Disputation i psykologi med Fara Tabrizi

Datum: 20 november kl. 9.00

Rum: F234, Campus Östersund

Avhandlingens titel: From polygenic risk for affective disorders to stress reactivity: Transcriptomic and metabolomic investigations

Respondent: Fara Tabrizi

Huvudhandledare: Fredrik Åhs

Biträdande handledare: Billy Jansson

Ordförande: Fredrik Åhs, professor i psykologi vid institutionen för pedagogik, psykologi och socialt arbete

Opponent: Eleni Tzavara, PhD, MD, Université de Paris, Frankrike. 

Betygsnämnd:

Docent Anna Andreasson, Stockholms universitet

Professor Fang Fang, Karolinska institutet

Professor Stig Vinberg, Mittuniversitetet

Delta via Zoom

Anmälan till mingel efter disputationen

Abstract

Anxiety and depressive disorders are common, frequently co-occurring and heritable. Diathesis-stress models propose that inherited liability shapes how individuals respond to stress, but this proposition has rarely been examined using molecular measures of the acute stress response. This thesis investigated whether genetic liability for affective disorders is associated with individual differences in acute psychosocial stress responsivity, from gene expression to affective experience, and characterized the metabolic response to the same stressor. 

Study I included 13,699 twins from the Swedish Twin Registry, of whom 12,792 were genotyped. Anxiety and depressive outcomes were defined from prescription records and validated against clinical diagnoses. Twin models and 42 polygenic scores were used to characterize genetic liability and identify a score for the experimental studies. Studies II and III recruited participants from the Swedish Twin Registry into a laboratory protocol using the Trier Social Stress Test, with recruitment stratified on the depressive symptoms polygenic score to increase the range of genetic liability. Study II included 302 twins, of whom 203 had a polygenic score. Whole-blood gene expression, leukocyte subsets, cortisol, heart rate and state anxiety were measured repeatedly from baseline to 90 minutes after stress. Study III applied untargeted metabolomics to 561 plasma metabolites in 33 participants at three timepoints to establish a general metabolic response profile and identify features for future targeted investigation. 

Study I found higher twin heritability for comorbid anxiety and depression, 79%, than for anxiety or depression alone, 41% and 50%. Polygenic prediction was also strongest for the comorbid outcome, and the depressive symptoms polygenic score was the strongest predictor across all three outcomes. For clinically diagnosed outcomes, the score explained 3.6% of variance in comorbid anxiety and depression, compared with 1.3% in anxiety and 1.5% in depression, and increased the area under the curve for comorbidity from .667 to .718. Individuals in the highest score decile had more than five times the odds of comorbid anxiety and depression compared with those in the lowest decile, but only around twice the odds of either disorder alone. These findings motivated carrying the depressive symptoms polygenic score forward and oversampling the extremes of its distribution in Study II. 

In Study II, the stressor produced marked changes across all measured response systems, with 12,368 genes differing from baseline. Twin models showed that heritability of transcriptional change was concentrated in the later response, with eight individual genes meeting criteria at 90 minutes. Expression of four of these genes, FKBP5, NSFL1C, SLA and CDK5R1, correlated positively with cortisol, with coefficients ranging from .37 to .61. Additive genetic influence was also detected for stress-induced changes in heart rate, anticipatory anxiety, neutrophil counts and cortisol. Twelve of twenty physiological and subjective change scores met the criteria for additive genetic influence, with heritability estimates between 23% and 46%. Neutrophil change was heritable throughout the post-stress period and showed positive genetic correlations with a composite score of anxious-depressive traits at 60 and 90 minutes. Although these correlations did not survive correction across all models, their confidence intervals excluded zero. Higher polygenic liability was associated with downward shifts in immunoglobulin complex and antigen binding gene sets at 30 and 90 minutes. These associations were detected by ranked gene-set enrichment, while no individual gene showed a significant polygenic score by timepoint interaction. 

Study III identified 250 metabolites with stress-related changes over time, with most changes largely resolved by 90 minutes. Three clusters emerged that suggested complementary functions. Fatty acids and acylcarnitines were consistent with rapid energy mobilization, sphingolipids and phospholipids with membrane remodeling and lipid signaling, and reductions in amino acids and nucleotide intermediates with reduced anabolic and biosynthetic activity. Many of the identified metabolites overlap with those previously linked to depression, anxiety, and PTSD, suggesting that stress-induced metabolic responses may engage biological pathways relevant to psychiatric disorders and represent an initial step toward understanding how acute stress responses relate to these conditions.

Taken together, the findings indicate that genetic liability is related to specific components and phases of the acute stress response rather than to a generally stronger or weaker response. Later neutrophil mobilization and coordinated adaptive immune transcription emerged as prioritized processes through which inherited liability may be expressed in affective phenotypes, while the metabolic findings establish a reference profile for future studies of individual differences. Overall, the thesis shows that the links between genetic liability, acute stress and affective disorder become visible when the response is resolved into specific biological systems and temporal phases rather than treated as a unitary reaction.