- Email: [email protected]
Dependency and Psychological Well-Being Among Late Adolescents
Journal of Adolescent Health xxx (2019) 1e6
www.jahonline.org Original article
Short-Term Longitudinal Relationships Between Smartphone Use/Dependency and Psychological Well-Being Among Late Adolescents Matthew A. Lapierre, Ph.D. *, Pengfei Zhao, M.A., and Benjamin E. Custer, M.A. Department of Communication, University of Arizona, Tucson, Arizona
Article history: Received January 17, 2019; Accepted June 5, 2019 Keywords: Smartphone dependency; Smartphone use; Depressive symptoms; Loneliness; Late adolescents
A B S T R A C T
Purpose: The aim of the study was to determine the short-term longitudinal pathways between smartphone use, smartphone dependency, depressive symptoms, and loneliness among late adolescents. Methods: A two-wave longitudinal survey was used using adolescents between the ages of 17 and 20 years. The interval between wave 1 and wave 2 was between 2.5 and 3 months. Using convenience sampling, the total number of participants who completed both waves of data collection was 346. Validated measures assessed smartphone dependency, smartphone use, depressive symptoms, and loneliness. The longitudinal model was tested using path modeling techniques. Results: Among the 346 participants (33.6% male, mean [standard deviation] age at wave 1, 19.11 [.75] years, 56.9% response rate), longitudinal path models revealed that wave 1 smartphone dependency predicted loneliness (b ¼ .08, standard error [SE] ¼ .05, p ¼ .043) and depressive symptoms (b ¼ .11, SE ¼ .05, p ¼ .010) at wave 2, loneliness at wave 1 predicted depressive symptoms at wave 2 (b ¼ .21, SE ¼ .05, p < .001), and smartphone use at wave 1 predicted smartphone dependency at wave 2 (b ¼ .08, SE ¼ .05, p ¼ .011). Conclusions: Considering the rates of smartphone ownership/use among late adolescents (95%), the association between smartphone use and smartphone dependency, and the deleterious effects of loneliness and depression within this population, health practitioners should communicate with patients and parents about the links between smartphone engagement and psychological well-being. Ó 2019 Society for Adolescent Health and Medicine. All rights reserved.
In just over a decade, the smartphone has become a technological necessity. The Pew Research Center indicates that 77% of American adults own a smartphone, and that such ownership is nearly universal (95%) for adolescents [1,2]. Yet, with the growing
Conflicts of interest: The authors have no conflicts of interest to disclose. * Address correspondence to: Matthew A. Lapierre, Ph.D., Department of Communication, University of Arizona, 1103 E University Blvd, P.O.Box 210025, Tucson, AZ 85721. E-mail address: [email protected] (M.A. Lapierre). 1054-139X/Ó 2019 Society for Adolescent Health and Medicine. All rights reserved. https://doi.org/10.1016/j.jadohealth.2019.06.001
IMPLICATIONS AND CONTRIBUTION
This research sheds light on existing studies by investigating the directionality between smartphone engagement and psychological well-being among late adolescents. The findings reveal that smartphone engagement is a significant predictor of subsequent depressive symptoms and loneliness. This research can help practitioners understand the detrimental effects of smartphone engagement on adolescent well-being.
importance of smartphones in teens’ daily lives, there is some concern that these devices are interfering with their overall health and well-being [3]. Of particular concern is just how smartphones are affecting users. Researchers have reported mixed findings for the association between smartphone use and well-being [4]. Some studies have suggested that smartphone use may be positively associated with well-being [5]. For example, George et al. found that the number of text messages was negatively associated with depressive symptoms [6]. However, other studies have suggested
2
M.A. Lapierre et al. / Journal of Adolescent Health xxx (2019) 1e6
that smartphone use contributes to poor mental health [3]. In addition, recent research has demonstrated that smartphone use is a predictor of smartphone dependency, which, in turn, negatively affects mental health [7]. Smartphone dependency (also known as “problematic smartphone use” or “smartphone addiction”), which is characterized by excessive psychological attachment on smartphones with negative functional consequences [8], has been suggested to warrant considerable concerns among adolescent populations [9]. Although smartphone dependency is not included in DSM-5 [10] because of further examinations required, existing studies have suggested that smartphone dependency should be considered as a behavioral addiction functioning similarly with substance addictions such as alcohol and drugs [11] with shared characteristics including loss of control, tolerance, and withdrawal [12]. Smartphone dependency may be less of a concern than smartphone use, but smartphone dependency has been associated with a number of problematic outcomes including mental health issues [13,14]. Among these mental health issues, depression and loneliness are of paramount concern because they are associated with increased prevalence of chronic diseases, early morbidity/mortality, and suicide [15,16]. Previous cross-sectional research in this area has found a consistent link between depressive symptoms and loneliness with smartphone dependency [17]. However, there is no consensus on the directionality of these relationships, mostly because few longitudinal examinations have been conducted. One recent study that examined the longitudinal relationship between depression and smartphone dependency found that depression at wave 1 predicted dependency at wave 2 (mediated via mindfulness); however, the authors did not account for smartphone dependency at wave 1 [18]. Moreover, this finding is inconsistent with previous longitudinal studies on depression and general Internet dependency (which should function similarly to smartphone dependency) that have found that Internet dependency predicts both future depression and loneliness [19]. Also, the inconsistent findings raise the possibility that the relationship between smartphone dependency and psychological well-being is bidirectional. Namely, depressive symptoms and loneliness predict increased smartphone dependency over time, and increased smartphone dependency exacerbates depressive symptoms and loneliness over time. However, more longitudinal studies are needed to shed light on this issue. This set of findings is of particular importance because late adolescents (aged 18e20 years) are vulnerable to poor mental health, such as depression [20]; consequently, the focus of the present study is to determine the directionality between smartphone engagement (i.e., use, dependency) and psychological well-being (i.e., depressive symptoms, loneliness) among late adolescents. Previous studies have established that there is a clear relationship between smartphone dependency and reduced psychological well-being [17]. However, what is unknown is whether smartphone dependency fuels diminished well-being or if the relationship works in the other direction (i.e., depressive symptoms/loneliness breeds smartphone dependency). In addition, because of mixed findings, the association and directionality between smartphone use and well-being are also underdetermined. To that end, this study uses a longitudinal examination of late adolescents’ smartphone use, smartphone dependency, loneliness, and depressive symptoms to help answer these important questions.
Methods Recruitment College students enrolled in communication and sociology courses were recruited from three large public universities in the American Southwest, Midwest, and Southeast. The research was approved by the authors’ institutional review board (protocol# 18-002-COMM). The first wave of data collection occurred at the end of January and the beginning of February 2018. The second wave of data collection took place during the last 2 weeks of April 2018. To ensure the waves of data matched, participants were asked to create a unique and anonymous identifier. Sample The total of potential participants contacted for participation in this study was 816. In the first wave of the study, 585 participants (72%) successfully completed the survey; at the second wave, 590 participants completed the survey. However, there were participants who completed the survey at wave 1 who did not take it at wave 2 (and vice versa), and the total number of participants who completed both waves of data collection was 464; namely, 79% of participants who completed the survey at wave 1 also completed the follow-up. To keep the focus of this study on late adolescents, all participants aged 21 years were removed from the sample resulting in 346 total respondents. The sample was 33.6% male (n ¼ 116), and the mean age of participants in the sample at wave 1 was 19.11 years (standard deviation [SD] ¼ .75). Participants in the study racially identified as white (n ¼ 230, 66.5%), African American (n ¼ 14, 4.0%), Asian/ Pacific Islander (n ¼ 51, 14.7%), Latino/a (n ¼ 25, 7.2%), Native American (n ¼ 1, .34%), “other” (n ¼ 2, .6%), and multiracial (n ¼ 23, 6.6%). Among the universities that participated, 50.3% (n ¼ 174) came from the university in the Midwest region, 39% (n ¼ 135) from the southwestern university, and 10.7% (n ¼ 37) from the southeastern university. Measures Smartphone dependency. Smartphone dependency was measured using the smartphone addiction proneness scale by Kim et al. [8]. This is a 15-item scale where participants are asked to indicate their level of agreement with statements regarding their smartphone engagement (e.g., “I panic when I cannot use my smartphone”). All questions were asked on a 4-point scale (0 ¼ strongly disagree to 3 ¼ strongly agree; W1: mean ¼ 1.21, SD ¼ .45, a ¼ .88; W2: mean ¼ 1.19, SD ¼ .47, a ¼ .89). Smartphone use. Participants were asked to estimate their daily smartphone use in minutes across eight common uses for smartphones. These were talking on the phone, texting, using e-mail, surfing the Web, using social networks (e.g., Instagram and Facebook), using FaceTime or other video chat, playing games, and/or using news apps. Participants’ use across these eight categories were summed to create a measure of smartphone use in minutes per day. Because of significant positive skew, the measure was transformed using a log transformation (W1: mean ¼ 2.54, SD ¼ .29; W2: mean ¼ 2.52, SD ¼ .28). Loneliness. The measure of loneliness for this study was the UCLA Loneliness scale [21]. This is an 8-item measure that asks
M.A. Lapierre et al. / Journal of Adolescent Health xxx (2019) 1e6
participants to indicate how often certain statements apply to them, such as “I feel left out” and “I lack companionship.” Each question was measured on a 4-point scale (0 ¼ never to 3 ¼ always) with scores averaged for participants (W1: mean ¼ 1.00, SD ¼ .61, a ¼ .83; W2: .98, SD ¼ .63, a ¼ .85). Depressive symptoms. The 10-item Center for Epidemiologic Studies Depression Scale was used as the measure for depressive symptoms [22]. The scale asks participants to indicate how many days in the last week they felt a particular way. For example, participants are asked if they were “bothered by things that usually do not bother me.” All questions were answered using a 4-point scale (0 ¼ rarely or none of the time [less than one] to 3 ¼ most of all of the time [5e7]) and were summed to create the measure (W1: mean ¼ 9.58, SD ¼ 5.07, a ¼ .80; W2: mean ¼ 9.71, SD ¼ 5.52, a ¼ .83). Analysis strategy Table 1 shows all zero-order correlations for the variables of interest. The longitudinal model was tested using path modeling techniques in AMOS 23. The initial test used a saturated path model paths not significant at the one-tailed p .10 subsequently removed [23]. To evaluate the appropriateness of the path model, chi-square, the root mean squared error of approximation (RMSEA), the comparative fit index, and the normed fit index were used to assess the model fit. For the path model, the standards used to determine fit were RMSEA at .05, a non-significant c2, and goodness-of-fit indices above .90 [23]. Finally, to guard against multicollinearity in the model, both measures of loneliness and depressive symptoms were centered because of their moderately large correlation [24]. Results The initial saturated model revealed that the following paths were not significant: wave 1 depressive symptoms to wave 2 smartphone dependency (b ¼ 0, SE ¼ .04, p > .99), wave 1 depressive symptoms to wave 2 smartphone use (b ¼ .01, SE ¼ .03, p ¼ .92), wave 1 smartphone dependency to wave 2 smartphone use (b ¼ .02, SE ¼ .02, p ¼ .65), wave 1 loneliness to wave 2 smartphone use (b ¼ .04, SE ¼ .02, p ¼ .40), wave 1 smartphone use to wave 2 loneliness (b ¼ .01, SE ¼ .08, p ¼ .87), and wave 1 smartphone use to wave 2 depressive symptoms (b ¼ .01, SE ¼ .08, Table 1 Zero order correlations for variables of interest Variable
2
3
4
.139** 1. W1: smartphone .215*** .012 use 2. W1: smartphone .241*** .297*** dependency 3. W1: loneliness .628*** 4. W1: depressive symptoms 5. W2: smartphone use 6. W2: smartphone dependency 7. W2: loneliness 8. W2: depressive Symptoms N ¼ 346; W1 ¼ Wave 1; W2 ¼ Wave 2. * p < .05; **p < .01; ***p < .001.
5
6
7
.722*** .243*** .017
8 .090
.183**
.786*** .250*** .295***
.039 .136*
.250*** .743*** .524*** .278*** .500*** .631*** .268*** .030
.107*
.356*** .409*** .666***
3
p ¼ .89). In addition, the wave 1 association between loneliness and smartphone use was not significant (b ¼ .01, SE ¼ .01, p ¼ .82), and the error term associations between wave 2 smartphone use to both depressive symptoms (b ¼ .02, SE ¼ .004, p ¼ .68) and loneliness at wave 2 (b ¼ .04, SE ¼ .004, p ¼ .43) were not significant. The adjusted model with nonsignificant paths removed demonstrated good model fit c2(10) ¼ 4.24, p ¼ .94, as both goodness-of-fit indices were above recommended thresholds (comparative fit index ¼ 1.000; normed fit index ¼ .997) and the RMSEA < .001. There was one path that was not significant as the link from wave 1 loneliness to wave 2 smartphone dependency was only marginally significant (b ¼ .06, SE ¼ .03, p ¼ .060; Figure 1 for final model). Of particular note, the following paths were significant: smartphone dependency at wave 1 predicted loneliness at wave 2 (b ¼ .08, SE ¼ .05, p ¼ .043), smartphone dependency at wave 1 predicted depressive symptoms at wave 2 (b ¼ .11, SE ¼ .05, p ¼ .010), loneliness at wave 1 was associated with depressive symptoms at wave 2 (b ¼ .21, SE ¼ .05, p < .001), and smartphone use at wave 1 predicted smartphone dependency at wave 2 (b ¼ .08, SE ¼ .05, p ¼ .011). Discussion The aim of this study was to determine the longitudinal pathways between smartphone use, smartphone dependency, depressive symptoms, and loneliness among late adolescents. Depression and loneliness represent serious impediments to adolescent health [15,16] with evidence that depression becomes more severe as teens age [20]. This is particularly concerning when considering the increased prevalence and use of smartphones within this population [1], as a number of cross-sectional studies have found that smartphone use/dependency may be linked to loneliness and depressive symptoms [17]; however, the directionality of these relationships was still undetermined. Our longitudinal results showed that smartphone dependency predicted increased depressive symptoms 3 months later for late adolescents. This finding is inconsistent with the previous longitudinal study on smartphone dependency and depression, which demonstrated that depression leads to later smartphone dependency via mindfulness [18]. But, it is important to note that the authors of this previous study did not account for dependency at wave 1. However, our findings do align with previous longitudinal studies on depression and general Internet dependency. For example, Muusses et al. conducted a five-wave longitudinal study and found that compulsive Internet use (i.e., being incapable of regulating one’s Internet use) predicted later depression [19]. Previous studies provide several reasons to explain this relationship. One is based on the displacement hypothesis, which states that time spent on the Internet displaces the time spent on offline communication with close friends and family members [25]. Compared with offline communication, online interaction may sometimes engender less social support and intimacy [26], leading to feelings of social isolation and detachment. Another reason might be that Internet dependency causes vocational impairment, such as loss of productivity in academic or occupational performance [27], which, in turn, makes one experience negative emotions. Smartphone dependency may predict depressive symptoms via these same underlying mechanisms such as lack of social support from strong ties or vocational impairment, which should be investigated in future studies. Nevertheless, the findings from
4
M.A. Lapierre et al. / Journal of Adolescent Health xxx (2019) 1e6
Figure 1. Final longitudinal model.
the present study are concerning. Specifically, although Internet dependency certainly shares some characteristics with smartphone dependency, Internet dependency is associated with different constructs than smartphone dependency [28], and there are qualitative differences regarding how smartphones are involved in our daily lives. Internet dependency refers to problematic use of the Internet chats, games, and information seeking and has typically been associated with general computer use [29]. Smartphones offer similar affordances but extend the potential risk via their portability, wireless data access, and wide range of applications that make them “indispensable” devices [30]. As such, the potential risks they pose to the mental health of young people might be amplified, yet the possible amplified effect requires further examination. Our results further demonstrated that smartphone dependency significantly predicted later loneliness among participants. Previous research has shown that loneliness and smartphone dependency are linked [31,32]; however, this is the first known study to show that smartphone dependency precedes loneliness. There is some overlap with previous research on Internet dependency and loneliness. For example, Jia et al. found that Internet dependency did predict later loneliness among Chinese university students [33]. Yet, in that study, they also found that loneliness was predictive of later Internet dependency. The present study only found a marginally significant link between wave 1 loneliness and wave 2 smartphone dependency. Our findings suggest that smartphones are not a favorable surrogate compared with offline interaction in terms of providing social support and intimacy [26]; consequently, reliance on smartphones potentially removes people from
meaningful interpersonal connections, which may engender feelings of increased loneliness [34]. It is also important to discuss the magnitude of these effects for smartphone dependency on loneliness/depressive symptoms and their practical importance for the health and well-being of adolescents. As seen in Figure 1, the magnitude of these effects would traditionally be considered small [35]. However, there are two things to note regarding the size of these effects. First, the effects between smartphone dependency and depressive symptoms are larger than what has been previously found in research looking at Internet dependency and this same variable (b ¼ .11 vs. b ¼ .06) [19]. Second, because we controlled for both loneliness and depressive symptoms at wave 1 in the analyses, the amount of unaccounted for variance at wave 2 for these variables is much smaller; this happens because of the autocorrelations between waves and the fact that loneliness and depressive symptoms are highly correlated with one another (r ¼ .63). As such, the amount of residual variance left for smartphone dependency to predict either of these variables is quite small. However, this amplification effect for dependency on these variables is important, particularly considering the tremendous negative outcomes associated with both loneliness and depressive symptoms [15,16] as any increase in either of these variables can lead to serious problems in adolescents. The role that smartphone use played in the longitudinal model was interesting but not unexpected, as previous crosssectional research has shown that smartphone use and smartphone dependency are linked [7,36]. In the present study, smartphone use at wave 1 was associated with dependency at wave 2 but was not associated with depressive symptoms or
M.A. Lapierre et al. / Journal of Adolescent Health xxx (2019) 1e6
loneliness in the model in either direction. This finding stands in contrast to recent theorizing about smartphone use’s influence on psychological well-being, which has suggested that use, on its own, is associated with such outcomes [3,5]. In other words, although smartphone ownership is nearly universal for adolescents, smartphone use itself may not be detrimental to mental health. Rather, only those who develop smartphone dependency from smartphone use may experience negative outcomes with respect to psychological well-being. Existing research has suggested that smartphone use may lead to smartphone dependency via habitual smartphone use [7]. Namely, the multifunction, accessibility, and automatic checking of smartphone might develop smartphone use to be a habit, which contributes to later smartphone dependency. Therefore, these results suggest that if researchers wish to understand the link between smartphone use and well-being, it is critical to explore the role played by smartphone dependency. However, the present study focused on negative constructs (i.e., depressive symptoms and loneliness), so future research should explore the directionality of the associations between smartphone engagement and positive constructs (e.g., social support and social capital). Implications How might practitioners use the findings from the present study to improve adolescent health? First, it is necessary to acknowledge that late adolescents are vulnerable to becoming depressed and lonely, particularly when many of these young people are making significant life changes such as joining the workforce or moving away to college [37]. Therefore, understanding the directionality of the relationship between smartphone engagement and depressive symptoms and loneliness may shed light on late adolescents’ mental health maintenance. Pediatricians and other practitioners should consult with teen patients about their smartphone use and help them understand the links between use, dependency, and psychological wellbeing. Practitioners could also recommend that their patients use smartphone applications that help to limit the time spent with these devices (e.g., Moment and Flipd) [38].
5
completed the first wave, those who dropped out after wave 1 scored lower on the depressive symptoms measure than participants who finished both data collections (mean ¼ .83 vs. mean ¼ .96; Hedges’ g ¼ .24, 95% confidence interval [.50, .02]). However, it is unclear how this would challenge the overall findings because the present study was looking at within-person change on this measure over time. Fourth, this study did not use clinical thresholds as an indicator of depression, and the means of some constructs such as depressive symptoms and loneliness were low. However, this study did find that the variation of smartphone dependency can contribute to the variations of depressive symptoms and loneliness. Namely, smartphone dependency can be a predictor for subsequent depressive symptoms and loneliness among older adolescents. The final limitations with the study were the nature of the smartphone use measure and the validity of the smartphone dependency measure. First, the present study used a self-report measure rather than direct observations to assess the extent to which participants used their smartphone. Future research should use direct observations of smartphone use (i.e., using smartphone applications such as Moment) to further validate the findings of the present study. Second, the smartphone dependency measurement this study used was developed and validated on South Korean samples. To our knowledge, because the phenomenon of smartphone dependency is relatively new, there is no well-established scale of smartphone dependency specific to the U.S. context. Although this study reported high reliability of this measure, future studies should develop and validate a smartphone dependency measure using U.S. samples. This research sheds lights on existing studies by investigating the directionality between smartphone engagement and psychological well-being (i.e., depressive symptoms and loneliness) among late adolescents. The results revealed that smartphone dependency predicted depressive symptoms and loneliness over time, and loneliness marginally predicted smartphone dependency over time. In addition, smartphone use was associated with later smartphone dependency yet smartphone use did not have a direct relationship with either depressive symptoms or loneliness. In other words, smartphone dependency is critical to understanding outcomes associated with smartphone use.
Limitations and future research A few limitations with the present study should be addressed. First, the sample used to test these relationships was a convenience sample composed entirely of college students. In addition, the sample was disproportionately female and white, and the response rate of this study was slightly low. Although there is little reason to believe that these findings are not generalizable to the larger population of late adolescents, future research should test these relationships using proportionate samples with late adolescents who are not attending college. Second, the interval of the two waves of data collection was between 2.5 and 3 months. Although this short longitudinal study sufficiently offered the preliminary results about the directionality between smartphone engagement and psychological well-being, future studies should use multiple-wave data over a longer period to test these relationships. Third, there was some attrition in the study as not all the potential participants completed both waves of the data collection. In particular, in an examination of participants who completed both waves of data collection versus those who only
Funding Sources This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. References [1] Pew Research Center. Mobile fact sheet. Available at: http://www. pewinternet.org/fact-sheet/mobile/. Accessed August 8, 2018. [2] Pew Research Center. Teens, social media & technology 2018. Washington, DC: Pew Research Center; 2018. [3] Twenge JM, Martin GN, Campbell WK. Decreases in psychological wellbeing among American adolescents after 2012 and links to screen time during the rise of smartphone technology. Emotion 2018;18:765e80. [4] Odgers C. Smartphones are bad for some teens, not all. Nature 2018;554: 432e4. [5] Bae SM. The relationship between smartphone use for communication, social capital, and subjective well-being in Korean adolescents: Verification using multiple latent growth modeling. Child Youth Serv Rev 2019;96:93e9. [6] George MJ, Russell MA, Piontak JR, et al. Concurrent and subsequent associations between daily digital technology use and high-risk adolescents’ mental health symptoms. Child Dev 2018;89:78e88.
6
M.A. Lapierre et al. / Journal of Adolescent Health xxx (2019) 1e6
[7] Van Deursen AJAM, Bolle CL, Hegner SM, et al. Modeling habitual and addictive smartphone behavior: The role of smartphone usage types, emotional intelligence, social stress, self-regulation, age, and gender. Comput Hum Behav 2015;45:411e20. [8] Kim D, Lee Y, Lee J, et al. Development of Korean smartphone addiction proneness scale for youth. PLoS One 2014;9:e97920. [9] Seo DG, Park Y, Kim MK, et al. Mobile phone dependency and its impacts on adolescents’ social and academic behaviors. Comput Hum Behav 2016;63: 282e92. [10] American Psychiatric Association. Diagnostic and statistical manual of mental disorders (DSM-5). 5th ed. Arlington, VA: American Psychiatric Association; 2013. [11] Billieux J. Problematic use of the mobile phone: A literature review and a pathways model. Curr Psychiatry Rev 2012:299e307. [12] Billieux J, Gay P, Rochat L, et al. The role of urgency and its underlying psychological mechanisms in problematic behaviours. Behav Res Ther 2010;48:1085e96. [13] Wolniewicz CA, Tiamiyu MF, Weeks JW, et al. Problematic smartphone use and relations with negative affect, fear of missing out, and fear of negative and positive evaluation. Psychiatry Res 2018;262:618e23. [14] Elhai JD, Tiamiyu M, Weeks J. Depression and social anxiety in relation to problematic smartphone use: The prominent role of rumination. Internet Res 2018;28:315e32. [15] Hawkley LC, Cacioppo JT. Loneliness matters: A theoretical and empirical review of consequences and mechanisms. Ann Behav Med 2010;40: 218e27. [16] Kandel DB, Raveis VH, Davies M. Suicidal ideation in adolescence: Depression, substance use, and other risk factors. J Youth Adolesc 1991;20: 289e309. [17] Aker S, S¸ahin MK, Sezgin S, et al. Psychosocial factors affecting smartphone addiction in university students. J Addict Nurs 2017;28:215e9. [18] Elhai JD, Levine JC, O’Brien KD, et al. Distress tolerance and mindfulness mediate relations between depression and anxiety sensitivity with problematic smartphone use. Comput Hum Behav 2018;84: 477e484. [19] Muusses LD, Finkenauer C, Kerkhof P, et al. A longitudinal study of the association between compulsive Internet use and wellbeing. Comput Hum Behav 2014;36:21e8. [20] Petersen AC, Sarigiani PA, Kennedy RE. Adolescent depression: Why more girls? J Youth Adolesc 1991;20:247e71. [21] Hays RD, DiMatteo MR. A short-form measure of loneliness. J Pers Assess 1987;51:69e87.
[22] Radloff LS. The CES-D scale: A self-report depression scale for research in the general population. Appl Psychol Meas 1977;1:385e401. [23] Kline RB. Principles and practice of structural equation modeling. 3rd ed. New York, NY: Guilford Press; 2010. [24] Iacobucci D, Schneider MJ, Popovich DL, et al. Mean centering helps alleviate “micro” but not “macro” multicollinearity. Behav Res Methods 2016; 48:1308e17. [25] Kraut R, Patterson M, Lundmark V, et al. Internet paradox: A social technology that reduces social involvement and psychological well-being? Am Psychol 1998;52:1017e31. [26] Cummings JN, Butler B, Kraut R. The quality of online social relationships. Commun ACM 2002;45:103e8. [27] Tokunaga RS. A unique problem or the manifestation of a preexisting disorder? The mediating role of problematic Internet use in the relationships between psychosocial problems and functional impairment. Communic Res 2014;41:531e60. [28] Choi S-W, Kim D-J, Choi J-S, et al. Comparison of risk and protective factors associated with smartphone addiction and internet addiction. J Behav Addict 2015;4:308e14. [29] Elhai JD, Dvorak RD, Levine JC, et al. Problematic smartphone use: A conceptual overview and systematic review of relations with anxiety and depression psychopathology. J Affect Disord 2017;207:251e9. [30] Pew Research Center. The smartphone difference. Washington, DC: Pew Research Center; 2015. [31] Lapierre MA. Smartphones and loneliness in love: Testing links between smartphone engagement, loneliness, and relational health. Psychol Pop Media Cult 2019. https://doi.org/10.1037/ppm0000230. [32] Bian M, Leung L. Smartphone addiction: Linking loneliness,shyness, symptoms and patterns of use to social capital. Soc Sci Comput Rev 2013; 41:159e76. [33] Jia Z, Wang Y, Yang Y, et al. Chinese university students’ loneliness and generalized pathological internet use: A longitudinal cross-lagged analysis. Soc Behav Personal Int J 2018;46:861e70. [34] Yao MZ, Zhong ZJ. Loneliness, social contacts and Internet addiction: A cross-lagged panel study. Comput Hum Behav 2014;30:164e70. [35] Cohen J. A power primer. Psychol Bull 1992;112:155e9. [36] Lapierre MA, Lewis MN. Should it stay or should it go now? Smartphones and relational health. Psychol Pop Media Cult 2018;7:384e95. [37] Greenberg JS. A study of stressors in the college student population. Health Educ 1981;12:8e12. [38] Aboujaoude E. Problematic internet use two decades later: Apps to wean us off apps. CNS Spectr 2019;24:371e3.
www.jahonline.org Original article
Short-Term Longitudinal Relationships Between Smartphone Use/Dependency and Psychological Well-Being Among Late Adolescents Matthew A. Lapierre, Ph.D. *, Pengfei Zhao, M.A., and Benjamin E. Custer, M.A. Department of Communication, University of Arizona, Tucson, Arizona
Article history: Received January 17, 2019; Accepted June 5, 2019 Keywords: Smartphone dependency; Smartphone use; Depressive symptoms; Loneliness; Late adolescents
A B S T R A C T
Purpose: The aim of the study was to determine the short-term longitudinal pathways between smartphone use, smartphone dependency, depressive symptoms, and loneliness among late adolescents. Methods: A two-wave longitudinal survey was used using adolescents between the ages of 17 and 20 years. The interval between wave 1 and wave 2 was between 2.5 and 3 months. Using convenience sampling, the total number of participants who completed both waves of data collection was 346. Validated measures assessed smartphone dependency, smartphone use, depressive symptoms, and loneliness. The longitudinal model was tested using path modeling techniques. Results: Among the 346 participants (33.6% male, mean [standard deviation] age at wave 1, 19.11 [.75] years, 56.9% response rate), longitudinal path models revealed that wave 1 smartphone dependency predicted loneliness (b ¼ .08, standard error [SE] ¼ .05, p ¼ .043) and depressive symptoms (b ¼ .11, SE ¼ .05, p ¼ .010) at wave 2, loneliness at wave 1 predicted depressive symptoms at wave 2 (b ¼ .21, SE ¼ .05, p < .001), and smartphone use at wave 1 predicted smartphone dependency at wave 2 (b ¼ .08, SE ¼ .05, p ¼ .011). Conclusions: Considering the rates of smartphone ownership/use among late adolescents (95%), the association between smartphone use and smartphone dependency, and the deleterious effects of loneliness and depression within this population, health practitioners should communicate with patients and parents about the links between smartphone engagement and psychological well-being. Ó 2019 Society for Adolescent Health and Medicine. All rights reserved.
In just over a decade, the smartphone has become a technological necessity. The Pew Research Center indicates that 77% of American adults own a smartphone, and that such ownership is nearly universal (95%) for adolescents [1,2]. Yet, with the growing
Conflicts of interest: The authors have no conflicts of interest to disclose. * Address correspondence to: Matthew A. Lapierre, Ph.D., Department of Communication, University of Arizona, 1103 E University Blvd, P.O.Box 210025, Tucson, AZ 85721. E-mail address: [email protected] (M.A. Lapierre). 1054-139X/Ó 2019 Society for Adolescent Health and Medicine. All rights reserved. https://doi.org/10.1016/j.jadohealth.2019.06.001
IMPLICATIONS AND CONTRIBUTION
This research sheds light on existing studies by investigating the directionality between smartphone engagement and psychological well-being among late adolescents. The findings reveal that smartphone engagement is a significant predictor of subsequent depressive symptoms and loneliness. This research can help practitioners understand the detrimental effects of smartphone engagement on adolescent well-being.
importance of smartphones in teens’ daily lives, there is some concern that these devices are interfering with their overall health and well-being [3]. Of particular concern is just how smartphones are affecting users. Researchers have reported mixed findings for the association between smartphone use and well-being [4]. Some studies have suggested that smartphone use may be positively associated with well-being [5]. For example, George et al. found that the number of text messages was negatively associated with depressive symptoms [6]. However, other studies have suggested
2
M.A. Lapierre et al. / Journal of Adolescent Health xxx (2019) 1e6
that smartphone use contributes to poor mental health [3]. In addition, recent research has demonstrated that smartphone use is a predictor of smartphone dependency, which, in turn, negatively affects mental health [7]. Smartphone dependency (also known as “problematic smartphone use” or “smartphone addiction”), which is characterized by excessive psychological attachment on smartphones with negative functional consequences [8], has been suggested to warrant considerable concerns among adolescent populations [9]. Although smartphone dependency is not included in DSM-5 [10] because of further examinations required, existing studies have suggested that smartphone dependency should be considered as a behavioral addiction functioning similarly with substance addictions such as alcohol and drugs [11] with shared characteristics including loss of control, tolerance, and withdrawal [12]. Smartphone dependency may be less of a concern than smartphone use, but smartphone dependency has been associated with a number of problematic outcomes including mental health issues [13,14]. Among these mental health issues, depression and loneliness are of paramount concern because they are associated with increased prevalence of chronic diseases, early morbidity/mortality, and suicide [15,16]. Previous cross-sectional research in this area has found a consistent link between depressive symptoms and loneliness with smartphone dependency [17]. However, there is no consensus on the directionality of these relationships, mostly because few longitudinal examinations have been conducted. One recent study that examined the longitudinal relationship between depression and smartphone dependency found that depression at wave 1 predicted dependency at wave 2 (mediated via mindfulness); however, the authors did not account for smartphone dependency at wave 1 [18]. Moreover, this finding is inconsistent with previous longitudinal studies on depression and general Internet dependency (which should function similarly to smartphone dependency) that have found that Internet dependency predicts both future depression and loneliness [19]. Also, the inconsistent findings raise the possibility that the relationship between smartphone dependency and psychological well-being is bidirectional. Namely, depressive symptoms and loneliness predict increased smartphone dependency over time, and increased smartphone dependency exacerbates depressive symptoms and loneliness over time. However, more longitudinal studies are needed to shed light on this issue. This set of findings is of particular importance because late adolescents (aged 18e20 years) are vulnerable to poor mental health, such as depression [20]; consequently, the focus of the present study is to determine the directionality between smartphone engagement (i.e., use, dependency) and psychological well-being (i.e., depressive symptoms, loneliness) among late adolescents. Previous studies have established that there is a clear relationship between smartphone dependency and reduced psychological well-being [17]. However, what is unknown is whether smartphone dependency fuels diminished well-being or if the relationship works in the other direction (i.e., depressive symptoms/loneliness breeds smartphone dependency). In addition, because of mixed findings, the association and directionality between smartphone use and well-being are also underdetermined. To that end, this study uses a longitudinal examination of late adolescents’ smartphone use, smartphone dependency, loneliness, and depressive symptoms to help answer these important questions.
Methods Recruitment College students enrolled in communication and sociology courses were recruited from three large public universities in the American Southwest, Midwest, and Southeast. The research was approved by the authors’ institutional review board (protocol# 18-002-COMM). The first wave of data collection occurred at the end of January and the beginning of February 2018. The second wave of data collection took place during the last 2 weeks of April 2018. To ensure the waves of data matched, participants were asked to create a unique and anonymous identifier. Sample The total of potential participants contacted for participation in this study was 816. In the first wave of the study, 585 participants (72%) successfully completed the survey; at the second wave, 590 participants completed the survey. However, there were participants who completed the survey at wave 1 who did not take it at wave 2 (and vice versa), and the total number of participants who completed both waves of data collection was 464; namely, 79% of participants who completed the survey at wave 1 also completed the follow-up. To keep the focus of this study on late adolescents, all participants aged 21 years were removed from the sample resulting in 346 total respondents. The sample was 33.6% male (n ¼ 116), and the mean age of participants in the sample at wave 1 was 19.11 years (standard deviation [SD] ¼ .75). Participants in the study racially identified as white (n ¼ 230, 66.5%), African American (n ¼ 14, 4.0%), Asian/ Pacific Islander (n ¼ 51, 14.7%), Latino/a (n ¼ 25, 7.2%), Native American (n ¼ 1, .34%), “other” (n ¼ 2, .6%), and multiracial (n ¼ 23, 6.6%). Among the universities that participated, 50.3% (n ¼ 174) came from the university in the Midwest region, 39% (n ¼ 135) from the southwestern university, and 10.7% (n ¼ 37) from the southeastern university. Measures Smartphone dependency. Smartphone dependency was measured using the smartphone addiction proneness scale by Kim et al. [8]. This is a 15-item scale where participants are asked to indicate their level of agreement with statements regarding their smartphone engagement (e.g., “I panic when I cannot use my smartphone”). All questions were asked on a 4-point scale (0 ¼ strongly disagree to 3 ¼ strongly agree; W1: mean ¼ 1.21, SD ¼ .45, a ¼ .88; W2: mean ¼ 1.19, SD ¼ .47, a ¼ .89). Smartphone use. Participants were asked to estimate their daily smartphone use in minutes across eight common uses for smartphones. These were talking on the phone, texting, using e-mail, surfing the Web, using social networks (e.g., Instagram and Facebook), using FaceTime or other video chat, playing games, and/or using news apps. Participants’ use across these eight categories were summed to create a measure of smartphone use in minutes per day. Because of significant positive skew, the measure was transformed using a log transformation (W1: mean ¼ 2.54, SD ¼ .29; W2: mean ¼ 2.52, SD ¼ .28). Loneliness. The measure of loneliness for this study was the UCLA Loneliness scale [21]. This is an 8-item measure that asks
M.A. Lapierre et al. / Journal of Adolescent Health xxx (2019) 1e6
participants to indicate how often certain statements apply to them, such as “I feel left out” and “I lack companionship.” Each question was measured on a 4-point scale (0 ¼ never to 3 ¼ always) with scores averaged for participants (W1: mean ¼ 1.00, SD ¼ .61, a ¼ .83; W2: .98, SD ¼ .63, a ¼ .85). Depressive symptoms. The 10-item Center for Epidemiologic Studies Depression Scale was used as the measure for depressive symptoms [22]. The scale asks participants to indicate how many days in the last week they felt a particular way. For example, participants are asked if they were “bothered by things that usually do not bother me.” All questions were answered using a 4-point scale (0 ¼ rarely or none of the time [less than one] to 3 ¼ most of all of the time [5e7]) and were summed to create the measure (W1: mean ¼ 9.58, SD ¼ 5.07, a ¼ .80; W2: mean ¼ 9.71, SD ¼ 5.52, a ¼ .83). Analysis strategy Table 1 shows all zero-order correlations for the variables of interest. The longitudinal model was tested using path modeling techniques in AMOS 23. The initial test used a saturated path model paths not significant at the one-tailed p .10 subsequently removed [23]. To evaluate the appropriateness of the path model, chi-square, the root mean squared error of approximation (RMSEA), the comparative fit index, and the normed fit index were used to assess the model fit. For the path model, the standards used to determine fit were RMSEA at .05, a non-significant c2, and goodness-of-fit indices above .90 [23]. Finally, to guard against multicollinearity in the model, both measures of loneliness and depressive symptoms were centered because of their moderately large correlation [24]. Results The initial saturated model revealed that the following paths were not significant: wave 1 depressive symptoms to wave 2 smartphone dependency (b ¼ 0, SE ¼ .04, p > .99), wave 1 depressive symptoms to wave 2 smartphone use (b ¼ .01, SE ¼ .03, p ¼ .92), wave 1 smartphone dependency to wave 2 smartphone use (b ¼ .02, SE ¼ .02, p ¼ .65), wave 1 loneliness to wave 2 smartphone use (b ¼ .04, SE ¼ .02, p ¼ .40), wave 1 smartphone use to wave 2 loneliness (b ¼ .01, SE ¼ .08, p ¼ .87), and wave 1 smartphone use to wave 2 depressive symptoms (b ¼ .01, SE ¼ .08, Table 1 Zero order correlations for variables of interest Variable
2
3
4
.139** 1. W1: smartphone .215*** .012 use 2. W1: smartphone .241*** .297*** dependency 3. W1: loneliness .628*** 4. W1: depressive symptoms 5. W2: smartphone use 6. W2: smartphone dependency 7. W2: loneliness 8. W2: depressive Symptoms N ¼ 346; W1 ¼ Wave 1; W2 ¼ Wave 2. * p < .05; **p < .01; ***p < .001.
5
6
7
.722*** .243*** .017
8 .090
.183**
.786*** .250*** .295***
.039 .136*
.250*** .743*** .524*** .278*** .500*** .631*** .268*** .030
.107*
.356*** .409*** .666***
3
p ¼ .89). In addition, the wave 1 association between loneliness and smartphone use was not significant (b ¼ .01, SE ¼ .01, p ¼ .82), and the error term associations between wave 2 smartphone use to both depressive symptoms (b ¼ .02, SE ¼ .004, p ¼ .68) and loneliness at wave 2 (b ¼ .04, SE ¼ .004, p ¼ .43) were not significant. The adjusted model with nonsignificant paths removed demonstrated good model fit c2(10) ¼ 4.24, p ¼ .94, as both goodness-of-fit indices were above recommended thresholds (comparative fit index ¼ 1.000; normed fit index ¼ .997) and the RMSEA < .001. There was one path that was not significant as the link from wave 1 loneliness to wave 2 smartphone dependency was only marginally significant (b ¼ .06, SE ¼ .03, p ¼ .060; Figure 1 for final model). Of particular note, the following paths were significant: smartphone dependency at wave 1 predicted loneliness at wave 2 (b ¼ .08, SE ¼ .05, p ¼ .043), smartphone dependency at wave 1 predicted depressive symptoms at wave 2 (b ¼ .11, SE ¼ .05, p ¼ .010), loneliness at wave 1 was associated with depressive symptoms at wave 2 (b ¼ .21, SE ¼ .05, p < .001), and smartphone use at wave 1 predicted smartphone dependency at wave 2 (b ¼ .08, SE ¼ .05, p ¼ .011). Discussion The aim of this study was to determine the longitudinal pathways between smartphone use, smartphone dependency, depressive symptoms, and loneliness among late adolescents. Depression and loneliness represent serious impediments to adolescent health [15,16] with evidence that depression becomes more severe as teens age [20]. This is particularly concerning when considering the increased prevalence and use of smartphones within this population [1], as a number of cross-sectional studies have found that smartphone use/dependency may be linked to loneliness and depressive symptoms [17]; however, the directionality of these relationships was still undetermined. Our longitudinal results showed that smartphone dependency predicted increased depressive symptoms 3 months later for late adolescents. This finding is inconsistent with the previous longitudinal study on smartphone dependency and depression, which demonstrated that depression leads to later smartphone dependency via mindfulness [18]. But, it is important to note that the authors of this previous study did not account for dependency at wave 1. However, our findings do align with previous longitudinal studies on depression and general Internet dependency. For example, Muusses et al. conducted a five-wave longitudinal study and found that compulsive Internet use (i.e., being incapable of regulating one’s Internet use) predicted later depression [19]. Previous studies provide several reasons to explain this relationship. One is based on the displacement hypothesis, which states that time spent on the Internet displaces the time spent on offline communication with close friends and family members [25]. Compared with offline communication, online interaction may sometimes engender less social support and intimacy [26], leading to feelings of social isolation and detachment. Another reason might be that Internet dependency causes vocational impairment, such as loss of productivity in academic or occupational performance [27], which, in turn, makes one experience negative emotions. Smartphone dependency may predict depressive symptoms via these same underlying mechanisms such as lack of social support from strong ties or vocational impairment, which should be investigated in future studies. Nevertheless, the findings from
4
M.A. Lapierre et al. / Journal of Adolescent Health xxx (2019) 1e6
Figure 1. Final longitudinal model.
the present study are concerning. Specifically, although Internet dependency certainly shares some characteristics with smartphone dependency, Internet dependency is associated with different constructs than smartphone dependency [28], and there are qualitative differences regarding how smartphones are involved in our daily lives. Internet dependency refers to problematic use of the Internet chats, games, and information seeking and has typically been associated with general computer use [29]. Smartphones offer similar affordances but extend the potential risk via their portability, wireless data access, and wide range of applications that make them “indispensable” devices [30]. As such, the potential risks they pose to the mental health of young people might be amplified, yet the possible amplified effect requires further examination. Our results further demonstrated that smartphone dependency significantly predicted later loneliness among participants. Previous research has shown that loneliness and smartphone dependency are linked [31,32]; however, this is the first known study to show that smartphone dependency precedes loneliness. There is some overlap with previous research on Internet dependency and loneliness. For example, Jia et al. found that Internet dependency did predict later loneliness among Chinese university students [33]. Yet, in that study, they also found that loneliness was predictive of later Internet dependency. The present study only found a marginally significant link between wave 1 loneliness and wave 2 smartphone dependency. Our findings suggest that smartphones are not a favorable surrogate compared with offline interaction in terms of providing social support and intimacy [26]; consequently, reliance on smartphones potentially removes people from
meaningful interpersonal connections, which may engender feelings of increased loneliness [34]. It is also important to discuss the magnitude of these effects for smartphone dependency on loneliness/depressive symptoms and their practical importance for the health and well-being of adolescents. As seen in Figure 1, the magnitude of these effects would traditionally be considered small [35]. However, there are two things to note regarding the size of these effects. First, the effects between smartphone dependency and depressive symptoms are larger than what has been previously found in research looking at Internet dependency and this same variable (b ¼ .11 vs. b ¼ .06) [19]. Second, because we controlled for both loneliness and depressive symptoms at wave 1 in the analyses, the amount of unaccounted for variance at wave 2 for these variables is much smaller; this happens because of the autocorrelations between waves and the fact that loneliness and depressive symptoms are highly correlated with one another (r ¼ .63). As such, the amount of residual variance left for smartphone dependency to predict either of these variables is quite small. However, this amplification effect for dependency on these variables is important, particularly considering the tremendous negative outcomes associated with both loneliness and depressive symptoms [15,16] as any increase in either of these variables can lead to serious problems in adolescents. The role that smartphone use played in the longitudinal model was interesting but not unexpected, as previous crosssectional research has shown that smartphone use and smartphone dependency are linked [7,36]. In the present study, smartphone use at wave 1 was associated with dependency at wave 2 but was not associated with depressive symptoms or
M.A. Lapierre et al. / Journal of Adolescent Health xxx (2019) 1e6
loneliness in the model in either direction. This finding stands in contrast to recent theorizing about smartphone use’s influence on psychological well-being, which has suggested that use, on its own, is associated with such outcomes [3,5]. In other words, although smartphone ownership is nearly universal for adolescents, smartphone use itself may not be detrimental to mental health. Rather, only those who develop smartphone dependency from smartphone use may experience negative outcomes with respect to psychological well-being. Existing research has suggested that smartphone use may lead to smartphone dependency via habitual smartphone use [7]. Namely, the multifunction, accessibility, and automatic checking of smartphone might develop smartphone use to be a habit, which contributes to later smartphone dependency. Therefore, these results suggest that if researchers wish to understand the link between smartphone use and well-being, it is critical to explore the role played by smartphone dependency. However, the present study focused on negative constructs (i.e., depressive symptoms and loneliness), so future research should explore the directionality of the associations between smartphone engagement and positive constructs (e.g., social support and social capital). Implications How might practitioners use the findings from the present study to improve adolescent health? First, it is necessary to acknowledge that late adolescents are vulnerable to becoming depressed and lonely, particularly when many of these young people are making significant life changes such as joining the workforce or moving away to college [37]. Therefore, understanding the directionality of the relationship between smartphone engagement and depressive symptoms and loneliness may shed light on late adolescents’ mental health maintenance. Pediatricians and other practitioners should consult with teen patients about their smartphone use and help them understand the links between use, dependency, and psychological wellbeing. Practitioners could also recommend that their patients use smartphone applications that help to limit the time spent with these devices (e.g., Moment and Flipd) [38].
5
completed the first wave, those who dropped out after wave 1 scored lower on the depressive symptoms measure than participants who finished both data collections (mean ¼ .83 vs. mean ¼ .96; Hedges’ g ¼ .24, 95% confidence interval [.50, .02]). However, it is unclear how this would challenge the overall findings because the present study was looking at within-person change on this measure over time. Fourth, this study did not use clinical thresholds as an indicator of depression, and the means of some constructs such as depressive symptoms and loneliness were low. However, this study did find that the variation of smartphone dependency can contribute to the variations of depressive symptoms and loneliness. Namely, smartphone dependency can be a predictor for subsequent depressive symptoms and loneliness among older adolescents. The final limitations with the study were the nature of the smartphone use measure and the validity of the smartphone dependency measure. First, the present study used a self-report measure rather than direct observations to assess the extent to which participants used their smartphone. Future research should use direct observations of smartphone use (i.e., using smartphone applications such as Moment) to further validate the findings of the present study. Second, the smartphone dependency measurement this study used was developed and validated on South Korean samples. To our knowledge, because the phenomenon of smartphone dependency is relatively new, there is no well-established scale of smartphone dependency specific to the U.S. context. Although this study reported high reliability of this measure, future studies should develop and validate a smartphone dependency measure using U.S. samples. This research sheds lights on existing studies by investigating the directionality between smartphone engagement and psychological well-being (i.e., depressive symptoms and loneliness) among late adolescents. The results revealed that smartphone dependency predicted depressive symptoms and loneliness over time, and loneliness marginally predicted smartphone dependency over time. In addition, smartphone use was associated with later smartphone dependency yet smartphone use did not have a direct relationship with either depressive symptoms or loneliness. In other words, smartphone dependency is critical to understanding outcomes associated with smartphone use.
Limitations and future research A few limitations with the present study should be addressed. First, the sample used to test these relationships was a convenience sample composed entirely of college students. In addition, the sample was disproportionately female and white, and the response rate of this study was slightly low. Although there is little reason to believe that these findings are not generalizable to the larger population of late adolescents, future research should test these relationships using proportionate samples with late adolescents who are not attending college. Second, the interval of the two waves of data collection was between 2.5 and 3 months. Although this short longitudinal study sufficiently offered the preliminary results about the directionality between smartphone engagement and psychological well-being, future studies should use multiple-wave data over a longer period to test these relationships. Third, there was some attrition in the study as not all the potential participants completed both waves of the data collection. In particular, in an examination of participants who completed both waves of data collection versus those who only
Funding Sources This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. References [1] Pew Research Center. Mobile fact sheet. Available at: http://www. pewinternet.org/fact-sheet/mobile/. Accessed August 8, 2018. [2] Pew Research Center. Teens, social media & technology 2018. Washington, DC: Pew Research Center; 2018. [3] Twenge JM, Martin GN, Campbell WK. Decreases in psychological wellbeing among American adolescents after 2012 and links to screen time during the rise of smartphone technology. Emotion 2018;18:765e80. [4] Odgers C. Smartphones are bad for some teens, not all. Nature 2018;554: 432e4. [5] Bae SM. The relationship between smartphone use for communication, social capital, and subjective well-being in Korean adolescents: Verification using multiple latent growth modeling. Child Youth Serv Rev 2019;96:93e9. [6] George MJ, Russell MA, Piontak JR, et al. Concurrent and subsequent associations between daily digital technology use and high-risk adolescents’ mental health symptoms. Child Dev 2018;89:78e88.
6
M.A. Lapierre et al. / Journal of Adolescent Health xxx (2019) 1e6
[7] Van Deursen AJAM, Bolle CL, Hegner SM, et al. Modeling habitual and addictive smartphone behavior: The role of smartphone usage types, emotional intelligence, social stress, self-regulation, age, and gender. Comput Hum Behav 2015;45:411e20. [8] Kim D, Lee Y, Lee J, et al. Development of Korean smartphone addiction proneness scale for youth. PLoS One 2014;9:e97920. [9] Seo DG, Park Y, Kim MK, et al. Mobile phone dependency and its impacts on adolescents’ social and academic behaviors. Comput Hum Behav 2016;63: 282e92. [10] American Psychiatric Association. Diagnostic and statistical manual of mental disorders (DSM-5). 5th ed. Arlington, VA: American Psychiatric Association; 2013. [11] Billieux J. Problematic use of the mobile phone: A literature review and a pathways model. Curr Psychiatry Rev 2012:299e307. [12] Billieux J, Gay P, Rochat L, et al. The role of urgency and its underlying psychological mechanisms in problematic behaviours. Behav Res Ther 2010;48:1085e96. [13] Wolniewicz CA, Tiamiyu MF, Weeks JW, et al. Problematic smartphone use and relations with negative affect, fear of missing out, and fear of negative and positive evaluation. Psychiatry Res 2018;262:618e23. [14] Elhai JD, Tiamiyu M, Weeks J. Depression and social anxiety in relation to problematic smartphone use: The prominent role of rumination. Internet Res 2018;28:315e32. [15] Hawkley LC, Cacioppo JT. Loneliness matters: A theoretical and empirical review of consequences and mechanisms. Ann Behav Med 2010;40: 218e27. [16] Kandel DB, Raveis VH, Davies M. Suicidal ideation in adolescence: Depression, substance use, and other risk factors. J Youth Adolesc 1991;20: 289e309. [17] Aker S, S¸ahin MK, Sezgin S, et al. Psychosocial factors affecting smartphone addiction in university students. J Addict Nurs 2017;28:215e9. [18] Elhai JD, Levine JC, O’Brien KD, et al. Distress tolerance and mindfulness mediate relations between depression and anxiety sensitivity with problematic smartphone use. Comput Hum Behav 2018;84: 477e484. [19] Muusses LD, Finkenauer C, Kerkhof P, et al. A longitudinal study of the association between compulsive Internet use and wellbeing. Comput Hum Behav 2014;36:21e8. [20] Petersen AC, Sarigiani PA, Kennedy RE. Adolescent depression: Why more girls? J Youth Adolesc 1991;20:247e71. [21] Hays RD, DiMatteo MR. A short-form measure of loneliness. J Pers Assess 1987;51:69e87.
[22] Radloff LS. The CES-D scale: A self-report depression scale for research in the general population. Appl Psychol Meas 1977;1:385e401. [23] Kline RB. Principles and practice of structural equation modeling. 3rd ed. New York, NY: Guilford Press; 2010. [24] Iacobucci D, Schneider MJ, Popovich DL, et al. Mean centering helps alleviate “micro” but not “macro” multicollinearity. Behav Res Methods 2016; 48:1308e17. [25] Kraut R, Patterson M, Lundmark V, et al. Internet paradox: A social technology that reduces social involvement and psychological well-being? Am Psychol 1998;52:1017e31. [26] Cummings JN, Butler B, Kraut R. The quality of online social relationships. Commun ACM 2002;45:103e8. [27] Tokunaga RS. A unique problem or the manifestation of a preexisting disorder? The mediating role of problematic Internet use in the relationships between psychosocial problems and functional impairment. Communic Res 2014;41:531e60. [28] Choi S-W, Kim D-J, Choi J-S, et al. Comparison of risk and protective factors associated with smartphone addiction and internet addiction. J Behav Addict 2015;4:308e14. [29] Elhai JD, Dvorak RD, Levine JC, et al. Problematic smartphone use: A conceptual overview and systematic review of relations with anxiety and depression psychopathology. J Affect Disord 2017;207:251e9. [30] Pew Research Center. The smartphone difference. Washington, DC: Pew Research Center; 2015. [31] Lapierre MA. Smartphones and loneliness in love: Testing links between smartphone engagement, loneliness, and relational health. Psychol Pop Media Cult 2019. https://doi.org/10.1037/ppm0000230. [32] Bian M, Leung L. Smartphone addiction: Linking loneliness,shyness, symptoms and patterns of use to social capital. Soc Sci Comput Rev 2013; 41:159e76. [33] Jia Z, Wang Y, Yang Y, et al. Chinese university students’ loneliness and generalized pathological internet use: A longitudinal cross-lagged analysis. Soc Behav Personal Int J 2018;46:861e70. [34] Yao MZ, Zhong ZJ. Loneliness, social contacts and Internet addiction: A cross-lagged panel study. Comput Hum Behav 2014;30:164e70. [35] Cohen J. A power primer. Psychol Bull 1992;112:155e9. [36] Lapierre MA, Lewis MN. Should it stay or should it go now? Smartphones and relational health. Psychol Pop Media Cult 2018;7:384e95. [37] Greenberg JS. A study of stressors in the college student population. Health Educ 1981;12:8e12. [38] Aboujaoude E. Problematic internet use two decades later: Apps to wean us off apps. CNS Spectr 2019;24:371e3.