Nomophobia, anxiety, and social comparison orientation: Associations with physical health symptoms among college students

Abstract

Objectives:

Nomophobia is characterized by distress about being without one’s mobile phone and is associated with negative health outcomes, particularly for college students. However, the unique contributions of nomophobia versus global anxiety and the role of social comparison orientation (SCO) is not clear.

Participants:

310 college students (M_Age = 19.43; 66% women; 34% racial/ethnic minority identity).

Methods:

Cross-sectional survey using validated measures of nomophobia, anxiety, SCO, and physical health symptoms.

Results:

Including nomophobia did not improve model fit relative to anxiety alone for predicting symptoms (ΔR² = 0.003, p = 0.32). There was no interaction between nomophobia and SCO (R² = 0.01, p = 0.12), though the interaction between anxiety and SCO was significant (R² = 0.03, p = 0.01).

Conclusions:

Global anxiety may be more useful for identifying students at risk for physical health symptoms than nomophobia. Future work elucidating whether nomophobia is an appropriate treatment target for managing anxiety-related and/or physical health symptoms could help improve the health of college students.

Introduction

The prevalence of mobile phone use continues to increase: as of 2024, 91% of individuals in the United States report owning a smartphone, with a rate of 98% among adults ages 18–29.¹ Although smartphones offer many helpful functions, including facilitating social connections,² some users develop problematic patterns of use, and these patterns are more common among college students than other groups.¹ For example, nomophobia (NO MObile PHOne phoBIA) is characterized by fear or distress resulting from being without one’s mobile phone or being unable to use one’s phone to connect to others (e.g., via text message or connection to the internet).^3,4 Nomophobia is prominent among college students, affecting up to 97% worldwide.⁵ Nomophobia may be especially common among women, though existing evidence is inconclusive.³ The prevalence of nomophobia is particularly alarming as it has been associated with smartphone addiction, anxiety symptoms, and insomnia.^6,7 These experiences often result in negative outcomes in the domains of relationships, self-esteem, stress, academic performance, and anxiety.^8,9

Nomophobia has been described as one of several potential anxiety disorders, which are broadly characterized as excessive worry that is difficult to control, causing restlessness, irritability, and sleep disturbances that lead to significant distress and/or functional impairment.¹⁰ Worry, distress, and negative physical health outcomes are implicated in many of these disorders, particularly Generalized Anxiety Disorder (GAD), though the target of one’s worry and distress is narrower for nomophobia. An underlying contributor to anxiety disorders is intolerance of uncertainty, or difficulty coping with uncertain situations and their implications (cf.¹¹) this tendency is widely understood to be a transdiagnostic risk factor for the development and maintenance of clinically significant anxiety and related disorders.^12,13 Nomophobia may be a specific manifestation of uncertainty intolerance, whereby an individual’s mobile phone represents a means to obtain (more) certain information about oneself, others, and the world; mobile phone use is then negatively reinforced through reductions in anxiety with greater (perceived) certainty.¹⁴ As illustrated in Figure 1, the inability to use a phone for this purpose removes a method to alleviate distress and produces symptoms of anxiety. Despite considerable overlap in their definitions and the relevance of the intolerance of uncertainty framework, the extent to which nomophobia adds to our understanding of more global anxiety, worry, and their correlates is not clear. Specifically, there is insufficient evidence to determine whether nomophobia is a distinct syndrome or a subtype of more general anxiety (or of GAD). An improved understanding of the unique contributions and impacts of nomophobia would indicate the utility of targeted assessment and treatment approaches beyond those already existing for global anxiety.

Figure 1.

Proposed model of intolerance of uncertainty, anxiety, and nomophobia.

Anxiety, nomophobia, and physical health symptoms

Extant evidence indicates that anxiety is associated with physical health symptoms,¹⁵ though few studies have examined the unique association between nomophobia and physical health symptoms³ – particularly among high-prevalence groups such as college students. An increasing number of undergraduate students in the United States reported poor physical health,¹⁶ and anxiety and underlying intolerance of uncertainty are known risk factors. For instance, intolerance of uncertainty is positively associated with somatic symptoms and negatively associated with quality of life in this population.^17,18 Nomophobia may also increase risk for physical health symptoms and represent a unique treatment target for improving student health outcomes, though less is known about the link between nomophobia and physical health among college students.

Many students who experience nomophobia endorse physical symptoms associated with heightened anxiety, such as rapid heartbeat, excessive sweating, tremors, and shaking due to heightened arousal.¹⁹ However, investigation beyond anxiety-related somatic symptoms is scarce, and most research focuses on associations between physical health symptoms and broader technology-related behaviors such as excessive smartphone use.³ Few studies have examined the association between nomophobia and a broader range of physical health symptoms; those that have note straining and watering of the eye, fatigue, headaches, trouble sleeping, and musculoskeletal problems such as joint pain, “text neck syndrome,” and “Short Message Service (SMS) thumb”.³ Additional knowledge about these associations beyond the contribution of global anxiety could help to determine whether nomophobia is indeed a distinct condition, and whether students who endorse this experience are at uniquely heightened risk for physical health problems²⁰ and consequent disruptions to their academic, social, and occupational functioning.⁵

Intolerance of uncertainty in relative self-evaluation: Nomophobia and social comparison orientation

Up to 93% of U.S. adults ages 18–29 report using at least one social media site or app.²¹ These platforms provide opportunities for social comparison (i.e., evaluating oneself in relation to another person²²) by exposing users to information about other people. On social media and elsewhere, users can engage in upward comparisons (where the target is considered better off than the self) and downward comparisons (where the target is considered worse off than the self^23,24). Among individuals who have difficulty tolerating uncertainty, exposure to information about others and the resulting self-evaluations could reduce uncertainty and thus, anxiety and distress. Consistent with this idea, individuals who endorse higher intolerance of uncertainty also report more frequent upward comparisons,²⁵ though this association appears to be cyclical over the long-term.²⁶ Further, nomophobia has been explicitly linked to the use of social media platforms²⁷ and their availability, such that outages in social media access are associated with increased symptoms of nomophobia.²⁸

It is possible that a broader tendency to compare with others contributes to or exacerbates the consequences of nomophobia, as information about others is readily accessible on social media platforms that are often accessed via mobile phones.^29,30 Social comparison orientation (SCO; the general tendency to compare oneself to others) is positively associated with social anxiety, social phobia, social media addiction, and problematic smartphone use, including among college students,^31–34 and is associated with poor physical health outcomes in a wide range of groups (e.g., sleep disruption^35–38). At present, however, the role of SCO in the association between nomophobia and physical health symptoms (beyond more general anxiety) is not clear.

Aims of the present study

Physical health symptoms have been associated with problematic smartphone use, though there has been little investigation of unique associations with nomophobia or the role of social comparison orientation (SCO). Establishing these connections is crucial for understanding how nomophobia, global anxiety, and social comparison may contribute to negative health outcomes among college students, and can clarify whether nomophobia is a specific (i.e., unique) and appropriate intervention target for this population. Toward these goals, the primary aim of the present study was to determine the association between nomophobia (over and above the contribution of more global anxiety) and physical health symptoms among college students. A secondary aim was to determine associations between nomophobia and SCO (independent of anxiety) and whether SCO moderates the relation between nomophobia and physical health symptoms.

Methods

Participants and procedures

Students enrolled in a psychology course at a large, public university in the northeastern U.S. were eligible to participate in a large-scale survey study in 2022–2023, if they endorsed: 1) being at least 18 years of age, 2) using digital technology daily (e.g., smartphones, social media, etc.), and 3) fluency in English. Students were recruited via the SONA online scheduling system; on this platform, they are able to view and self-select to enroll in one or more active studies in exchange for course credit. Of the 310 participants who enrolled in this study and completed the relevant measures, the majority identified as women (65.81%), and 33.66% identified with racial/ethnic minority backgrounds; 55.2% were first-year students and 58.0% lived on campus with roommates (see Table 1 for additional participant demographics). The measures below were selected for their strong psychometric properties for assessing the constructs of interest via self-report. Procedures were approved by the relevant Institutional Review Board and all participants provided electronic documentation of consent. Participants completed the survey electronically via Qualtrics and received course credit for their participation. To minimize social desirability bias, students were informed that their survey responses would be deidentified (identifiable information would not be included in data analyses and dissemination), and that their course instructor would not have access to their responses.

Table 1.

Participant demographics (N = 310).

	M (SD)
Age	19.43 (2.11)
	n (%)
Gender
Woman	204 (65.81%)
Man	99 (31.94%)
Non-binary	5 (1.61%)
Transgender	2 (0.65%)
Year*
Freshman	170 (55.19%)
Sophomore	59 (19.16%)
Junior	43 (13.96%)
Senior	34 (11.04%)
Graduate student	2 (0.65%)
Ethnic identity*
Hispanic/Latinx	46 (14.94%)
Non-Hispanic/Latinx	262 (85.06%)
Racial Identity**
Black or African American	41 (13.40%)
East Asian	11 (3.59%)
Native American or Alaska Native	1 (0.33%)
Native Hawaiian or Other Pacific Islander	4 (1.31%)
White	203 (66.34%)
Multiracial	19 (6.21%)
Other	27 (8.82%)
Living/housing
On campus with roommate(s)	180 (58.06%)
On campus single	15 (4.84%)
Off campus with parent(s)	79 (25.48%)
Off campus with spouse/significant other	7 (2.26%)
Off campus with roommate(s)	25 (8.06%)
Off campus single	4 (1.29%)

Note:

^*n = 308 (2 did not answer);

^**n = 306 (4 did not answer).

Measures

Nomophobia Questionnaire (NMP-Q) is a 20-item measure that assesses the severity of nomophobia.⁴ Items start with the base prompt of “If I did not have my smartphone with me,” followed by an additional clause to complete the statement (e.g., “I would feel uncomfortable without constant access to information through my smartphone”). Items are rated on a 7-point scale from “strongly disagree” (1) to “strongly agree” (7). Scoring involves summing item ratings and interpretation follows total score thresholds: ≤20 indicates the absence of nomophobia, scores 21–59 (inclusive) indicate mild nomophobia, scores 60–99 (inclusive) indicate moderate nomophobia, and scores 100–140 indicate severe nomophobia.⁴ A recent systematic review indicated that the NMP-Q has an excellent level of internal consistency (ɑ = 0.92) and no significant age or gender differences.³⁹

Emotional Distress-Anxiety – Short Form 6a (PROMIS). To assess anxiety symptoms, respondents are asked to complete six items using a 5-point scale (1 = Never; 2 = Rarely, 3 = Sometimes, 4 = Often, 5 = Always).⁴⁰ All items ask about anxiety symptoms in the last 7 days; these include fear, difficulty concentrating, nervousness, uneasiness, and a sense of “needing help for my anxiety.” We selected this measure of anxiety, rather than alternative options such as the Beck Anxiety Inventory, because it does not include physical symptoms associated with anxiety (e.g., tension, headaches) and thus, would not assess the same symptoms as the CHIPS. In accordance with scoring guidelines, raw scores were converted to t-scores and then were converted to final scores.^40,41 Previous work demonstrated validity equivalent to other established measures of anxiety in clinical samples (e.g., GAD-7).⁴²

Cohen-Hoberman Inventory of Physical Symptoms (CHIPS) uses 33 items to assess the presence and intensity of distressing physical symptoms over the last 2 weeks.^43,44 Items start with “How much were you bothered by:” followed by physical symptoms (e.g., back pain, faintness, and blurred vision). Respondents rate each symptom on a scale from 0 (“not been bothered by the problem”) to 4 (“problem has been an extreme bother”). Summary scores use sums of ratings across the 33 items, with higher scores indicating more severe symptoms. Reliability and validity were established using samples of college students,⁴⁵ and previous work has shown strong internal validity of the scale (ɑ = 0.92).⁴³

Iowa Netherlands Comparison Orientation Measure (INCOM) is a 23-item measure of social comparison orientation (SCO).⁴⁶ It uses 11 core non-directional comparison items, six upward comparison items, and six downward comparison items. Each item uses a five-point scale from “I disagree strongly” (1) to “I agree strongly” (5). Each subscale score is calculated by summing across relevant items and higher scores indicate stronger SCO. This measure has demonstrated strong psychometric properties in initial and subsequent validation studies, including among college students (ɑ = 0.82).^46,47

Data analysis

Analyses were conducted using RStudio Version 4.4.1. We ran descriptive statistics to determine participant characteristics, particularly with respect to levels of the predictors and outcome of interest (i.e., nomophobia, anxiety, SCO, and physical health symptoms). We tested for bivariate correlations between scores on measures of these constructs and tested gender differences in these constructs to determine whether to control for gender as a covariate. We used a general linear model to determine whether adding nomophobia improved model fit for predicting physical health symptoms, beyond the contribution of anxiety. The reduced model was specified as: physical health symptoms = anxiety + gender. The full model was specified as: physical health symptoms = anxiety + gender + nomophobia. Finally, we added an interaction to our model to determine whether SCO moderated the association between nomophobia and physical symptoms, controlling for anxiety (model: physical health symptoms = anxiety + gender + nomophobia + SCO + anxiety*SCO + nomophobia*SCO).

A priori, the full models were considered to outperform the reduced models if the ΔR² (change in R²) > 0.02.⁴⁸ We also report additional model fit statistics (i.e., AIC, BIC, and Bayes Factor) and p-values, for reference. We selected this approach to maximize the utility of our sample of 310 students; this sample size affords adequate power to detect main effects in general linear models (at p < 0.05), though power for interaction effects is limited. Effect sizes such as ΔR² and model fit statistics are not as dependent on sample size as p-values and are informative in moderate sample sizes.⁴⁹

Results

Only 0.65% of participants did not experience nomophobia symptoms (NMP_Q score ≤20). In contrast, 13% reported mild symptoms, 56% reported moderate symptoms, and 30.97% reported severe symptoms. Thus, the majority of students in the present sample reported moderate to severe symptoms of nomophobia. With respect to global anxiety, 25% of participants reported no to slight anxiety, 26% reported mild anxiety, 34% reported moderate anxiety, and 15% reported severe anxiety. These experiences co-occurred in a subset of students: 47% endorsed moderate-to-severe levels of both nomophobia and global anxiety, though 39% reported moderate-to-severe nomophobia but none-to-mild anxiety and 2% indicated moderate-to-severe anxiety but none-to-mild nomophobia.

The average total physical symptoms (CHIPS) score was 23.15 (SD = 19.20) and ranged from 0 to 109. The average INCOM (social comparison orientation; SCO) score was 36.77 (SD = 7.22) and ranged from 15 to 55. As shown in Table 2, we observed positive correlations between anxiety and nomophobia (r = 0.42), between nomophobia and physical health symptoms (r = 0.24), and among anxiety, nomophobia, physical health symptoms, and SCO (rs = 0.23–0.39, ps < 0.001). In addition, there were gender differences in nomophobia, physical health symptoms, and global anxiety, such that women reported more symptoms than men in all three domains (ps < 0.001), and a model controlling for gender outperformed the full model with the variables of interest only (e.g., physical health symptoms, anxiety, and nomophobia; F[1,259] = 10.46, ΔR² = 0.03, p < 0.01). As such, models outlined below controlled for gender.

Table 2.

Descriptive statistics, correlations, and model comparison results.

	Mean	SD	Range
Anxiety	15.61	5.89	6–30
Nomophobia	86.27	24.15	20–140
Physical symptoms	23.15	19.20	0–109
Social comparison orientation	36.77	7.22	15–55
	Anxiety	Nomophobia	Physical symptoms	Social comparison orientation
Anxiety	–
Nomophobia	0.42*	–
Physical symptoms	0.39*	0.24*	–
Social comparison orientation	0.33*	0.24*	0.23*	–
Model comparisons	R 2	AIC	BIC	Bayes Factor
Full	0.188	2246.10	2263.94	0.10
Reduced	0.185	2245.09	2259.36	9.86
ΔR2 = 0.003
F[1,259] = 0.98, p = 0.32

Note:

^*p < 0.001.

Full: physical health symptoms = anxiety + nomophobia + gender.

Reduced: physical health symptoms = anxiety + gender.

Model comparisons showed no significant difference between the full model (physical health symptoms = anxiety + gender + nomophobia; F[3,258] = 19.97, R² = 0.188, p < 0.01) and the reduced model (physical health symptoms = anxiety + gender; F[2,259] = 29.46, R² = 0.185, p < 0.01), suggesting that nomophobia did not meaningfully add to variance attribution in physical health symptoms when controlling for anxiety and gender (F[1,259] = 0.98, ΔR² = 0.003, p = 0.32). Specifically, anxiety (controlling for gender) accounted for 18.5% of the variability in physical health symptoms, and increased to 18.8% with the addition of nomophobia. Additionally, the AIC, BIC and Bayes Factor favored the reduced model (see Table 2).

A general linear model also showed that, controlling for anxiety and gender, the individual main effects of nomophobia (t[254] = 0.08, SE = 0.25, R²<0.01, p = 0.94) and SCO (t[254] = −1.44, SE = 0.55, R² = 0.02, p = 0.15) did not significantly predict physical health symptoms. The interaction between nomophobia and SCO was not significant (t[254] = 0.12, SE = 0.01, R² < 0.01, p = 0.91), but the interaction effect between anxiety and SCO was significant (t[254] = 2.54, SE = 0.03, R² = 0.03, p = 0.01). Finally, the main effect of anxiety was not significant (t[254] = −1.58, SE = 1.05, R² = 0.15, p = 0.12) but gender was a significant predictor of physical health symptoms in this model (t[254] = −3.38, SE = 1.67, R² = 0.04, p < 0.01; see Table 2). This model with SCO and the interactions between anxiety and SCO, as well as nomophobia and SCO, outperformed a model without SCO and these interaction effects (ΔR² = 0.04, p < 0.01). Thus, results suggest that considering SCO, particularly its interaction with anxiety, is important in predicting physical health symptoms among college students.

Discussion

For the present report, our goals were to elucidate the association between nomophobia and physical health symptoms among college students (controlling for anxiety), identify associations between nomophobia and social comparison orientation (SCO; independent of anxiety), and determine whether SCO moderates the relation between nomophobia and physical health symptoms. Physical symptom and SCO scores reported in our sample were comparable to previous research with college students,^50–52 though global anxiety scores were lower in our sample compared to previous studies with college students.⁵³ Almost all (308/310; 99.35%) of the students in our sample reported at least mild nomophobia symptoms, indicating a striking level of distress related to being away from their phones. The prevalence of nomophobia in our sample may reflect our eligibility criterion of using digital technology daily, but are also consistent with previous findings that nomophobia is common among college students.⁵ These prevalence rates are worrisome, especially among women, as it is associated with a range of negative mental health outcomes,^8,9 physical health symptoms,¹⁹ and musculoskeletal problems.³ Indeed, the moderate association between nomophobia and physical health symptoms in the present study lends further support to the growing body of evidence that problematic mobile phone-related distress and dependency could be an indicator of risk for negative physical and psychological symptoms. Future studies should clarify the potential mechanisms underlying this association, to inform targeted approaches for prevention and treatment.

Nomophobia showed a moderate co-occurrence and bivariate association with global anxiety in the present sample, suggesting that the two constructs are related but distinct as measured; this is consistent with the intolerance of uncertainty model of anxiety and the way nomophobia is treated in existing work. However, nomophobia did not account for meaningful unique variance in physical health symptoms. Rather, self-reported global anxiety symptoms offered meaningful predictive value for physical health symptoms among college students. Specifically, global anxiety and gender as the predictors led to comparable model fit as a model that added nomophobia. This may be because students’ anxiety involves concerns their physical health: the prevalence of health or illness anxiety among college students has increased from 1985 to 2017,⁵⁴ and is not only associated with anxiety, but also measures of possible addiction to using the Internet for health purposes.⁵⁵

Together, these findings remain consistent with the notion that intolerance of uncertainty contributes to health-specific anxiety, such as illness anxiety and possible Internet addiction for health purposes specifically. Recent work has found that intolerance of uncertainty moderated the relation between health anxiety and the frequency of Internet searches for medical information,⁵⁶ as well as catastrophic health appraisals,⁵⁷ such that the association was stronger when among those with higher (vs. lower) intolerance of uncertainty. While few studies have discussed nomophobia-specific treatment, cognitive-behavioral therapy has been identified as effective in addressing nomophobia,⁵⁸ as well as global anxiety symptoms.^59,60 Focusing on concerns such as isolation, loneliness, and distress in response to uncertainty – which can contribute to different forms of anxiety⁶¹ and underlie these related conditions – may be more efficient for addressing general underlying schemas and core beliefs than focusing on phone-specific beliefs alone.

Notably, our findings indicate that nomophobia did not uniquely explain physical health symptoms, over and above the contribution of more global anxiety. However, as a considerable number of participants reported moderate-to-severe nomophobia but only none-to-mild anxiety symptoms, future research could explore whether nomophobia uniquely contributes to other concerns (e.g., mental health symptoms). More broadly, the present findings raise an interesting question about whether nomophobia should be considered a pathological condition that is distinct from anxiety. Pathologized psychological disorders tend to indicate a pattern of symptoms that uniquely implicate clinically significant deviance, distress, dysfunction, and/or danger for an individual.^62,63 Although nomophobia is associated with significant distress and impaired functioning (including among college students), it may represent one of many manifestations of global anxiety and thus, be better classified as a subtype than a distinct experience. Additional research that builds on the present study is needed to determine the optimal conceptualization and treatment of nomophobia among college students.

Finally, we did not observe a meaningful contribution from SCO with respect to predicting physical health symptoms among college students. The construct of SCO is broad and does not indicate how people respond to comparisons, which may have greater explanatory value for their physical health symptoms. Specifically, although some comparisons are associated with negative self-perceptions and affect, some are associated with positive experiences.^23,35 Thus, the role of comparisons may be complex and depend on the extent to which people experience immediate positive and negative outcomes in response, and investigation of associations with nomophobia may benefit from more nuance. For instance, the fear of missing out (FOMO) may be relevant to both nomophobia and social comparison:^64,65 seeing oneself missing out on an exciting or interesting social event that someone else is participating in may be a specific instance of focusing on the difference between oneself and an upward (better-off) other.⁶⁶

This study used a robust analytic approach to compare the predictive values of nomophobia and global anxiety among college students as well as the potential contributing role of SCO. As this represents only an initial step to determine the unique predictive value of nomophobia for physical health symptoms, however, there is need for replication of the present findings in larger, nationally representative samples of college students. These results may not generalize to other health outcomes or populations, and additional investigation of the contribution of nomophobia to physical health is warranted. Future work should also control for or otherwise address more general anxiety and should directly test whether nomophobia is associated with conditions such as social anxiety, social phobia, and other mental health conditions. As many symptoms of anxiety are physiological (e.g., hands shaking) and could be either characteristics of anxiety or downstream physical consequences, clarifying the association and causal pathway between psychological anxiety and physical symptoms remains a high priority for promoting college health.

This study had additional noteworthy limitations, including a cross-sectional design, a sample size limiting power to detect interaction effects (though we did not rely on p-values for interpretation), and eligibility criteria requiring daily use of digital technology (which may explain the high prevalence of nomophobia and limited range of variability), which should be addressed in future work. Additionally, all participants were recruited from psychology courses at one institution, which may limit the generalizability of these findings. Overall, however, the present findings suggest that nomophobia may not represent a unique target for preventing or improving physical health among college students (as it may not address their more global anxiety or underlying intolerance of uncertainty), and these findings point to specific avenues for future work in this area.