Why Anxious Teens Spiral: Rumination, Shame, and the Fear of Fear Itself

Social anxiety is one of the most common psychological struggles of adolescence, and a new study published in Current Psychology offers one of the clearest maps yet of how it takes root in the teenage mind. Researchers at the University of Mohaghegh Ardabili in Iran set out to answer a deceptively simple question: why do some high school students, once they begin worrying about social judgment, spiral into full-blown social anxiety while others do not? Their answer centers on a trio of cognitive and emotional processes—rumination, internalized shame, and anxiety sensitivity—wired together in a pathway that could reshape how clinicians think about prevention and treatment. The work, led by Amirhossein Abdoli, Esmaeil Sadri Damirchi, and Mohammadreza Noroozi Homayoon, used structural equation modeling to test whether anxiety sensitivity acts as the crucial middleman between darker self-focused thoughts and the fear of being watched, judged, or humiliated in front of peers.

The study focused on students in Ardabil, Iran, during the 2024–2025 academic year. From a population of high schoolers who met cutoff criteria for social anxiety, the team recruited 279 participants through cluster random sampling. Each student completed a battery of well-validated questionnaires: the Social Phobia Inventory, or SPIN, to measure social anxiety symptoms; the Ruminative Responses Scale to capture the tendency to brood repetitively over negative feelings; the Internalized Shame Scale to assess the deep, stable sense of being fundamentally flawed or unworthy; and the Anxiety Sensitivity Index, which measures the fear of anxiety-related sensations themselves—the belief that a racing heart or trembling hands signal something dangerous or catastrophic. Because all participants were under eighteen, the researchers obtained written informed consent from parents or legal guardians alongside the students’ own assent, and the protocol was approved by the university’s ethics committee in accordance with the Declaration of Helsinki.

The theoretical logic behind the model draws on decades of clinical theory. Rumination, the habit of chewing over past failures and imagined future embarrassments, has long been implicated in social anxiety; a 2024 meta-analysis confirmed that post-event rumination—replaying a social encounter after it ends—strongly maintains social fear. Internalized shame operates on an even deeper level. Unlike situational embarrassment, internalized shame is a global, painful conviction that the self is defective, a state that psychologists such as Paul Gilbert have linked to perceptions of low social rank and submissive standing within a group. Anxiety sensitivity, meanwhile, is the tendency to interpret the physical and mental symptoms of anxiety—palpitations, blushing, mind going blank—as harmful in their own right. The researchers hypothesized that both rumination and shame feed into this fear-of-fear mechanism, which then amplifies social anxiety, while also exerting direct effects of their own.

The statistical machinery behind the study was rigorous. The team employed a descriptive-correlational design analyzed through structural equation modeling, a technique that allows researchers to test an entire network of hypothesized causal pathways simultaneously rather than examining relationships one pair at a time. To verify the mediating role of anxiety sensitivity, they used a bootstrap procedure with 2,000 resamples, a method widely regarded as the gold standard for estimating indirect effects because it does not assume the indirect effect follows a normal distribution. They also computed Pearson correlations among all variables and evaluated the model against a full panel of goodness-of-fit indices, including the comparative fit index, the Tucker-Lewis index, the incremental fit index, the normed fit index, and the root mean square error of approximation.

The results were striking in their consistency. Rumination predicted anxiety sensitivity with a standardized coefficient of 0.46, and internalized shame predicted it with a coefficient of 0.44, both statistically significant at the p < 0.001 level. All three variables also directly predicted social anxiety: rumination at 0.45, internalized shame at 0.42, and—most powerfully—anxiety sensitivity at 0.52. In other words, the fear of anxiety symptoms themselves emerged as the single strongest direct predictor of social anxiety in this sample of adolescents. The bootstrap analysis then confirmed the central hypothesis: anxiety sensitivity partially mediated the link between rumination and social anxiety, with a standardized indirect effect of 0.238 and a 95 percent confidence interval of 0.140 to 0.340, and likewise mediated the link between internalized shame and social anxiety, with an indirect effect of 0.226 and a confidence interval of 0.130 to 0.330. Because these confidence intervals exclude zero, the indirect pathways are statistically reliable.

Just as important as the significant pathways was the quality of the model itself. Every fit index landed in the range that methodologists consider excellent: the ratio of chi-square to degrees of freedom was 1.244, well below the conventional threshold of 3; the comparative fit index reached 0.988 and the incremental fit index 0.989, both far above the 0.95 benchmark; the Tucker-Lewis index stood at 0.983, the normed fit index at 0.944, and the parsimony-adjusted RMSEA was just 0.030 with a PCLOSE value of 0.908, indicating no meaningful discrepancy between the model and the observed data. In structural equation modeling, where complex models frequently strain against fit criteria, results this clean are uncommon, and they suggest the proposed architecture—shame and rumination feeding anxiety sensitivity, which in turn feeds social anxiety, with direct paths running alongside—captures something real about how these processes interlock in adolescence.

The word partial in partial mediation matters. Anxiety sensitivity did not fully explain the relationship between rumination, shame, and social anxiety; substantial direct pathways remained. This means that brooding thoughts and internalized shame also fuel social anxiety through other channels—perhaps by shaping negative self-imagery, anticipatory processing, or avoidance behavior—while simultaneously being funneled through the fear-of-fear mechanism. For clinicians, this dual finding is arguably the study’s most actionable insight. Interventions that target only one node in the network may leave other pathways intact. Conversely, a treatment strategy that reduces anxiety sensitivity while also addressing ruminative habits and shame-based self-evaluation could, in principle, attack the problem from multiple angles at once, weakening both the direct routes and the mediated route to social fear.

The findings arrive at a moment of growing concern about adolescent mental health. A 2024 systematic review and meta-analysis estimated substantial global prevalence of social anxiety disorder among children, adolescents, and young adults, and longitudinal research from Norway has linked adolescent social anxiety to reduced school completion and lower progression to higher education. Social anxiety in the teenage years is not a benign phase that everyone simply outgrows; it can derail education, friendships, and the developmental task of building a stable social identity. Prior work has also shown that rumination functions as a transdiagnostic process—one that cuts across depression, anxiety, and related conditions—and that co-rumination with peers can elevate anxiety sensitivity in adolescents. The new study ties these threads together into a single, testable structure applicable to a school-age population.

Several caveats frame the conclusions. The design is correlational, so the model describes patterns of association consistent with a causal story rather than proving causation; longitudinal and experimental work would be needed to establish directionality definitively. The sample comprised Iranian high school students who already met social anxiety cutoff criteria, and cultural context may shape how shame and rumination are experienced and expressed, so replication in other populations is warranted. The data were self-reported, which introduces the usual possibility of response bias. The authors also note that the underlying data are available from the corresponding author upon reasonable request, and the research received no specific external funding. Still, the study’s practical implications are clear: screening adolescents for anxiety sensitivity could help identify those most at risk, and interventions—whether cognitive-behavioral, compassion-focused, or aimed at interrupting ruminative loops—may gain leverage by explicitly teaching teenagers that a pounding heart before a class presentation is uncomfortable but not dangerous. In the architecture of adolescent social fear, the fear of fear itself may be the keystone worth pulling first.

Subject of Research: A structural equation modeling study of rumination, internalized shame, and anxiety sensitivity as predictors of social anxiety in high school students

Article Title: Developing a model of social anxiety based on rumination and internalized shame with the mediating role of anxiety sensitivity among high school students

Article References: Developing a model of social anxiety based on rumination and internalized shame with the mediating role of anxiety sensitivity among high school students. (n.d.). https://doi.org/10.1007/s12144-026-10179-5

Image Credits: AI Generated

DOI: 10.1007/s12144-026-10179-5

Keywords: social anxiety, rumination, internalized shame, anxiety sensitivity, adolescents, high school students, structural equation modeling, mediation analysis, Current Psychology, school psychology, cognitive vulnerability, bootstrap mediation