.png){kind=icon}
.png){kind=icon}
Return
Dynamics of Cognitions and Symptoms during Treatment: A Prospective Test of Beck's Cognitive Theory of Depression
.png){kind=icon}
PRE
AI
O
K
A
DOI：10.1159/000551172.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Introduction: Beck's cognitive theory posits that maladaptive beliefs causally drive depressive symptoms. Yet empirical support for the cognition-to-symptoms pathway is mixed, often due to sparse sampling, conflation of within- and between-person variance, and limited temporal modeling. We, therefore, aimed to address these limitations by testing reciprocal cognition-symptom dynamics in routine care using a more robust within-person longitudinal approach drawing on session-to-session assessments. Methods: Adults in Norwegian primary care receiving CBT-oriented treatment (baseline N = 1,564; PHQ-9 >= 10) completed eight assessments across approximate to 16 weeks (PHQ-9; three-item negative self-beliefs measure). Primary analyses were observation-level dynamic-panel models (DPMs); residual-level models served as corroboration. We modeled non-linearity and time trends, tested equality-over-time and varying intervals, and reported model-implied long-run effects. Results: Symptoms and negative self-beliefs declined non-linearly. Time-lagged models supported reciprocity: lower negative self-beliefs predicted later symptom reductions, and lower symptoms predicted later belief change. In DPMs, a one-point improvement in negative self-beliefs forecast a long-run 0.46-point PHQ-9 reduction (95% CI 0.22-0.70); a one-point PHQ-9 improvement forecast a 0.16-point belief change (95% CI 0.09-0.22). Standardized long-run effects were 0.36 (cognition -> symptoms; 95% CI 0.13-0.42) and 0.21 (symptoms -> cognition; 95% CI 0.12-0.26); magnitudes did not differ (chi 2(1) = 1.77, p = 0.18). Findings were robust across specifications. Conclusions: This study provides credible longitudinal, within-person evidence in line with Beck's model: improvements in negative self-beliefs precede and accumulate into subsequent symptom relief, while symptoms also feed back on beliefs. As such, it fills an important empirical gap by demonstrating a robust cognition-to-symptoms pathway under rigorous temporal modeling. Depression often involves negative beliefs about the self, such as I'm worthless or I'm a failure. Cognitive behavioral therapy (CBT) aims to reduce depression partly by changing these beliefs. However, it is not clear whether improvements in negative self-beliefs usually come before improvements in depressive symptoms, or whether symptom improvement tends to lead to more positive self-beliefs. We followed 1,564 adults with elevated depressive symptoms who received CBT-oriented treatment in Norwegian primary care. Over about 16 weeks, participants completed eight brief assessments. Depressive symptoms were measured with a standard questionnaire covering problems such as low mood, sleep difficulties, and loss of interest. Negative self-beliefs were measured with three statements about the self. We analyzed the repeated assessments to test whether a person's scores at one time point predicted their scores at the next time point, while also taking into account that people generally improve over the course of treatment. Both depressive symptoms and negative self-beliefs decreased during treatment. Lower negative self-beliefs at one assessment were followed by lower depressive symptoms at the next assessment. Symptom improvement also tended to be followed by further reductions in negative self-beliefs. The two directions of association were similar in size and the results were consistent across additional analyses. Because this was an observational study, the findings cannot prove that one change caused the other, but they show that symptom change and negative self-belief change are closely linked over time during treatment.
Keywords:
Cognitive behavioral therapy
Symptoms of depression
Depressive cognitions
Negative self-beliefs
Within-person effects
Cross-lagged panel model
Journal
IF:
17.4
Papers:
2.6K
Citations:
6.5K
Organization
Cited Papers