FOMO’s Underlying Dimensions Are Differentially Related to Problematic Smartphone Use but Equally Related to Depression, Anxiety, Rumination, and Distress Tolerance

Fear of missing out has long been treated as a single psychological entity, but a new study suggests that its two underlying components behave very differently when it comes to problematic smartphone use—while remaining indistinguishable in their ties to depression, anxiety, rumination, and poor distress tolerance. The research, published in the International Journal of Mental Health and Addiction, provides some of the most granular evidence to date that the apprehension of missing out and the desire to stay constantly connected are not interchangeable facets of the FOMO experience.

Since Andrew Przybylski and colleagues first formalized the construct in 2013, fear of missing out has been defined by two features: a pervasive apprehension that others might be having rewarding experiences from which one is absent, and the desire to stay continually connected with what others are doing. The first feature is largely cognitive and emotional in character, resembling worry or repetitive negative thinking about social exclusion. The second is more behavioral, functioning much like a safety behavior—frequently checking social media and messaging apps to reduce anxiety about missing rewarding experiences. Researchers have labeled these dimensions “apprehension of missing out” and “constant connection desire,” and while they are conceptually and statistically related, they are increasingly understood as distinct constructs.

The distinction matters because the two dimensions, if truly separable, might respond to different interventions and predict different downstream behaviors. A purely emotional apprehension about exclusion might be expected to travel with mood disturbance, while a behavioral drive to remain connected might be expected to manifest in compulsive device checking. Whether the empirical data support this tidy division has, until recently, been unclear, since most studies have measured FOMO as a single total score and correlated it with outcomes in aggregate.

The study was led by Jon D. Elhai of the University of Toledo, working with Tali Gazit of Bar-Ilan University, Silvia Casale of the University of Florence, and Christian Montag of the University of Macau. Between January and December 2025, the team recruited 388 undergraduate students from Introductory Psychology courses at a medium-sized Midwestern US university. Participants completed a cross-sectional web survey hosted on Psychdata.com and received course research points for taking part. The university’s Institutional Review Board approved the study before data collection began, and all participants provided informed consent.

The researchers applied careful data-cleaning procedures to ensure the quality of their sample. From an initial 422 rows of data, they excluded 14 participants who provided substantial missing data, six duplicate responses, four overly fast completers who finished in under 215 seconds, six participants younger than 18, one participant flagged by longstring detection for giving 23 consecutive identical answers, and three participants who reported not using social media at all. The final sample averaged 20.06 years of age (SD = 4.18), was roughly two-thirds female (65.7%), and was predominantly Caucasian (69.8%), with Black, Asian, and Hispanic/Latino participants also represented. Most participants were freshmen, and most worked part-time or not at all.

Participants completed a battery of validated self-report measures. FOMO was assessed with the ten-item FOMO Scale, whose items the researchers mapped onto seven apprehension items and three connection desire items following procedures established in prior work. Problematic smartphone use was measured with the Smartphone Addiction Scale–Short Version, and problematic social media use with the Bergen Social Media Addiction Scale. Negative affectivity was captured through the depression and anxiety subscales of the Depression Anxiety Stress Scale–21, the brooding subscale of the Ruminative Responses Scale–10, and the Distress Tolerance Scale—which the team rescored so that higher values indicated worse distress tolerance, aligning the direction of all measures toward pathological responding. Internal consistency was strong across most scales, ranging from coefficient alpha 0.82 to 0.92, with one notable exception: the three-item connection desire subscale showed relatively low reliability (alpha = 0.60, omega = 0.62), a limitation the authors attribute to its brevity.

Analysis proceeded in two stages. After imputing small amounts of missing item-level data—only 0.63% of item values were missing, and the missingness was statistically consistent with data missing completely at random—the team ran confirmatory factor analyses in Mplus, treating questionnaire items as ordinal given their limited response options. They first tested a single-factor FOMO model against the two-factor apprehension/connection desire model, using chi-square difference testing to compare fit. They then built combined models pairing the two FOMO factors with each other construct, using Wald chi-square tests with parameter constraints to determine whether the two FOMO dimensions correlated differentially with problematic smartphone use, problematic social media use, depression, anxiety, brooding rumination, or distress tolerance. This two-step logic—establishing the dimensional structure before testing differential correlates—guards against drawing conclusions about distinct facets from a measure that may not in fact be multidimensional.

The psychometric results supported treating FOMO as two-dimensional, though the advantage was modest. The two-factor model fit significantly better than the single-factor model (χ² difference test p = 0.0007), but the improvement in fit was small—for example, a difference of only 0.002 in the comparative fit index. This echoes earlier work by Elhai and colleagues in 2025, which had found more robust incremental fit for the two-factor structure across American and Italian samples.

The central finding concerned problematic smartphone use. Using Wald chi-square tests, the researchers found that PSU severity correlated more strongly with FOMO’s connection desire factor (r = 0.600) than with its apprehension factor (r = 0.493), a statistically significant difference (p = 0.009). This replicated the earlier 2025 findings and supports the theoretical prediction that the behavioral, safety-behavior-like dimension of FOMO is the one most tightly bound to compulsive phone use. Problematic social media use, however, told a different story: it was equally related to both dimensions (r = 0.552 for apprehension; r = 0.600 for connection desire; p = 0.122), a departure from the earlier work that had found connection desire more strongly tied to PSMU as well.

The authors offer a possible explanation for this divergence that hinges on the device itself. Problematic smartphone use, by definition, involves the smartphone, a device uniquely characterized by instant push notifications—often social in nature—that interrupt daily activities and lure users back to their screens. Recent findings suggest FOMO may serve as a mechanism linking notification receipt to smartphone addiction, potentially creating a feedback loop in which connection desire drives smartphone use, which in turn intensifies FOMO. Consistent with this, participants in the current study reported especially high use of mobile-first platforms built around frequent social updates and push notifications: 93% used Instagram, 83% Snapchat, and 74% TikTok. The authors caution, however, that this platform interpretation is descriptive and exploratory, and that the data cannot establish causality.

Equally striking is what the study did not find. Despite the apprehension dimension’s explicitly cognitive and emotional character—which might suggest a special affinity with negative affect—both FOMO dimensions correlated equally with depression (r = 0.455 vs. 0.408), anxiety (r = 0.527 vs. 0.517), brooding rumination (r = 0.503 vs. 0.453), and distress intolerance (r = 0.497 vs. 0.537), with all Wald tests far from significance. By extending the equal-relationship pattern beyond depression and anxiety to rumination and distress tolerance, the study corroborates and broadens the earlier work of Elhai, Casale, and Bond. The authors suggest that apprehension and connection desire may function as cognitive/emotional and behavioral markers, respectively, of a broader negative affectivity dimension—a view consistent with prior research showing that people high in FOMO experienced greater stress and negative emotion during a real-world social media outage.

That outage finding is worth underscoring, because it offers rare real-world evidence rather than laboratory simulation. When platforms go dark, individuals high in FOMO cannot simply check more often to soothe their worry, and their distress spikes accordingly. Such results imply that the emotional core of FOMO is not incidental to its behavioral expression but inseparable from it—precisely the pattern the equal correlations with negative affectivity suggest.

The findings are framed within two theoretical traditions. Self-determination theory conceptualizes FOMO as arising from unmet social relatedness needs, with problematic internet use serving as a maladaptive attempt to satisfy belonging. The Interaction of Person-Affect-Cognition-Execution model, meanwhile, positions internet-related cognitions like FOMO as risk factors for problematic use, and predicts that the specific technology overused will reflect the nature of the underlying cognition—here, the socially driven connection desire aligning with the notification-saturated smartphone. The authors also note that severe negative affectivity is probably a risk factor for FOMO rather than the reverse, though bidirectional relations are likely, and short-term studies have shown FOMO prospectively driving negative mood.

The study carries important limitations. Its cross-sectional design precludes causal inference, and the authors call for longitudinal and experience-sampling research. The sample—Midwestern US undergraduates in introductory psychology—limits generalizability to other ages, cultures, and settings. All measures were self-report, with no structured diagnostic interviews and no objective smartphone or social media usage data. The two-factor FOMO model showed only minimal fit improvement over the single-factor model, and the connection desire subscale’s low internal consistency complicates interpretation. The team also modified the time-frames of several instruments—the BSMAS and DASS-21 among them—to maintain consistency across measures. Still, the data, survey protocol, and analysis code are publicly available in a Mendeley repository, supporting transparency and replication.

The broader implications are twofold. Clinically and conceptually, the results suggest that experiencing FOMO may have less to do with genuine social deficits and more to do with an underlying emotional vulnerability that saturates both of its dimensions. Practically, they point toward the design of interventions: if constant connection desire is the facet most implicated in problematic smartphone use, efforts to reduce compulsive phone behavior—such as managing notification exposure—may need to target the behavioral checking loop specifically, rather than the anxious apprehension that accompanies it. As smartphone notifications continue to fragment attention across mobile-first platforms, understanding which face of FOMO drives which behavior may prove essential to addressing problematic technology use at its roots.