, such as video-sharing sites and games, have high levels of distracting enhancements and advertising.^19-21 Research regarding mobile device use and EF has been limited to date; 1 study found that parent-reported duration of the use of mobile apps was associated with reduced overall EF 1 year later.²² Our prior work has shown cross-sectional links between reduced parent-reported EF and the use of devices for calming purposes,²³ which suggests that more impulsive children might develop heavier device use practices, and this may displace opportunities to practice goal-directed behavior.^24,25

Similarly, research on emotion regulation—such as early externalizing behavior,^26,27 social-emotional screening scores,²⁸ and preschooler behavioral problems^29,30—has largely examined TV or screen time, not mobile device use. The American Academy of Pediatrics (AAP) recommends limiting the use of mobile devices for emotional calming purposes,³¹ based on limited cross-sectional evidence that infants with difficult temperament,³² fussing,³³ self-regulation problems,³⁴ and social-emotional delays³⁵ have greater TV and mobile device use, but no longitudinal studies have studied the practice of using devices for calming, to our knowledge.

Parents may wonder whether using mobile devices to settle down a young child has long-term consequences for development, or whether it is a benign temporary parenting strategy that reduces household stress. Particularly in early childhood, the frequent use of devices to calm distress may displace opportunities for development of independent and alternative self-regulation strategies.

This study, therefore, aimed to examine longitudinal, bidirectional associations between the use of mobile devices for calming and EF and emotional reactivity over the course of 6 months in young children, testing moderation by child sex and temperament. Current theoretical models emphasize the importance of children’s individual differences in biological sensitivity to context,³⁶ including media contexts.³⁷ Moreover, emotion socialization practices³⁸ and media habits differ by child sex³⁹ and temperament traits (eg, surgency)^40,41; thus, studies that investigate these differences are needed to craft more precise guidance about when, and for whom, the use of mobile devices for calming purposes should be avoided.

Data for this longitudinal cohort study were collected from August 2018 to January 2020 through web-based surveys at baseline (T1), 3-month follow-up (T2), and 6-month follow-up (T3) waves. Parents provided online written consent for themselves and their child and received up to $150 for completion of all 3 waves. This study was approved by the institutional review board of the University of Michigan. We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

The sample and study methods have been described in detail previously.⁴² A community-based convenience sample of young children was recruited via flyers posted in childcare centers, pediatric clinics, the university’s online participant registry, and social media advertisements. The eligibility criteria included (1) parent or legal guardian of a 3- to 5-year-old child; (2) lives with the child more than 5 days per week; (3) English-speaking; and (4) family owned at least 1 Android or iOS device.

To improve generalizability, children did not need to regularly use mobile devices to be included in the study. Exclusion criteria included presence of child developmental delays or use of psychotropic medication. Following the STROBE cohort study reporting guidelines, participant characteristics are shown by data collection wave in eTable 1 in the Supplement; 422 participants provided data at T1, 375 (88.9%) at T2, and 366 (86.7%) at T3. The retained participants at T2 were significantly more likely to be non-Hispanic White individuals and girls and to have higher parental educational attainment and a higher income, while the retained participants at T3 were more likely to be girls and to have higher parental educational attainment.

Parents completed the Comprehensive Assessment of Family Media Exposure (CAFE) Consortium Qualtrics survey,⁴³ which assesses child and family media use behaviors, context, and content. Use of mobile devices for calming purposes was assessed with the question “When your child is upset and needs calming down, how likely are you to give him/her a mobile device to use, like a smartphone or tablet?” Parents responded on a 5-point scale ranging from 0 (not at all likely) to 4 (very likely). This question format has been used in prior studies.^35,44

Parents completed additional surveys via REDCap,^45,46 including the Behavior Rating Inventory of Executive Function–Preschool Version (BRIEF-P),⁴⁷ a parent-reported measure of early childhood EF (eg, “is impulsive,” “needs help from adult to stay on task”) that correlates with observed EF assessments.⁴⁸ Raw scores were transformed into age- and sex-adjusted T-scores for the Global Executive Composite (GEC) (α = .96); higher scores indicate reduced EF.

Parents also completed the emotional reactivity subscale of the Child Behavior Checklist–Preschool (CBCL-P),⁴⁹ a valid and reliable measure of child behavior for ages 18 months to 5 years. The emotional reactivity subscale (T1, α = .75) sums 9 symptoms (eg, “rapid shifts between sadness and excitement,” “sudden changes in mood or feelings”) rated on a 3-point scale ranging from 0 (not true) to 2 (very true); higher scores indicate higher emotional reactivity.

Parents reported their child’s age, sex, race and ethnicity (Table 1 provides investigator-defined categories), preschool or childcare enrollment, and prematurity; their own age, sex, educational attainment, marital status, and employment status; and household income and size (from which we calculated the income to needs ratio). Race and ethnicity were examined as a covariate because of extensive research showing associations between child race and ethnicity and media use behaviors as well as developmental outcomes⁵⁰—both thought to be due to a systemic lack of opportunities, poverty, and structural racism. Parents self-reported their child’s race in investigator-determined categories and reported ethnicity separately (Hispanic or non-Hispanic). Due to the small sample sizes in the individual race and ethnicity categories, the variable was dichotomized as “underrepresented minority vs non-Hispanic White.” As an assessment of child surgent temperament, parents completed the Rothbart Child Behavior Questionnaire–Very Short Form,⁵¹ a valid and reliable measure of temperament for ages 3 to 7 years. We used the surgency subscale (α = .76, a mean of 12 items [eg, “often rushes into new situations,” “is full of energy, even in the evening”], rated on a 7-point scale from 1 [extremely untrue] to 7 [extremely true]) because surgency qualities of approach, interest in novelty, extroversion, persistence, and intensity have been associated with media use^40,41 and parent-child interaction quality.⁵²

The sample size (presuming 15.0% attrition per wave) was finalized based on estimates that we could detect standardized estimates for each path of small magnitude (0.14-0.16) with 80.0% power using 1-sided type I error α = .05 (P < .05).

Path models were constructed using Mplus, version 8.4 (Muthen & Muthen) to test concurrent, auto-correlations, and cross-lagged associations between the use of devices for calming purposes and BRIEF-P GEC T-score or CBCL-P emotional reactivity score at T1, T2, and T3, stratifying all analyses by males vs females (no children identified as nonbinary or other sex) and high vs low surgency score (split at the median value). We used maximum likelihood estimation techniques in Mplus to fit all cross-lagged models, using the comparative fit index (CFI)⁵³ and standardized root mean square residual (SRMR)⁵⁴ to assess model goodness of fit. We chose not to use a random intercept cross-lagged panel model because emotional reactivity and use of devices for calming would be expected to vary at different rates among children over time. The full information maximum likelihood method, which uses all available data, was used because it is the recommended method for addressing incomplete data when fitting structural equation models.⁵⁵

We controlled for potential confounders that were theoretically associated with EF, emotional reactivity, or media use; these included child age, sex, parent educational attainment (high school degree equivalent or less, some college or 2-year degree, 4-year college degree, or advanced degree), child preschool attendance (center-based preschool, home-based child care, or none), household income to needs ratio, and child race and ethnicity.

Among the 422 parents, 395 (93.6%) were females and the mean (SD) age of the entire sample was 34.0 (4.7) years. In addition, 254 parents (60.2%) had a college degree or higher. Among the 422 children sampled, 224 (53.1%) were males and the mean (SD) age of the entire sample was 3.8 years (0.5). In the sample, the race and ethnicity of 313 children (74.2%) were non-Hispanic White. Other sample characteristics are given in Table 1. At T1, 36 parents (8.5%) reported being very likely to use mobile devices to calm their child when upset; these numbers were 29 of 375 (7.7%) at T2 and 29 of 366 (7.9%) at T3.

Standardized regression coefficients for the full sample are shown in the Figure and eTable 2 in the Supplement. Because model fit for the overall sample was weak (ie, CFI < 0.9 and higher SRMR), only the stratified model results are presented in the subsequent section.

As shown in Table 2, BRIEF-P GEC T-scores showed within-person stability over time (r [standardized regression coefficient] = 0.80-0.82 in males, 0.79-0.82 in females), as did parent-reported use of devices for calming (0.63-0.72 in males, 0.50-0.58 in females). In males, EF showed a significant concurrent correlation with use of devices for calming purposes at baseline (T1) (r = 0.33, 95% CI, 0.21-0.46; P < .001), but not at T2 or T3, nor in cross-lagged associations between T1 to T2, or T2 to T3. Females, for whom the model fit was weaker, showed no significant concurrent correlations, but higher use of devices for calming purposes at T1 was associated with higher BRIEF-P GEC scores at T2 (r = 0.12; 95% CI, 0.03-0.22; P = .01).

The CBCL-P emotional reactivity score showed within-person stability over time (r = 0.68-0.70 in males, 0.57-0.65 in females). In males, emotional reactivity was significantly associated with use of devices for calming purposes concurrently (at T1, r = 0.22; 95% CI, 0.08-0.35; P < .001), and in a cross-lagged manner between T2 and T3. Higher use of mobile devices for calming purposes at T2 was associated with higher emotional reactivity at T3 (r = 0.20; 95% CI, 0.10-0.30; P < .001), while converse associations were not statistically significant (r = 0.10; 95% CI, −0.01 to 0.21; P = .06). Concurrent and cross-lagged associations were not significant for female participants.

As shown in Table 3, children with high surgency showed concurrent correlations between BRIEF-P GEC score and use of devices for calming purposes at T1 (r = 0.16; 95% CI, 0.02-0.30; P = .02). Children with low surgency also showed significant T1 concurrent correlations between the BRIEF-P GEC score and the use of devices for calming (r = 0.24; 95% CI, 0.10-0.38; P < .001). In children with high surgency, there was a substantial but insignificant path between weaker EF (higher BRIEF-P GEC score) and increased use of devices for calming between T1 and T2 (r = 0.11; 95% CI, −0.01 to 0.22; P = .07). In children with low surgency, use of devices for calming purposes at T1 was associated with reduced EF at T2 (r = 0.09; 95% CI, 0.01-0.18; P = .047).

Although T1 concurrent correlations between CBCL-P emotional reactivity score and device use for calming were not significant in children with high surgency, they were significant in children with low surgency (at T1, r = 0.19; 95% CI, 0.04-0.33; P = .01). Cross-lagged associations were significant only in children with high surgency between T2 and T3: higher use of mobile devices for calming purposes at T2 was associated with higher emotional reactivity at T3 (r = 0.11; 95% CI, 0.01-0.22; P = .03), while higher emotional reactivity at T2 was associated with higher use of mobile devices for calming purposes at T3 (r = 0.13; 95% CI, 0.02-0.24; P = .02). Goodness of fit for the cross-lagged model varied by moderating variables (Tables 2 and 3); models for males and children with high surgency showed good fit for both BRIEF-P and CBCL-Emotional Reactivity (CFI > 0.9 and SRMR < 0.1), but poorer model fit for females and children with low surgency.

In this longitudinal cohort study examining bidirectional associations between mobile device use and development in 3- to 5-year-old children, we found that both child EF and emotional reactivity had baseline concurrent associations with the use of devices to calm, but the latter showed more consistent association with device use for calming over time. Specifically in boys and children with higher temperamental surgency, between T2 (3-month follow-up) and T3 (6-month follow-up), higher emotional reactivity was associated with an increased use of devices for calming purposes, and an increased use of devices for calming with higher emotional reactivity.

Reasons for strong baseline correlations between EF, emotional reactivity, and use of devices for calming may reflect the fact that children were at their youngest age at that data collection wave. Because EF and emotion regulation mature throughout early childhood, use of devices for calming may have been associated with these developmental skills at a point when they were weakest. In other words, a higher degree of tantrums, intense emotionality, defiance, or lack of behavioral inhibition may have triggered the increased use of devices as a parenting strategy at this earlier time.

Device use for calming may have been associated with emotional reactivity more so than with EF for a few reasons. Decline in children’s negative affect in response to device use may be rapid and overt, and thus rewarding to both parents and children, establishing motivation for maintaining this cycle. Executive functioning skills such as distractibility and impulse inhibition may not be as responsive to device use in the moment.

Moreover, it is possible that cross-lagged correlations between device use for calming and emotional reactivity stood out between T2 and T3 because the habit of using devices to manage difficult behavior strengthened over time, as the parent and child had less practice in other emotion-regulation strategies. This proposed mechanism is consistent with the evidence that children’s media preferences strengthen over time,⁵⁶ so it is possible that parents allowed device use in response to children’s demands for this preferred activity.

Longitudinal associations between emotional reactivity and device use for calming were increased in boys compared with that in girls. Sex differences in emotion regulation among preschool-aged children have long been identified,⁵⁷ with boys often demonstrating delayed skills. Reasons for this can include differences in emotion socialization,⁵⁸ differences in response to parenting factors,⁵⁹ and generally greater emotional immaturity of young male children.⁵⁷ There may also be sex-related differences in device use habits based on media content and design. For example, boys and girls have been noted to use different types of digital media, with boys engaging in content that may be more stimulating⁶⁰ or engaging, which may lead to more intense demands for media.

We also found more associations between emotional reactivity and device use for calming in children with higher surgency at baseline. These findings are consistent with recent publications showing that high surgency correlates with media use for calming in cross-sectional data⁴¹ and moderates the associations between psychosocial stress and media use.⁴⁰ Mechanisms may include higher parenting stress in response to rapid shifts in child behavior, with a stronger reward cycle when child behavior is immediately calmed with a device. Children with more surgent temperaments also require more external scaffolding from parents to develop self-regulation,⁶¹ which is displaced when devices are used.

Our findings regarding EF were less consistent. Weaker EF was associated only with later use of devices for calming in children with higher surgency from T1 to T2 but was not statistically significant. In females and children with low surgency, increased use of devices for calming at baseline (T1) was associated with decreased EF at T2, although the fit indices were weaker in these models; therefore, these results should be interpreted with caution. Replication in future samples is needed.

These findings contribute to the literature on early childhood media use that is increasingly leveraging longitudinal samples to tease out the bidirectional influences between media exposure and child development while accounting for the between-participant confounders that plague prior observational work. For example, Madigan and colleagues²⁸ have found that longer screen media use is associated with poorer performance on developmental screening tests, but not the converse. McArthur and colleagues³⁰ identified bidirectional associations between media and parent-reported behavioral and developmental problems. However, prior research has examined that it is the duration of screen time, rather than how media and mobile devices are used throughout daily routines, that shape children’s social-emotional development. In this study, we considered the use of devices specifically for the purposes of calming a child showing negative emotion, a practice that many parents endorse.⁶²

The strengths of our study include examination of 2 different domains relative to child well-being finding distinct patterns that suggest emotional reactivity is more closely tied to this practice than is EF. The results support the AAP recommendations for the judicious use of mobile devices for calming purposes, for example, intentional use on airplane flights and long car rides.

A major limitation of this study is the use of a single question to assess the use of mobile devices for calming purposes. Although this measure has been used in prior studies,^35,44 it has not been validated against objective measures; new questionnaires that assess device use in multiple behavioral contexts are needed.

Future research should develop more objective ways of assessing regulatory uses of media, such as through audio recordings or ecological momentary assessment. In addition, our sample had a higher proportion of children of White race and ethnicity whose parents had higher educational attainment than the general US population, particularly in participants retained across all periods. In addition, residual confounding is possible, which we attempted to address by adjusting for multiple covariates. Replication in different cohorts and ages is needed.

Although using videos, apps, or photographs on a device may be effective in distracting or assuaging a young child’s distress in the moment, the results of this longitudinal cohort study suggest that this practice may become a more frequent habit with more emotionally reactive children, and that this may worsen their emotion-regulation skills over time. Particularly for boys and children with more surgent temperaments, pediatric health care professionals may wish to encourage alternative methods and therapeutic supports for emotion regulation from an early age.

Accepted for Publication: September 16, 2022.

Published Online: December 12, 2022. doi:10.1001/jamapediatrics.2022.4793

Corresponding Author: Jenny S. Radesky, MD, Department of Pediatrics, University of Michigan Medical School, 300 N Ingalls St, #1107, Ann Arbor, MI 48109 (jradesky@med.umich.edu).

Author Contributions: Drs Radesky and Kaciroti had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Radesky, Schaller, Miller.

Acquisition, analysis, or interpretation of data: Radesky, Kaciroti, Weeks.

Drafting of the manuscript: Radesky, Kaciroti, Miller.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Kaciroti, Weeks.

Obtained funding: Radesky, Kaciroti.

Administrative, technical, or material support: Radesky, Schaller, Miller.

Supervision: Radesky, Schaller.

Conflict of Interest Disclosures: Dr Radesky reported receiving personal fees from Noggin (Viacom/CBS) for serving on their scientific advisory board in 2021, and consulting fees from Melissa & Doug Toys outside the submitted work. Dr Kaciroti reported receiving grants from the University of Michigan during the conduct of the study. No other disclosures were reported.

Funding/Support: This research was funded by grant 1R21HD094051 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, awarded to Drs Radesky (principal investigator), Miller (coinvestigator), and Kaciroti (coinvestigator). REDCap and recruitment support was provided by the Michigan Institute for Clinical & Health Research (Clinical and Translational Science Award: grant number UL1TR002240).

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.