Clinical, genetic, and sociodemographic predictors of symptom severity after internet-delivered cognitive behavioural therapy for depression and anxiety

Clinical, genetic, and sociodemographic predictors of symptom severity after internet-delivered cognitive behavioural therapy for depression and anxiety | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Clinical, genetic, and sociodemographic predictors of symptom severity after internet-delivered cognitive behavioural therapy for depression and anxiety Olly Kravchenko, Julia Boberg, David Mataix-Cols, James Crowley, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4075444/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Internet-delivered cognitive behavioural therapy (ICBT) is an effective and accessible treatment for mild to moderate depression and anxiety disorders. However, up to 50% of patients do not experience sufficient symptom relief. Identifying patient characteristics predictive of higher post-treatment symptom severity is crucial for devising personalized interventions to avoid treatment failures and reduce healthcare costs. Using the new Swedish multimodal database MULTI-PSYCH, we expand upon established predictors of treatment outcome and assess the added benefit of utilizing polygenic risk scores (PRS) and nationwide register data in a combined sample of 2668 patients treated with ICBT for major depressive disorder ( n = 1300), panic disorder ( n = 727), and social anxiety disorder ( n = 641). We present two linear regression models: a baseline model using six well-established predictors and a full model incorporating six clinic-based, 32 register-based predictors, and PRS for seven psychiatric disorders and traits. First, we assessed predictor importance through bivariate associations and then compared the models based on the proportion of variance explained in post-treatment scores. Our analysis identified several novel predictors of higher post-treatment severity, including comorbid ASD and ADHD, receipt of financial benefits, and prior use of some psychotropic medications. The baseline model explained 27% of the variance in post-treatment symptom scores, while the full model offered a modest improvement, explaining 34%. Developing a machine learning model that can capture complex non-linear associations and interactions between high-quality multimodal input features is a viable next step to improve prediction of symptom severity post ICBT. Health sciences/Diseases/Psychiatric disorders/Depression Health sciences/Diseases/Psychiatric disorders/ADHD Health sciences/Diseases/Psychiatric disorders/Autism spectrum disorders Figures Figure 1 Introduction Major depressive disorder (MDD), panic disorder (PD), and social anxiety disorder (SAD) are highly prevalent mental health disorders 1 . MDD ranks as the first and anxiety disorders as the sixth leading cause of disability globally, measured by years lived with disability 2 . Cognitive behavioural therapy (CBT) is an effective first-line psychotherapeutic treatment for mild to moderate depression and anxiety disorders 3 , 4 . To improve accessibility, CBT is increasingly delivered in a therapist-guided internet-based format (ICBT) which has shown comparable efficacy 5 – 10 , superior accessibility for some patients 6 , and a high probability of cost-effectiveness 11 , 12 . However, CBT/ICBT is not universally efficacious, with up to 50% of patients not responding sufficiently to the treatment 13 – 15 . Therefore, it is crucial to identify reliable predictors of unfavourable treatment outcome and develop accurate predictive models that would assist clinicians in selecting appropriate, tailored care for these patients. Numerous clinical characteristics have been investigated as potential predictors of differential response to CBT. Baseline symptom severity is consistently shown to be the most robust predictor, but the direction of this association is unclear. Some studies suggest that a higher baseline symptom level is predictive of a poorer prognosis following treatment 13 , 16 – 20 . By contrast, others report greater/faster improvement in this patient group 21 – 25 , which may be explained by regression toward the mean (i.e., extreme values sampled from a random variable are likely to be closer to its mean with repeated sampling) 26 . Comorbidity with other mood and anxiety disorders 13 , 19 , 27 – 29 as well as personality disorders 27 , 30 – 32 , early onset of symptoms 17 , 27 , 33 , 34 , and family history of psychopathology 20 , 21 have also been identified as potential predictors of poorer response to CBT. Furthermore, several treatment-related variables have been linked to the outcome, such as treatment expectancy 29 , adherence 25 , 28 , 31 , homework completion 35 , and working alliance 36 . Lastly, among sociodemographic patient characteristics, being employed 20 , 28 , 31 – 33 , married 23 , and having higher education 22 are predictive of a lower post-treatment symptom severity. Genetic differences are implicated in a wide range of individual variation in human complex traits, including differential susceptibility and response to environmental stimuli. A new area known as therapygenetics views therapeutic interventions as such a stimulus, providing the basis for the hypothesis that genetic variation could, at least partially, explain treatment effect heterogeneity 37 . Several studies suggest that polygenic risk scores (PRS) can be used to predict response to pharmacological treatments of psychiatric disorders 38 – 40 . PRS are constructed as a weighted sum of risk alleles of single nucleotide polymorphisms (SNPs) using summary statistics from genome-wide association studies (GWAS). Preliminary evidence suggests a link between a higher polygenic loading for autism spectrum disorder (PRS ASD) and a poorer treatment response in patients undergoing ICBT for MDD 41 . In another study using the same data as 41 with machine learning modelling, higher PRS MDD and PRS for intelligence (PRS IQ) weakly predicted remission in ICBT for MDD 42 . However, the largest therapygenetic GWA meta-analysis to date ( n = 2724) did not identify any common genetic variants associated with the CBT treatment outcome for depression and anxiety 43 , consistent with a previous GWAS that had a similar sample size and was also likely underpowered 44 . Despite the abundance of literature on predictors of treatment outcome, there are some limitations that this study aimed to address. Most existing studies are based on a relatively small number of clinical predictors that are derived from secondary data analysis of randomized controlled trials (RCT), which are costly to conduct and rely on small samples of participants who are not representative of the broader patient population, thus providing inconsistent evidence on predictor importance. We contribute to the state of the field by leveraging a recently constructed MULTI-PSYCH cohort ( n = 2668) with a broad selection of potential predictors, integrating data routinely collected at the clinic with additional inputs such as PRS and a broad range of variables from Swedish population registries 45 . The aim of the present study was twofold: to identify patient characteristics predictive of higher post-treatment symptom severity and to evaluate whether incorporating genetic and register predictors alongside established clinical predictors explains additional variance in post-treatment symptom severity. Patients and methods Sample The MULTI-PSYCH cohort study comprises 2668 adult outpatients treated with 12 weeks of ICBT for MDD ( n = 1300), PD ( n = 727) or SAD ( n = 641) at the Internet psychiatry unit of the Psychiatric Clinic Southwest at Karolinska University Hospital Huddinge, Sweden, between 2008 and 2020 45 . Screening and treatment procedures are described in more detail elsewhere 46 . As part of the assessment, patients completed an online questionnaire and were interviewed by a licensed clinician (psychologist or psychiatrist), providing sociodemographic information along with details about their condition and medical history. Blood samples were collected for DNA extraction before treatment, and data from multiple nationwide registers were linked using each patient’s unique Swedish personal number. Variables Candidate predictors were divided into three groups based on their source: clinic-based (obtained as part of pre-treatment questionnaires and interviews, containing medical and sociodemographic data), genetic (PRS for seven traits derived from GWAS summary statistics and calculated for all genotyped patients), and register-based (sourced through record linkage across multiple national registers containing medical, socioeconomic, and demographic data). This division is helpful for assessing the relative contribution of predictors derived from different sources: genetic and register data are not readily available in the routine care setting, and thus, their inclusion must be justified by substantial added explanatory power. The purpose of this study was to identify predictors that have prognostic value, and thus, selected variables must be available pre-treatment. Process variables, e.g., adherence and homework completion, were therefore excluded. Clinic-based predictors were pre-selected based on literature review and established evidence to minimize predictor selection bias. Studies were eligible for inclusion if conducted in adults treated with internet-delivered or face-to-face CBT for MDD, PD or SAD. See Supplementary Information, S1 Table for a list of established predictors. By contrast, register data are rarely accessible for research purposes, leading to a scarcity of literature on established predictors. Thus, primary selection of register-based variables was guided by their face validity and subsequently validated through bivariate regression analyses. Those predictors that were significantly associated with the outcome were added to the final model. Clinic-based predictors Baseline symptom severity was measured right before ICBT start with the Montgomery–Åsberg Depression Rating Scale, self-rating version (MADRS-S) for MDD (test–retest reliability: r = 0.78; internal consistency reliability: α = 0.84) 47 , Panic Disorder Severity Scale, self-report (PDSS-SR) for PD (test–retest reliability: r = 0.81; internal consistency reliability: α = 0.92) 48 , and Liebowitz Social Anxiety Scale, self-report (LSAS-SR) for SAD (test–retest reliability: r = 0.83; internal consistency reliability: α = 0.95) 49 . In the present study, the three disorders along with their respective predictors and outcome measures were combined to increase statistical power. While this decision comes at the cost of introducing more heterogeneity, we deemed it conceptually reasonable given the magnitude of phenotypic and etiological overlap between MDD, PD, and SAD, the similarities in the three ICBT treatment protocols, and the growing support for applying transdiagnostic models of psychopathology 50 – 52 . Finally, a recent machine-learning study aimed at investigating the value of pooling intervention data for MDD, PD, and SAD into a single dataset and using a larger sample derived from the same source as the current study supported the superiority of this approach 53 . The disorder-specific scales were transformed to a harmonized 0-100 range and subsequently pooled into a single score 54 (see Supplementary Information). Comorbidity is a binary measure that indicates the presence of coexisting psychiatric disorders as assigned by a clinician during a screening interview. Family history captures self-reported confirmed or suspected cases of psychopathology among first-degree relatives selected from a list of 36 options. Sex was derived from the Swedish personal identity number and therefore refers to biological sex. Highest achieved level of education was self-reported at screening and for statistical analysis dichotomised into University (finished or unfinished) and No university (up to upper secondary school). Marital status is a self-reported civil status dichotomised into Married (officially married or de facto married, i.e., in a committed relationship (‘married/living together/living apart together’)) and Unmarried (Single , Separated , and Widowed). Finally, Parental status was retrieved from the screening question, "Do you have children?". Genetic predictors Genotyping was performed in three batches at LIFE & BRAIN GmbH in Bonn, Germany, with Illumina HumanCoreExome-12 v1.0, Infinium Global Screening Array-24 v1.0, and Infinium Global Screening Array-24 v2.0, respectively. Quality control and imputation were performed via the RICOPILI GWAS pipeline 55 . PRS for seven traits (MDD, attention deficit hyperactivity disorder (ADHD), ASD, bipolar disorder (BPAD), schizophrenia (SCZ), IQ, and educational attainment (EA)) were constructed using the PRS-CS method. A detailed description of genetic data pre-processing can be found elsewhere 56 . Register-based predictors The Swedish national population registration system provides high-quality data on important life events for all individuals who have been assigned a personal identity number and is unique in its almost complete coverage 57 . Register-based employment status was derived from the Register-based Employment Statistics (RAMS) 58 and has been dichotomised into Employed / Unemployed . Income was obtained from the Total Population Register 59 and is defined as annual disposable household income divided by consumption weight of the family, thus constituting an individual’s component of the household income adjusted for family composition. For statistical analyses, Income was divided into quintiles. Lastly, Financial benefits is receipt of any of eight benefits from the Swedish social insurance system (for a complete list, see Supplementary Information). Statistics on employment, income, and financial benefits are updated annually; hence, the most recent available record from the year preceding treatment was utilized. Prior psychiatric diagnoses were derived from the National Patient Register (NPR) 60 and the Stockholm regional healthcare data warehouse VAL (Vårdanalysdatabasen) 61 . NPR comprises data on specialized in- and outpatient care visits in Sweden, while VAL registers all primary care visits in Stockholm County, where most of the patients reside. In addition to ICD-10 Chap. 5 codes (e.g. F3: Mood [affective] disorders), a separate variable was constructed for BPAD to avoid conflation with depressive disorders which are the target in the studied ICBT treatment. Separate variables were also created for ASD and ADHD and, given their neurodevelopmental nature and persistent lifelong manifestation, the inclusion was not limited to diagnoses received prior to ICBT treatment. To commence the ICBT treatment, a patient must receive a relevant diagnosis, either from a general practitioner or a clinician at the Internet Psychiatry clinic, which is subsequently transferred to the NPR. Hence, to avoid capturing the diagnosis related to this very healthcare episode and only focus on the prior events, all records of MDD/PD/SAD within a month preceding the ICBT treatment were excluded from the analysis. Data on dispensation of prescription medicines were retrieved from the National Prescribed Drug Register 62 and included Antidepressants (N06A), Anxiolytics (N05B), Hypnotics and sedatives (N05C), and Antipsychotics (N05A). As with the psychiatric diagnoses, all records of dispensation within a month preceding ICBT treatment were excluded from the analysis. Binary variables for Education ( University / No university ), Marital status ( Married / Unmarried ), and Parental status ( Children / No children ) were constructed from both clinic-based and register-based data. The rationale for using the clinic-based measure in the full model, as well as a detailed description of data pre-processing, is provided in Supplementary Information. Outcome Treatment outcome is defined as post-treatment symptom level, measured by a self-rated, disorder-specific scale. Identically to the baseline symptom severity, harmonized scores for the three disorders were pooled. Statistical analysis Missing data Summary of missing data is presented in Supplementary Information, S2 Table. Clinic-based predictors with over 30% missing values were excluded. When available, missing pre-treatment symptom severity values ( n = 30, 1.1%) were replaced with those collected at screening, resulting in three missing values in the final dataset (0.1%). For all the PRS, 456 (17.1%) observations were missing due to being excluded during quality control procedures (poor matching and partial non-European ancestry). Among register-based predictors, there were six missing values in Employment, Income, and Financial benefits. The outcome variable, post-treatment symptom severity, had 524 missing observations (19.6%). The missingness mechanism was hypothesized to be missing at random (MAR), i.e., the missingness is conditional on other variables in the dataset, and the sensitivity analysis was performed to check the robustness against departures from this assumption (see the results in Supplementary Information). The missingness pattern suggested that non-completers were more likely to have higher post-treatment symptom values since the variables associated with missingness were also predictive of higher outcome values in the non-missing subset. Thus, listwise deletion (complete-case analysis) was ruled out for two reasons: it negatively affects the precision of estimates by decreasing power and, more importantly, if missing data are not MCAR, it is likely to introduce selection bias into inferences and cause too conservative estimates 63 – 69 . To address the uncertainty introduced by missingness, missing values were imputed with multiple imputation using the R package mice 70 . A total of m = 20 datasets were imputed over 20 iterations. In accordance with the recommended procedure, 25 auxiliary variables not included in the analytical sample were added to the imputation dataset to improve the quality of imputations through increasing the plausibility of the MAR assumption 66 . Missing values were imputed using the default settings of the mice package. For numeric data, predictive mean matching ( pmm ) was used as the imputation method, replacing a missing value with a randomly selected observed donor that has a regression-predicted value closest to the missing value based on a simulated regression model. It is a superior method to regression in that it does not rely on the assumption of joint multivariate normal distribution and creates realistic imputed values. For binary data, logistic regression imputation ( logreg ) was used and for unordered categorical data, polytomous regression imputation ( polyreg ). To check the adequacy of the imputation models and assess convergence, both graphic and numeric diagnostic methods were employed, namely kernel density plots of the distributions of the observed and m = 20 imputed datasets and the Kolmogorov-Smirnov test to assess departures from the assumptions made in the imputation model 68 . Main analysis All statistical analyses were performed using R (version 4.3.1). First, univariate analyses were performed on all predictors to examine their distribution and properties. Next, two models were developed: a baseline model (clinic-based data only) and a full model (clinic-based, genetic, and register data). The same steps of statistical analysis were repeated for both models: first, separate bivariate linear regression models were fit to each predictor; next, all covariates were entered simultaneously into the multiple linear regression model. The purpose of conducting the analysis in two steps was dictated by the research questions. Given the complex nature of the studied phenomenon and the known associations between the covariates from existing literature on latent human traits, at least partial collinearity is inevitable. Thus, to evaluate independent associations between individual predictors and the outcome, assessment of simple regression coefficients is necessary to avoid misinterpreting small effect sizes, whereby a variable important in a bivariate model might be attenuated in a multiple model by other collinear predictors. For the second goal of the overall model appraisal, the accuracy of point estimates is not paramount. Therefore, variable selection was theory-driven, and all pre-selected variables were retained in the multiple regression model irrespective of their statistical significance. This was done to avoid the flaws related to the algorithm of stepwise variable selection whereby variables are added and removed based on their significance level 71 . A total of 45 predictors were used in the full model. All variables with zero variance were excluded from the analysis. Variables with very low variance (meeting both criteria: frequency of the second most common value of < 1% over the sample and percentage of unique values of < 10% of the total number of data points) ( n = 46) were excluded as independent predictors but retained within relevant composite variables. For a full list of included and excluded variables, see Supplementary Information. For prior diagnoses and medication, both composite and individual components were assessed through bivariate analyses for the sake of potentially discovering novel predictors. For example, using a composite variable that encompasses any prior psychotropic medication may be justified in the interest of model parsimony, but it may miss what specific medication drives the effect size of the association. However, only individual predictors were used in multiple regression analysis to avoid multicollinearity. Lastly, the baseline and the full model were evaluated based on their goodness of fit using adjusted coefficients of determination ( \({R}_{adj}^{2}\) ). All additional covariates are likely to increase \({R}^{2}\) , regardless of their true explanatory power. In contrast, \({R}_{adj}^{2}\) imposes a penalty for this inflation when comparing models with a varying number of predictors. Consequently, the assessment of the full model performance compared to the baseline model relied on the extra variance explained by additional predictors and was complemented with the appraisal of the Akaike information criterion (AIC) and root-mean-square error (RMSE) of the two models. Results Descriptive statistics Sample characteristics are presented in Table 1 (for more details, see Supplementary Information, S4 Table). Descriptive statistics for the observed, imputed, and complete dataset, as well as the distributional discrepancy between observed and imputed outcome values, are provided in Supplementary Information, S5 Table and Supplementary Information, Figure S1, respectively. Table 1 Descriptive statistics for the total sample and stratified by disorder Variable Total ( n = 2668) MDD ( n = 1300) PD ( n = 727) SAD ( n = 641) Pre-treatment symptom severity, mean (SD) 47.6 (17.8) 51.6 (15.7) 40.2 (18.8) 47.7 (17.9) Post-treatment symptom severity, mean (SD) 28.2 (18.8) 30.3 (18.7) 17.8 (16.1) 35.3 (17.2) Relative symptom change, % (SD) -41.3 (37.9) -41.7 (34.3) -53.4 (47.0) -28.1 (26.5) Age, mean (SD) 35.6 (11.4) 37.6 (11.9) 34.6 (10.8) 32.7 (10.3) Male sex, n (%) 1014 (38.0%) 443 (34.1%) 292 (40.2%) 279 (43.6%) Psychiatric comorbidities, n (%) 862 (33.5%) 394 (31.4%) 265 (37.9%) 203 (32.9%) Family history of psychopathology, n (%) 1812 (70.6%) 888 (71%) 488 (69.7%) 436 (70.7%) Self-reported married, n (%) 1566 (58.8%) 727 (56.1%) 477 (65.7%) 362 (56.7%) Self-reported having children, n (%) 1102 (41.4%) 603 (46.4%) 302 (41.5%) 199 (31.0%) Self-reported university education, n (%) 1693 (63.6%) 890 (68.8%) 406 (55.9%) 397 (62.1%) Register-based employed, n (%) 2448 (92.0%) 1205 (92.7%) 674 (92.8%) 571 (89.1%) Annual income (SEK), mean (SD) 241,713 (201,190) 256,187 (202,155) 244,258 (232,968) 209,591 (150,742) Financial benefits in the past year, n (%) 556 (20.9%) 302 (23.3%) 148 (20.4%) 106 (16.5%) Prior psychiatric diagnosis, n (%) 1793 (67.2%) 895 (68.8%) 518 (71.3%) 380 (59.3%) Prior psychotropic medication, n (%) 1665 (62.4%) 863 (66.4%) 472 (64.9%) 330 (51.5%) SD, standard deviation; SEK, Swedish krona; MDD, Major depressive disorder; PD, Panic disorder; SAD, Social anxiety disorder Predictor importance Parameter estimates for clinic-based, genetic, and register-based predictors are displayed in Table 2 . Table 2 Association between clinic-based, genetic, and register-based predictors and treatment outcome. Results from bivariate and multiple regression analyses in the full model Predictor Bivariate Multiple Estimate (95% CI) p -value Estimate (95% CI) p -value Pre-treatment score 0.57 (0.53, 0.61) < .001*** 0.51 (0.47, 0.55) < .001*** Comorbidities 5.20 (3.54, 6.86) < .001*** 0.42 (-1.03, 1.88) .569 Family history 2.75 (0.92, 4.58) .003** 0.19 (-1.41, 1.79) .816 Sex (ref: male) 0.58 (-1.06, 2.23) .488 -2.56 (-4.03, -1.09) < .001*** Unmarried 4.62 (2.89, 6.34) < .001*** 1.67 (0.08, 3.26) .040* No children 2.28 (0.72, 3.85) .004** 0.72 (-0.84, 2.28) .364 No university 4.18 (2.52, 5.85) < .001*** 1.95 (0.49, 3.42) .009** Unemployed 7.45 (4.15, 10.75) < .001*** 2.98 (-0.01, 5.97) .051 Income (quintiles) -1.90 (-2.47, -1.34) < .001*** -0.81 (-1.38, -0.25) .005** Any financial benefits 6.24 (4.23, 8.25) < .001*** 2.24 (0.37, 4.11) .019* PRS MDD 0.71 (-0.19, 1.60) .120 0.41 (-0.39, 1.21) .310 PRS ASD 0.58 (-0.37, 1.53) .228 0.43 (-0.45, 1.32) .335 PRS ADHD 0.68 (-0.33, 1.69) .182 0.14 (-0.71, 0.99) .741 PRS BPAD -0.21 (-1.07, 0.66) .640 -0.73 (-1.57, 0.12) .091 PRS Education -0.30 (-1.24, 0.64) .527 0.24 (-0.73, 1.20) .630 PRS IQ 0.37 (-0.59, 1.33) .444 0.38 (-0.58, 1.34) .434 PRS SCZ 0.28 (-0.63, 1.19) .544 0.46 (-0.45, 1.38) .315 Any prior psychiatric diagnosis 5.54 (3.88, 7.20) < .001*** Prior F1 3.16 (-0.07, 6.39) .055 -1.76 (-4.63, 1.11) .230 Prior F3 (excl. BPAD) 7.61 (5.89, 9.34) < .001*** 2.29 (0.56, 4.02) .010* Prior F4 3.29 (1.69, 4.90) < .001*** 0.93 (-0.61, 2.48) .235 Prior F5 7.84 (5.05, 10.64) < .001*** 2.70 (0.11, 5.30) .041* Prior F6 15.70 (9.35, 22.06) < .001*** 5.55 (-0.11, 11.20) .055 ASD 24.98 (18.35, 31.61) < .001*** 10.35 (4.17, 16.53) .001** ADHD 15.98 (12.24, 19.71) < .001*** 6.38 (2.91, 9.85) < .001*** Any prior medication 4.52 (2.89, 6.16) < .001*** Antidepressants 6.30 (4.66, 7.94) < .001*** Bupropion 14.16 (9.75, 18.58) < .001*** 6.04 (1.85, 10.24) .005** Duloxetine 10.84 (6.07, 15.61) < .001*** -0.49 (-4.87, 3.90) .827 Fluoxetine 10.52 (6.97, 14.07) <.001*** 4.18 (0.91, 7.44) .012* Venlafaxine 9.44 (5.77, 13.10) < .001*** 2.78 (-0.62, 6.18) .108 Mirtazapine 8.31 (5.12, 11.50) < .001*** 0.48 (-2.62, 3.59) .759 Amitriptyline 7.93 (2.89, 12.96) .002** 4.04 (-0.27, 8.35) .066 Sertraline 6.51 (4.56, 8.47) < .001*** 2.70 (0.88, 4.52) .004** Escitalopram 6.42 (3.63, 9.21) < .001*** 1.77 (-0.72, 4.27) .163 Citalopram 3.74 (1.60, 5.87) < .001*** 2.20 (0.22, 4.18) .030* Clomipramine 1.71 (-5.96, 9.38) .662 -2.16 (-8.73, 4.41) .518 Paroxetine -1.49 (-5.97, 2.98) .512 -0.09 (-3.95, 3.77) .963 Anxiolytics 3.14 (1.40, 4.89) < .001*** Buspirone 19.41 (11.63, 27.20) < .001*** 7.73 (0.61, 14.84) .033* Alprazolam 4.00 (-1.60, 9.59) .161 1.70 (-3.23, 6.62) .498 Hydroxyzine 3.16 (1.26, 5.05) .001** -1.78 (-3.72, 0.15) .070 Diazepam 2.59 (-1.36, 6.55) .198 -1.15 (-4.60, 2.29) .512 Oxazepam 2.25 (-0.02, 4.53) .052 -0.84 (-3.00, 1.32) .444 Hypnotics and sedatives 6.20 (4.40, 7.99) < .001*** Melatonin 11.02 (5.74, 16.30) < .001*** 2.56 (-2.31, 7.42) .301 Zolpidem 6.23 (3.46, 9.01) < .001*** 0.91 (-1.73, 3.56) .498 Zopiclone 5.79 (3.40, 8.18) < .001*** 0.03 (-2.49, 2.43) .984 Propiomazine 5.18 (2.80, 7.56) < .001*** -1.56 (-3.91, 0.79) .193 Antipsychotics 5.98 (1.37, 10.59) .011* -3.63 (-7.79, 0.53) .087 PRS, polygenic risk score; MDD, Major depressive disorder; ASD, Autism spectrum disorder; ADHD, Attention deficit hyperactivity disorder; BPAD, Bipolar affective disorder; IQ, intelligence quotient; SCZ, Schizophrenia; F1, Mental and behavioural disorders due to psychoactive substance use; F3, Mood [affective] disorders; F4, Anxiety, dissociative, stress-related, somatoform and other nonpsychotic mental disorders; F5, Behavioural syndromes associated with physiological disturbances and physical factors; F6, Disorders of adult personality and behaviour; CI, confidence interval Significance levels: *p < .05, **p < .01, ***p < .001 As expected, pre-treatment symptom level was a strong predictor of post-treatment score severity (β = 0.57, 95% CI [0.53, 0.61], p < .001), remaining almost unattenuated in the adjusted model. No significant association between PRS and the outcome was observed. Socioeconomic predictors (no university education, unemployment, receipt of financial benefits, and lower income quintile) were strongly associated with higher post-treatment symptom severity and remained significant in the multiple regression model. Being childless was associated with a higher post-treatment score in the bivariate model, but the effect disappeared in the adjusted model. By contrast, female sex showed association with lower post-treatment score in the adjusted model only. Having a history of psychiatric diagnoses predicted higher post-treatment symptom severity. In addition to previous records of depressive and anxiety disorders, eating and personality disorders were strongly associated with poorer treatment outcome. A particularly large effect sizes were found for comorbid ASD (β = 24.98, 95% CI [18.35, 31.61], p < .001) and ADHD (β = 15.98, 95% CI [12.24, 19.71], p < .001), the observed relationship persisted in the adjusted model. The other significant predictor was prior use of almost all psychotropic medications. In addition to the substantial effect of most antidepressants and some anxiolytics (particularly buspirone), it is worth emphasizing quite a strong association with the prior use of hypnotics and sedatives (with a notably large effect of melatonin) as well as antipsychotics. Model performance Estimates of comparative performance of the models are presented in Table 3 and visualized in Fig. 1 . \({R}_{adj}^{2}\) of the baseline model was .27, thus explaining 27% of the variance in post-treatment symptom severity, which is almost entirely driven by pre-treatment symptom severity. The full model yielded \({R}_{adj}^{2}\) of .34. Addition of PRS did not contribute to the model performance, increasing \({R}_{adj}^{2}\) of the full model by only 0.2 percentage points above clinic-based and register-based predictors. The full model was also superior when comparing the two models’ AIC and RMSE. These results suggest that including rich register data provided some additional explanatory power beyond the clinic-based predictors used in the baseline model, accounting for a further 7% of the variance. Table 3 Adjusted R-squared ( \({\varvec{R}}_{\varvec{a}\varvec{d}\varvec{j}}^{2}\) ), Akaike information criterion (ΔAIC), and Root mean square error (RMSE) values of baseline and full models Model \({R}_{adj}^{2}\) ΔAIC RMSE # of predictors Baseline model 0.27 22613.5 15.2 6 Full model 0.34 22411.2 14.3 45 \({R}_{adj}^{2}\) , Adjusted R-squared; ΔAIC, Akaike information criterion; RMSE, Root mean square error Discussion The first goal of this study was to identify predictors of post-treatment symptom severity in patients with mild to moderate MDD, PD, and SAD treated with highly standardized ICBT protocols and using a large and diverse pool of clinical, genetic, and register predictors. Our findings support the previously established importance of pre-treatment symptom severity, psychiatric comorbidities, family history of psychopathology, and socioeconomic status. In line with existing research on the link between marital status and mental health, being single, divorced or widowed was predictive of a poorer treatment outcome. Furthermore, thanks to unique data sources available in this study, we were able to assess a variety of predictors that have not been investigated before: the role of prior and concurrent psychiatric diagnoses and specific medications is a novel addition of this study, as are socioeconomic predictors such as income and receipt of financial benefits. Identification of comorbid ASD and ADHD as strong hindering factors bears potential clinical relevance. The need to adjust psychotherapeutic interventions for depression and anxiety to the patient’s ASD symptomatology has long been acknowledged. Socio-communication impairment, difficulties with introspection, and limited cognitive flexibility have been suggested as impeding treatment effectiveness 72 . Consequently, multiple CBT adaptations for patients with comorbid ASD have been developed 73 . The discovery that in this patient group, the average post-treatment score was nearly twice as high as for patients without ASD potentially highlights absence of relevant modifications in the studied ICBT treatment. In patients with comorbid ADHD, the findings may reflect a similar lack of treatment accommodation. Perhaps unsurprisingly, maintaining attention and organizing oneself over a 12-week treatment period, which involves extensive homework and the absence of immediate therapist guidance that is characteristic of ICBT, may pose a challenge for this patient group and calls for adaptive treatment strategies to avoid treatment failure 74 . The application of PRS in psychiatry is promising as it is a constant throughout lifetime and can be used for long-term prediction. Yet, their limitation is that on a population level, PRS tend to follow a Gaussian distribution with significant overlap between cases and controls, and they only explain a small fraction of genetic variation, which in turn explains around 50% of phenotypic variation. Failure to detect the association between PRS and post-treatment symptom severity in our study is perhaps not surprising. We used PRS related to psychopathology and personality, and, while genetic liability to these traits is likely to be shared with that of treatment outcome, it does not necessarily follow that the same PRS will be helpful in predicting it. Thus, a specific PRS constructed from GWAS of treatment response, which is still missing, could potentially yield a stronger effect. The second goal of the study was to assess whether employing a wider range of predictors would be advantageous in identifying patients who will have higher post-treatment symptom severity. We found that the full model with multimodal predictors had a superior performance in explaining the variance in post-treatment severity compared to the baseline model. However, the improvement in explained variance was modest. One potential interpretation is that even though the evaluated predictors encompass a distinct range of properties, they might, to a certain degree, be considered as capturing a shared underlying construct and providing little independent information. For example, genetic predictors will inevitably be correlated due to vertical pleiotropy, whereby the same genetic variant affects multiple different traits, e.g., a relevant SNP influences intelligence, which serves as a mediator influencing educational attainment, SES, and, in turn, post-treatment symptom severity. While the collinearity of independent variables does not negatively affect the explanatory power of the whole model, it does not improve it either, since many variables explain an overlapping proportion of variance in the outcome. Further, the full model would likely benefit from the inclusion of additional strong predictors. For example, in the exploratory complete-case analysis, duration of symptoms was strongly associated with the outcome in its effect size and variance explained but could not be included as it surpassed the predefined threshold for the allowed amount of missingness. Moreover, human complex traits are dominated by stochasticity and are often attributed to idiosyncratic factors which remain largely unmeasured. Another possibility is that pooling the three disorders makes the studied phenotype more heterogeneous and possibly dilutes the findings. Also, while the current sample size exceeds those in the earlier studies, it may still not be sufficiently powered. Finally, the relationship between predictors and the outcome may be non-linear and contain underlying interaction effects, making classical ordinary least squares modelling unsuitable. Strengths and limitations The main strength of this study is its relatively large sample size and diversity of evaluated predictors. Moreover, register-based data have almost no missing values and are highly reliable. Another strength is a homogenous group of patients that completed a highly protocolized treatment, which allows for meaningful comparisons of the outcome. Deriving the data from routine care has additional benefits since most studies of treatment outcome constitute a primary or secondary analysis of RCT data, where user characteristics are subject to stringent inclusion criteria, thus introducing selection bias and limiting the generalisability 75 . However, a potential limitation in the applicability of the findings to other populations needs to be mentioned. All the study participants lived in Sweden, were more educated than the general population, were fluent in Swedish, and exhibited mild to moderate symptom severity. In addition, there is a significant self-selection bias, with most patients self-referring to the Internet psychiatry clinic. Finally, it must be emphasized that the findings of this study cannot be viewed as supporting the causal nature of the relationship between any of the predictors and the outcome. While the terms treatment outcome and treatment response do semantically assume, at least partially, that the post-treatment symptom measure is conditional upon and a direct consequence of the intervention, no such claims can be made given the non-experimental design of the study. However, causal language is widely and loosely applied across the literature and, in the interest of brevity and recognition, is also sometimes used in this paper, more so when referring to the previous findings rather than the current study, where we try to adapt the more observational term post-treatment symptom severity . This operationalization of the response variable was chosen as it is deemed by the authors as the most appropriate measure to be predicted at the baseline. Since the goal of the future predictive model is not to assess treatment effectiveness, relative symptom change was not of interest here. Moreover, even when the percentage change is substantial and passes some predefined threshold, the patient may still be symptomatic beyond what is intended. Choosing remission as the outcome was rejected due to its somewhat arbitrary cut-off, loss of information, and diminished statistical power that ensues from dichotomization. Conclusions The current study adds to the growing literature on predictors of treatment outcome by supporting established ones and suggesting some novel predictors. It also proposes that a statistical model based on a few relatively easily obtainable measures explains a comparable proportion of the variance in the outcome as a model containing a much broader array of multimodal data. As the next step, a machine learning model should be developed to address non-linear associations and higher-order interactions between input features to perform predictions at individual patient level. Additional future endeavours may include leveraging all available SNP information beyond what meets a strict GWAS significance threshold and appraising an even larger and more diverse pool of register-based predictors. Declarations Code availability R code can be provided upon request to the corresponding author. Acknowledgements This study was funded by the Söderström-König Foundation, FORTE, the Swedish Research Council, and the Centre for innovative medicine. We gratefully acknowledge the contribution of the psychiatric research nurse Monica Hellberg to the collection of blood samples and of Bjorn Roelstraete to the dataset preparation. Author contributions OK, JW, and CR designed the study; OK conducted data analysis and wrote the manuscript; JJC and MH conducted pre-processing of genetic data and constructed PRS; all authors contributed to and approved the manuscript. Conflict of interest The authors declare that they have no conflict of interest. References Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry 2005; 62 : 593–602. World Health Organization. Depression and Other Common Mental Disorders: Global Health Estimates. World Health Organization, 2017. https://www.jstor.org/stable/resrep28026.1 (accessed 7 Mar 2024). American Psychological Association. Clinical Practice Guideline for the Treatment of Depression Across Three Age Cohorts. https://www.apa.org. https://www.apa.org/depression-guideline (accessed 4 Oct 2023). Practice Guideline for the Treatment of Patients With Panic Disorder. In: APA Practice Guidelines for the Treatment of Psychiatric Disorders: Comprehensive Guidelines and Guideline Watches . American Psychiatric Association: Arlington, VA, 2006 doi:10.1176/appi.books.9780890423363.51396. Carlbring P, Andersson G, Cuijpers P, Riper H, Hedman-Lagerlöf E. Internet-based vs. face-to-face cognitive behavior therapy for psychiatric and somatic disorders: an updated systematic review and meta-analysis. Cogn Behav Ther 2018; 47 : 1–18. Hedman E, Ljótsson B, Lindefors N. Cognitive behavior therapy via the Internet: a systematic review of applications, clinical efficacy and cost–effectiveness. Expert Rev Pharmacoecon Outcomes Res 2012; 12 : 745–764. Kiropoulos LA, Klein B, Austin DW, Gilson K, Pier C, Mitchell J et al. Is internet-based CBT for panic disorder and agoraphobia as effective as face-to-face CBT? J Anxiety Disord 2008; 22 : 1273–1284. Hedman E, Andersson G, Ljótsson B, Andersson E, Rück C, Mörtberg E et al. Internet-Based Cognitive Behavior Therapy vs. Cognitive Behavioral Group Therapy for Social Anxiety Disorder: A Randomized Controlled Non-inferiority Trial. PLoS ONE 2011; 6 : e18001. Andrews G, Davies M, Titov N. Effectiveness Randomized Controlled Trial of Face to Face Versus Internet Cognitive Behaviour Therapy for Social Phobia. Aust N Z J Psychiatry 2011; 45 : 337–340. Andrews G, Cuijpers P, Craske MG, McEvoy P, Titov N. Computer Therapy for the Anxiety and Depressive Disorders Is Effective, Acceptable and Practical Health Care: A Meta-Analysis. PLoS ONE 2010; 5 : e13196. Warmerdam L, Smit F, Van Straten A, Riper H, Cuijpers P. Cost-Utility and Cost-Effectiveness of Internet-Based Treatment for Adults With Depressive Symptoms: Randomized Trial. J Med Internet Res 2010; 12 : e53. Kraepelien M, Mattsson S, Hedman-Lagerlöf E, Petersson IF, Forsell Y, Lindefors N et al. Cost-effectiveness of internet-based cognitive–behavioural therapy and physical exercise for depression. BJPsych Open 2018; 4 : 265–273. Rozental A, Andersson G, Carlbring P. In the Absence of Effects: An Individual Patient Data Meta-Analysis of Non-response and Its Predictors in Internet-Based Cognitive Behavior Therapy. Front Psychol 2019; 10 : 589. Andersson G, Carlbring P, Rozental A. Response and Remission Rates in Internet-Based Cognitive Behavior Therapy: An Individual Patient Data Meta-Analysis. Front Psychiatry 2019; 10 : 749. Loerinc AG, Meuret AE, Twohig MP, Rosenfield D, Bluett EJ, Craske MG. Response rates for CBT for anxiety disorders: Need for standardized criteria. Clin Psychol Rev 2015; 42 : 72–82. Andersson G, Bergström J, Holländare F, Ekselius L, Carlbring P. Delivering Cognitive Behavioural Therapy for Mild to Moderate Depression via the Internet: Predicting Outcome at 6-Month Follow-Up. Verhaltenstherapie 2004; 14 : 185–189. Hamilton KE, Dobson KS. Cognitive therapy of depression. Clin Psychol Rev 2002; 22 : 875–893. Jarrett RB, Eaves GG, Grannemann BD, Rush AJ. Clinical, cognitive, and demographic predictors of response to cognitive therapy for depression: A preliminary report. Psychiatry Res 1991; 37 : 245–260. Black DW, Wesner RB, Gabel J, Bowers W, Monahan P. Predictors of short-term treatment response in 66 patients with panic disorder. J Affect Disord 1994; 30 : 233–241. De Graaf LE, Hollon SD, Huibers MJH. Predicting outcome in computerized cognitive behavioral therapy for depression in primary care: A randomized trial. J Consult Clin Psychol 2010; 78 : 184–189. El Alaoui S, Hedman E, Ljótsson B, Lindefors N. Long-term effectiveness and outcome predictors of therapist-guided internet-based cognitive–behavioural therapy for social anxiety disorder in routine psychiatric care. BMJ Open 2015; 5 : e007902. Edmonds M, Hadjistavropoulos HD, Schneider LH, Dear BF, Titov N. Who benefits most from therapist-assisted internet-delivered cognitive behaviour therapy in clinical practice? Predictors of symptom change and dropout. J Anxiety Disord 2018; 54 : 24–32. Button KS, Wiles NJ, Lewis G, Peters TJ, Kessler D. Factors associated with differential response to online cognitive behavioural therapy. Soc Psychiatry Psychiatr Epidemiol 2012; 47 : 827–833. Hadjistavropoulos HD, Pugh NE, Hesser H, Andersson G. Predicting Response to Therapist-Assisted Internet-Delivered Cognitive Behavior Therapy for Depression or Anxiety Within an Open Dissemination Trial. Behav Ther 2016; 47 : 155–165. Nordgreen T, Havik OE, Öst LG, Furmark T, Carlbring P, Andersson G. Outcome predictors in guided and unguided self-help for social anxiety disorder. Behav Res Ther 2012; 50 : 13–21. Copay AG, Subach BR, Glassman SD, Polly DW, Schuler TC. Understanding the minimum clinically important difference: a review of concepts and methods. Spine J 2007; 7 : 541–546. Mululo SCC, De Menezes GB, Vigne P, Fontenelle LF. A review on predictors of treatment outcome in social anxiety disorder. Rev Bras Psiquiatr 2012; 34 : 92–100. Hedman E, Andersson E, Ljótsson B, Andersson G, Andersson E, Schalling M et al. Clinical and genetic outcome determinants of Internet- and group-based cognitive behavior therapy for social anxiety disorder: Predictors and moderators of CBT for SAD 1. Acta Psychiatr Scand 2012; 126 : 126–136. Chambless DL, Tran GQ, Glass CR. Predictors of Response to Cognitive-Behavioral Group Therapy for Social Phobia. J Anxiety Disord 1997; 11 : 221–240. Prasko J, Ociskova M, Latalova K, Kamaradova D, Grambal A. Psychological factors and treatment effectiveness in resistant anxiety disorders in highly comorbid inpatients. Neuropsychiatr Dis Treat 2016; Volume 12 : 1539–1551. El Alaoui S, Ljótsson B, Hedman E, Svanborg C, Kaldo V, Lindefors N. Predicting Outcome in Internet-Based Cognitive Behaviour Therapy for Major Depression: A Large Cohort Study of Adult Patients in Routine Psychiatric Care. PLOS ONE 2016; 11 : e0161191. Schat A, Van Noorden MS, Noom MJ, Giltay EJ, Van Der Wee NJA, Vermeiren RRJM et al. Predictors of outcome in outpatients with anxiety disorders: The Leiden routine outcome monitoring study. J Psychiatr Res 2013; 47 : 1876–1885. Kolek A, Prasko J, Ociskova M, Holubova M, Vanek J, Grambal A et al. Severity of panic disorder, adverse events in childhood, dissociation, self-stigma and comorbid personality disorders Part 2: Therapeutic effectiveness of a combined cognitive behavioural therapy and pharmacotherapy in treatment-resistant inpatients. Neuro Endocrinol Lett 2019; 40 : 271–283. Kamaradova D PJ. Therapeutic response to complex cognitive-behavioral and pharmacological treatment in patients with social phobia. Mausbach BT, Moore R, Roesch S, Cardenas V, Patterson TL. The Relationship Between Homework Compliance and Therapy Outcomes: An Updated Meta-Analysis. Cogn Ther Res 2010; 34 : 429–438. Bur OT, Bielinski LL, Krauss S, Häfliger A, Guggisberg J, Krieger T et al. Working alliance and adherence mediate the effect of guidance in a web-based program for participants with mild to moderate depressive symptoms: A secondary mediation analysis. Internet Interv 2022; 30 : 100593. Eley TC. The future of therapygenetics: where will studies predicting psychological treatment response from genomic markers lead?: The Cutting Edge: The Future of Therapygenetics. Depress Anxiety 2014; 31 : 617–620. Fusar-Poli L, Rutten BPF, Van Os J, Aguglia E, Guloksuz S. Polygenic risk scores for predicting outcomes and treatment response in psychiatry: hope or hype? Int Rev Psychiatry 2022; 34 : 663–675. Meerman JJ, Ter Hark SE, Janzing JGE, Coenen MJH. The Potential of Polygenic Risk Scores to Predict Antidepressant Treatment Response in Major Depression: A Systematic Review. J Affect Disord 2022; 304 : 1–11. Ward J, Graham N, Strawbridge RJ, Ferguson A, Jenkins G, Chen W et al. Polygenic risk scores for major depressive disorder and neuroticism as predictors of antidepressant response: Meta-analysis of three treatment cohorts. PLOS ONE 2018; 13 : e0203896. Andersson E, Crowley JJ, Lindefors N, Ljótsson B, Hedman-Lagerlöf E, Boberg J et al. Genetics of response to cognitive behavior therapy in adults with major depression: a preliminary report. Mol Psychiatry 2019; 24 : 484–490. Wallert J, Boberg J, Kaldo V, Mataix-Cols D, Flygare O, Crowley JJ et al. Predicting remission after internet-delivered psychotherapy in patients with depression using machine learning and multi-modal data. Transl Psychiatry 2022; 12 : 357. Rayner C, Coleman JRI, Purves KL, Hodsoll J, Goldsmith K, Alpers GW et al. A genome-wide association meta-analysis of prognostic outcomes following cognitive behavioural therapy in individuals with anxiety and depressive disorders. Transl Psychiatry 2019; 9 : 150. Coleman JRI, Lester KJ, Keers R, Roberts S, Curtis C, Arendt K et al. Genome-wide association study of response to cognitive–behavioural therapy in children with anxiety disorders. Br J Psychiatry 2016; 209 : 236–243. The MULTI-PSYCH cohort | Karolinska Institutet. https://ki.se/en/cns/the-multi-psych-cohort (accessed 4 Oct2023). Titov N, Dear B, Nielssen O, Staples L, Hadjistavropoulos H, Nugent M et al. ICBT in routine care: A descriptive analysis of successful clinics in five countries. Internet Interv 2018; 13 : 108–115. Fantino B, Moore N. The self-reported Montgomery-Åsberg depression rating scale is a useful evaluative tool in major depressive disorder. BMC Psychiatry 2009; 9 : 26. Houck PR, Spiegel DA, Shear MK, Rucci P. Reliability of the self-report version of the panic disorder severity scale. Depress Anxiety 2002; 15 : 183–185. Baker SL, Heinrichs N, Kim H-J, Hofmann SG. The Liebowitz social anxiety scale as a self-report instrument: a preliminary psychometric analysis. Behav Res Ther 2002; 40 : 701–715. Lahey BB, Krueger RF, Rathouz PJ, Waldman ID, Zald DH. A hierarchical causal taxonomy of psychopathology across the life span. Psychol Bull 2017; 143 : 142–186. Pettersson E, Larsson H, D’Onofrio BM, Bölte S, Lichtenstein P. The general factor of psychopathology: a comparison with the general factor of intelligence with respect to magnitude and predictive validity. World Psychiatry 2020; 19 : 206–213. Kotov R, Krueger RF, Watson D, Achenbach TM, Althoff RR, Bagby RM et al. The Hierarchical Taxonomy of Psychopathology (HiTOP): A dimensional alternative to traditional nosologies. J Abnorm Psychol 2017; 126 : 454–477. Zantvoort K, Hentati Isacsson N, Funk B, Kaldo V. Data set size vs homogeneity – A Machine Learning study on pooling intervention data in E-Mental Health dropout predictions. 2024 [submitted for publication] Mataix-Cols D, Fernández De La Cruz L, Monzani B, Rosenfield D, Andersson E, Pérez-Vigil A et al. D-Cycloserine Augmentation of Exposure-Based Cognitive Behavior Therapy for Anxiety, Obsessive-Compulsive, and Posttraumatic Stress Disorders: A Systematic Review and Meta-analysis of Individual Participant Data. JAMA Psychiatry 2017; 74 : 501. Lam M, Awasthi S, Watson HJ, Goldstein J, Panagiotaropoulou G, Trubetskoy V et al. RICOPILI: Rapid Imputation for COnsortias PIpeLIne. Bioinformatics 2020; 36 : 930–933. Boberg J, Kaldo V, Mataix-Cols D, Crowley JJ, Roelstraete B, Halvorsen M et al. Swedish multimodal cohort of patients with anxiety or depression treated with internet-delivered psychotherapy (MULTI-PSYCH). BMJ Open 2023; 13 : e069427. Ludvigsson JF, Almqvist C, Bonamy A-KE, Ljung R, Michaëlsson K, Neovius M et al. Registers of the Swedish total population and their use in medical research. Eur J Epidemiol 2016; 31 : 125–136. Stat. SCB. Labour statistics based on administrative sources. https://www.scb.se/en/finding-statistics/statistics-by-subject-area/labour-market/employment-and-working-hours/labour-statistics-based-on-administrative-sources/ (accessed 7 Mar2024). Statistics Sweden. Total population register. https://www.scb.se/contentassets/8f66bcf5abc34d0b98afa4fcbfc0e060/rtb-bar-2016-eng.pdf. Socialstyrelsen. National Patient Register. 2023.https://www.socialstyrelsen.se/en/statistics-and-data/registers/national-patient-register/ (accessed 7 Mar2024). VAL-databaserna – Region Stockholm. https://www.folkhalsokollen.se/datakallor/val-databaserna/ (accessed 7 Mar2024). Socialstyrelsen. National Prescribed Drug Register. 2024.https://www.socialstyrelsen.se/en/statistics-and-data/registers/national-prescribed-drug-register/ (accessed 7 Mar2024). Rubin DB. Inference and missing data. Biometrika 1976; 63 : 581–592. Bell ML, Fairclough DL. Practical and statistical issues in missing data for longitudinal patient-reported outcomes. Stat Methods Med Res 2014; 23 : 440–459. Moons KGM, Donders RART, Stijnen T, Harrell FE. Using the outcome for imputation of missing predictor values was preferred. J Clin Epidemiol 2006; 59 : 1092–1101. Van Buuren S. Flexible Imputation of Missing Data, Second Edition . 2nd ed. Chapman and Hall/CRC: Second edition. | Boca Raton, Florida : CRC Press, [2019] |, 2018 doi:10.1201/9780429492259. Sterne JAC, White IR, Carlin JB, Spratt M, Royston P, Kenward MG et al. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ 2009; 338 : b2393–b2393. Abayomi K, Gelman A, Levy M. Diagnostics for Multivariate Imputations. J R Stat Soc Ser C Appl Stat 2008; 57 : 273–291. Hernán MA, Hernández-Díaz S, Robins JM. A Structural Approach to Selection Bias: Epidemiology 2004; 15 : 615–625. Buuren SV, Groothuis-Oudshoorn K. mice : Multivariate Imputation by Chained Equations in R . J Stat Softw 2011; 45 . doi:10.18637/jss.v045.i03. Harrell FE. Regression Modeling Strategies: With Applications to Linear Models, Logistic Regression, and Survival Analysis . Springer New York: New York, NY, 2001 doi:10.1007/978-1-4757-3462-1. Spain D, Happé F. How to Optimise Cognitive Behaviour Therapy (CBT) for People with Autism Spectrum Disorders (ASD): A Delphi Study. J Ration-Emotive Cogn-Behav Ther 2020; 38 : 184–208. Menezes M, Harkins C, Robinson MF, Mazurek MO. Treatment of Depression in Individuals with Autism Spectrum Disorder: A Systematic Review. Res Autism Spectr Disord 2020; 78 : 101639. Eddy LD, Knouse LE, Safren SA. Adapting CBT for Treating Anxiety Disorders and Depression in Adults with ADHD. In: Todd G, Branch R (eds). Evidence-Based Treatment for Anxiety Disorders and Depression: A Cognitive Behavioral Therapy Compendium . Cambridge University Press: Cambridge, 2022, pp 553–572. Taylor S, Abramowitz JS, McKay D. Non-adherence and non-response in the treatment of anxiety disorders. J Anxiety Disord 2012; 26 : 583–589. Supplementary Information Supplementary Figure and Supplementary Tables are not available with this version. Additional Declarations The authors have declared there is NO conflict of interest to disclose Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4075444","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":278695784,"identity":"f0be4dbc-532e-4ca6-8029-4fee3aeb6a72","order_by":0,"name":"Olly Kravchenko","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA6klEQVRIiWNgGAWjYDACduYGBoYDIBYPA4ghJ8FMSAszI6oWY9K1JM4gpIOfmbFNguHMPXmD42cPPi44Y5M+s52B+cMHPFokm0FabhQbbjiTl2w840Za7mxmBjZJfFYZHAZp+ZDAOHMGj5k0z4fDufOAWph5iNBiD9Ri/huoJV2OmYH58x+CWm4kJPZL8Jgx89w4nCANDDFpfN4H+qXZIuFMQnI/T46xNM+ZNMOZQN9J9uDRws/efPDGh2MJtm3sZww/8xyzkZc4f/jwhx/4rGFgYJFIQBUAxRR+wIwvEkbBKBgFo2AUMDAAAGMRSYToaeNGAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0009-0007-9482-8739","institution":"Karolinska Institutet","correspondingAuthor":true,"prefix":"","firstName":"Olly","middleName":"","lastName":"Kravchenko","suffix":""},{"id":278695785,"identity":"30d3896e-2de9-4ba5-8342-8372144f8f38","order_by":1,"name":"Julia Boberg","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Julia","middleName":"","lastName":"Boberg","suffix":""},{"id":278695786,"identity":"8695a25d-a771-4dea-85bc-9108b9a745a2","order_by":2,"name":"David Mataix-Cols","email":"","orcid":"https://orcid.org/0000-0002-4545-0924","institution":"Karolinska Institutet","correspondingAuthor":false,"prefix":"","firstName":"David","middleName":"","lastName":"Mataix-Cols","suffix":""},{"id":278695787,"identity":"01a177d7-7718-470d-b9bb-f734cf0de391","order_by":3,"name":"James Crowley","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"James","middleName":"","lastName":"Crowley","suffix":""},{"id":278695788,"identity":"9175a239-e30e-4d6d-a91a-9b057d047425","order_by":4,"name":"Matthew Halvorsen","email":"","orcid":"https://orcid.org/0000-0002-6707-2418","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Matthew","middleName":"","lastName":"Halvorsen","suffix":""},{"id":278695789,"identity":"0a52e8c8-8c35-4d4c-9bf0-f5106f51fbaf","order_by":5,"name":"Patrick Sullivan","email":"","orcid":"https://orcid.org/0000-0002-6619-873X","institution":"University of North Carolina","correspondingAuthor":false,"prefix":"","firstName":"Patrick","middleName":"","lastName":"Sullivan","suffix":""},{"id":278695790,"identity":"a8ae8882-8393-4ad1-8a45-7491f8602930","order_by":6,"name":"John Wallert","email":"","orcid":"https://orcid.org/0000-0002-1473-4916","institution":"Karolinska Institutet","correspondingAuthor":false,"prefix":"","firstName":"John","middleName":"","lastName":"Wallert","suffix":""},{"id":278695791,"identity":"7358ee66-8b2f-4626-92a8-4fa49101f956","order_by":7,"name":"Christian Rück","email":"","orcid":"https://orcid.org/0000-0002-8742-0168","institution":"Karolinska Institutet","correspondingAuthor":false,"prefix":"","firstName":"Christian","middleName":"","lastName":"Rück","suffix":""}],"badges":[],"createdAt":"2024-03-11 14:11:37","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4075444/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4075444/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":53748489,"identity":"5581f05e-de48-4c69-95bc-6f4fadc9b101","added_by":"auto","created_at":"2024-03-29 18:23:59","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":91525,"visible":true,"origin":"","legend":"\u003cp\u003eCompared variance explained by baseline and full model by predictor type\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-4075444/v1/83c7cd632eab71b0daeb0037.jpeg"},{"id":64739635,"identity":"c4881e02-33a2-4121-bbe5-10321806bb8a","added_by":"auto","created_at":"2024-09-18 08:40:47","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":976039,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4075444/v1/55047320-0890-46dd-8a93-c926c4582eac.pdf"}],"financialInterests":"The authors have declared there is \u003cb\u003eNO\u003c/b\u003e conflict of interest to disclose","formattedTitle":"Clinical, genetic, and sociodemographic predictors of symptom severity after internet-delivered cognitive behavioural therapy for depression and anxiety","fulltext":[{"header":"Introduction","content":"\u003cp\u003eMajor depressive disorder (MDD), panic disorder (PD), and social anxiety disorder (SAD) are highly prevalent mental health disorders\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. MDD ranks as the first and anxiety disorders as the sixth leading cause of disability globally, measured by years lived with disability\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Cognitive behavioural therapy (CBT) is an effective first-line psychotherapeutic treatment for mild to moderate depression and anxiety disorders\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. To improve accessibility, CBT is increasingly delivered in a therapist-guided internet-based format (ICBT) which has shown comparable efficacy\u003csup\u003e\u003cspan additionalcitationids=\"CR6 CR7 CR8 CR9\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e, superior accessibility for some patients\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e, and a high probability of cost-effectiveness\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. However, CBT/ICBT is not universally efficacious, with up to 50% of patients not responding sufficiently to the treatment\u003csup\u003e\u003cspan additionalcitationids=\"CR14\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Therefore, it is crucial to identify reliable predictors of unfavourable treatment outcome and develop accurate predictive models that would assist clinicians in selecting appropriate, tailored care for these patients.\u003c/p\u003e \u003cp\u003eNumerous clinical characteristics have been investigated as potential predictors of differential response to CBT. Baseline symptom severity is consistently shown to be the most robust predictor, but the direction of this association is unclear. Some studies suggest that a higher baseline symptom level is predictive of a poorer prognosis following treatment\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan additionalcitationids=\"CR17 CR18 CR19\" citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. By contrast, others report greater/faster improvement in this patient group\u003csup\u003e\u003cspan additionalcitationids=\"CR22 CR23 CR24\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e, which may be explained by regression toward the mean (i.e., extreme values sampled from a random variable are likely to be closer to its mean with repeated sampling)\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. Comorbidity with other mood and anxiety disorders\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan additionalcitationids=\"CR28\" citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e as well as personality disorders\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e,\u003cspan additionalcitationids=\"CR31\" citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e, early onset of symptoms\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e,\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e,\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e, and family history of psychopathology\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e have also been identified as potential predictors of poorer response to CBT. Furthermore, several treatment-related variables have been linked to the outcome, such as treatment expectancy\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e, adherence\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e, homework completion\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e, and working alliance\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. Lastly, among sociodemographic patient characteristics, being employed\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e,\u003cspan additionalcitationids=\"CR32\" citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e, married\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e, and having higher education\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e are predictive of a lower post-treatment symptom severity.\u003c/p\u003e \u003cp\u003eGenetic differences are implicated in a wide range of individual variation in human complex traits, including differential susceptibility and response to environmental stimuli. A new area known as therapygenetics views therapeutic interventions as such a stimulus, providing the basis for the hypothesis that genetic variation could, at least partially, explain treatment effect heterogeneity\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. Several studies suggest that polygenic risk scores (PRS) can be used to predict response to pharmacological treatments of psychiatric disorders\u003csup\u003e\u003cspan additionalcitationids=\"CR39\" citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. PRS are constructed as a weighted sum of risk alleles of single nucleotide polymorphisms (SNPs) using summary statistics from genome-wide association studies (GWAS). Preliminary evidence suggests a link between a higher polygenic loading for autism spectrum disorder (PRS ASD) and a poorer treatment response in patients undergoing ICBT for MDD\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. In another study using the same data as\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e with machine learning modelling, higher PRS MDD and PRS for intelligence (PRS IQ) weakly predicted remission in ICBT for MDD\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. However, the largest therapygenetic GWA meta-analysis to date (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2724) did not identify any common genetic variants associated with the CBT treatment outcome for depression and anxiety\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e, consistent with a previous GWAS that had a similar sample size and was also likely underpowered\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eDespite the abundance of literature on predictors of treatment outcome, there are some limitations that this study aimed to address. Most existing studies are based on a relatively small number of clinical predictors that are derived from secondary data analysis of randomized controlled trials (RCT), which are costly to conduct and rely on small samples of participants who are not representative of the broader patient population, thus providing inconsistent evidence on predictor importance. We contribute to the state of the field by leveraging a recently constructed MULTI-PSYCH cohort (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2668) with a broad selection of potential predictors, integrating data routinely collected at the clinic with additional inputs such as PRS and a broad range of variables from Swedish population registries\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. The aim of the present study was twofold: to identify patient characteristics predictive of higher post-treatment symptom severity and to evaluate whether incorporating genetic and register predictors alongside established clinical predictors explains additional variance in post-treatment symptom severity.\u003c/p\u003e"},{"header":"Patients and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eSample\u003c/h2\u003e \u003cp\u003eThe MULTI-PSYCH cohort study comprises 2668 adult outpatients treated with 12 weeks of ICBT for MDD (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1300), PD (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;727) or SAD (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;641) at the Internet psychiatry unit of the Psychiatric Clinic Southwest at Karolinska University Hospital Huddinge, Sweden, between 2008 and 2020\u003csup\u003e45\u003c/sup\u003e. Screening and treatment procedures are described in more detail elsewhere\u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e. As part of the assessment, patients completed an online questionnaire and were interviewed by a licensed clinician (psychologist or psychiatrist), providing sociodemographic information along with details about their condition and medical history. Blood samples were collected for DNA extraction before treatment, and data from multiple nationwide registers were linked using each patient\u0026rsquo;s unique Swedish personal number.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eVariables\u003c/h2\u003e \u003cp\u003eCandidate predictors were divided into three groups based on their source: clinic-based (obtained as part of pre-treatment questionnaires and interviews, containing medical and sociodemographic data), genetic (PRS for seven traits derived from GWAS summary statistics and calculated for all genotyped patients), and register-based (sourced through record linkage across multiple national registers containing medical, socioeconomic, and demographic data). This division is helpful for assessing the relative contribution of predictors derived from different sources: genetic and register data are not readily available in the routine care setting, and thus, their inclusion must be justified by substantial added explanatory power.\u003c/p\u003e \u003cp\u003eThe purpose of this study was to identify predictors that have prognostic value, and thus, selected variables must be available pre-treatment. Process variables, e.g., adherence and homework completion, were therefore excluded. Clinic-based predictors were pre-selected based on literature review and established evidence to minimize predictor selection bias. Studies were eligible for inclusion if conducted in adults treated with internet-delivered or face-to-face CBT for MDD, PD or SAD. See Supplementary Information, S1 Table for a list of established predictors. By contrast, register data are rarely accessible for research purposes, leading to a scarcity of literature on established predictors. Thus, primary selection of register-based variables was guided by their face validity and subsequently validated through bivariate regression analyses. Those predictors that were significantly associated with the outcome were added to the final model.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eClinic-based predictors\u003c/h2\u003e \u003cp\u003eBaseline symptom severity was measured right before ICBT start with the Montgomery\u0026ndash;\u0026Aring;sberg Depression Rating Scale, self-rating version (MADRS-S) for MDD (test\u0026ndash;retest reliability: \u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.78; internal consistency reliability: α\u0026thinsp;=\u0026thinsp;0.84)\u003csup\u003e47\u003c/sup\u003e, Panic Disorder Severity Scale, self-report (PDSS-SR) for PD (test\u0026ndash;retest reliability: \u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.81; internal consistency reliability: α\u0026thinsp;=\u0026thinsp;0.92)\u003csup\u003e48\u003c/sup\u003e, and Liebowitz Social Anxiety Scale, self-report (LSAS-SR) for SAD (test\u0026ndash;retest reliability: \u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.83; internal consistency reliability: α\u0026thinsp;=\u0026thinsp;0.95)\u003csup\u003e49\u003c/sup\u003e. In the present study, the three disorders along with their respective predictors and outcome measures were combined to increase statistical power. While this decision comes at the cost of introducing more heterogeneity, we deemed it conceptually reasonable given the magnitude of phenotypic and etiological overlap between MDD, PD, and SAD, the similarities in the three ICBT treatment protocols, and the growing support for applying transdiagnostic models of psychopathology\u003csup\u003e\u003cspan additionalcitationids=\"CR51\" citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e. Finally, a recent machine-learning study aimed at investigating the value of pooling intervention data for MDD, PD, and SAD into a single dataset and using a larger sample derived from the same source as the current study supported the superiority of this approach\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e. The disorder-specific scales were transformed to a harmonized 0-100 range and subsequently pooled into a single score\u003csup\u003e\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e (see Supplementary Information).\u003c/p\u003e \u003cp\u003eComorbidity is a binary measure that indicates the presence of coexisting psychiatric disorders as assigned by a clinician during a screening interview. Family history captures self-reported confirmed or suspected cases of psychopathology among first-degree relatives selected from a list of 36 options. Sex was derived from the Swedish personal identity number and therefore refers to biological sex. Highest achieved level of education was self-reported at screening and for statistical analysis dichotomised into \u003cem\u003eUniversity\u003c/em\u003e (finished or unfinished) and \u003cem\u003eNo university\u003c/em\u003e (up to upper secondary school). Marital status is a self-reported civil status dichotomised into \u003cem\u003eMarried\u003c/em\u003e (officially married or de facto married, i.e., in a committed relationship (\u0026lsquo;married/living together/living apart together\u0026rsquo;)) and \u003cem\u003eUnmarried (Single\u003c/em\u003e, \u003cem\u003eSeparated\u003c/em\u003e, and \u003cem\u003eWidowed).\u003c/em\u003e Finally, Parental status was retrieved from the screening question, \"Do you have children?\".\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eGenetic predictors\u003c/h2\u003e \u003cp\u003eGenotyping was performed in three batches at LIFE \u0026amp; BRAIN GmbH in Bonn, Germany, with Illumina HumanCoreExome-12 v1.0, Infinium Global Screening Array-24 v1.0, and Infinium Global Screening Array-24 v2.0, respectively. Quality control and imputation were performed via the RICOPILI GWAS pipeline\u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e. PRS for seven traits (MDD, attention deficit hyperactivity disorder (ADHD), ASD, bipolar disorder (BPAD), schizophrenia (SCZ), IQ, and educational attainment (EA)) were constructed using the PRS-CS method. A detailed description of genetic data pre-processing can be found elsewhere\u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eRegister-based predictors\u003c/h2\u003e \u003cp\u003eThe Swedish national population registration system provides high-quality data on important life events for all individuals who have been assigned a personal identity number and is unique in its almost complete coverage\u003csup\u003e\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eRegister-based employment status was derived from the Register-based Employment Statistics (RAMS)\u003csup\u003e\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e and has been dichotomised into \u003cem\u003eEmployed\u003c/em\u003e/\u003cem\u003eUnemployed\u003c/em\u003e. Income was obtained from the Total Population Register\u003csup\u003e\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e and is defined as annual disposable household income divided by consumption weight of the family, thus constituting an individual\u0026rsquo;s component of the household income adjusted for family composition. For statistical analyses, Income was divided into quintiles. Lastly, Financial benefits is receipt of any of eight benefits from the Swedish social insurance system (for a complete list, see Supplementary Information). Statistics on employment, income, and financial benefits are updated annually; hence, the most recent available record from the year preceding treatment was utilized.\u003c/p\u003e \u003cp\u003ePrior psychiatric diagnoses were derived from the National Patient Register (NPR)\u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e and the Stockholm regional healthcare data warehouse VAL (V\u0026aring;rdanalysdatabasen)\u003csup\u003e\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e\u003c/sup\u003e. NPR comprises data on specialized in- and outpatient care visits in Sweden, while VAL registers all primary care visits in Stockholm County, where most of the patients reside. In addition to ICD-10 Chap.\u0026nbsp;5 codes (e.g. F3: Mood [affective] disorders), a separate variable was constructed for BPAD to avoid conflation with depressive disorders which are the target in the studied ICBT treatment. Separate variables were also created for ASD and ADHD and, given their neurodevelopmental nature and persistent lifelong manifestation, the inclusion was not limited to diagnoses received prior to ICBT treatment. To commence the ICBT treatment, a patient must receive a relevant diagnosis, either from a general practitioner or a clinician at the Internet Psychiatry clinic, which is subsequently transferred to the NPR. Hence, to avoid capturing the diagnosis related to this very healthcare episode and only focus on the \u003cem\u003eprior\u003c/em\u003e events, all records of MDD/PD/SAD within a month preceding the ICBT treatment were excluded from the analysis. Data on dispensation of prescription medicines were retrieved from the National Prescribed Drug Register\u003csup\u003e\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e\u003c/sup\u003e and included Antidepressants (N06A), Anxiolytics (N05B), Hypnotics and sedatives (N05C), and Antipsychotics (N05A). As with the psychiatric diagnoses, all records of dispensation within a month preceding ICBT treatment were excluded from the analysis.\u003c/p\u003e \u003cp\u003eBinary variables for Education (\u003cem\u003eUniversity\u003c/em\u003e/\u003cem\u003eNo university\u003c/em\u003e), Marital status (\u003cem\u003eMarried\u003c/em\u003e/\u003cem\u003eUnmarried\u003c/em\u003e), and Parental status (\u003cem\u003eChildren\u003c/em\u003e/\u003cem\u003eNo children\u003c/em\u003e) were constructed from both clinic-based and register-based data. The rationale for using the clinic-based measure in the full model, as well as a detailed description of data pre-processing, is provided in Supplementary Information.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eOutcome\u003c/h2\u003e \u003cp\u003eTreatment outcome is defined as post-treatment symptom level, measured by a self-rated, disorder-specific scale. Identically to the baseline symptom severity, harmonized scores for the three disorders were pooled.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cdiv id=\"Sec10\" class=\"Section3\"\u003e \u003ch2\u003eMissing data\u003c/h2\u003e \u003cp\u003eSummary of missing data is presented in Supplementary Information, S2 Table. Clinic-based predictors with over 30% missing values were excluded. When available, missing pre-treatment symptom severity values (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;30, 1.1%) were replaced with those collected at screening, resulting in three missing values in the final dataset (0.1%). For all the PRS, 456 (17.1%) observations were missing due to being excluded during quality control procedures (poor matching and partial non-European ancestry). Among register-based predictors, there were six missing values in Employment, Income, and Financial benefits.\u003c/p\u003e \u003cp\u003eThe outcome variable, post-treatment symptom severity, had 524 missing observations (19.6%). The missingness mechanism was hypothesized to be missing at random (MAR), i.e., the missingness is conditional on other variables in the dataset, and the sensitivity analysis was performed to check the robustness against departures from this assumption (see the results in Supplementary Information). The missingness pattern suggested that non-completers were more likely to have higher post-treatment symptom values since the variables associated with missingness were also predictive of higher outcome values in the non-missing subset. Thus, listwise deletion (complete-case analysis) was ruled out for two reasons: it negatively affects the precision of estimates by decreasing power and, more importantly, if missing data are not MCAR, it is likely to introduce selection bias into inferences and cause too conservative estimates\u003csup\u003e\u003cspan additionalcitationids=\"CR64 CR65 CR66 CR67 CR68\" citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eTo address the uncertainty introduced by missingness, missing values were imputed with multiple imputation using the R package \u003cem\u003emice\u003c/em\u003e\u003csup\u003e\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e\u003c/sup\u003e. A total of \u003cem\u003em\u003c/em\u003e\u0026thinsp;=\u0026thinsp;20 datasets were imputed over 20 iterations. In accordance with the recommended procedure, 25 auxiliary variables not included in the analytical sample were added to the imputation dataset to improve the quality of imputations through increasing the plausibility of the MAR assumption\u003csup\u003e\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e\u003c/sup\u003e. Missing values were imputed using the default settings of the \u003cem\u003emice\u003c/em\u003e package. For numeric data, predictive mean matching (\u003cem\u003epmm\u003c/em\u003e) was used as the imputation method, replacing a missing value with a randomly selected observed donor that has a regression-predicted value closest to the missing value based on a simulated regression model. It is a superior method to regression in that it does not rely on the assumption of joint multivariate normal distribution and creates realistic imputed values. For binary data, logistic regression imputation (\u003cem\u003elogreg\u003c/em\u003e) was used and for unordered categorical data, polytomous regression imputation (\u003cem\u003epolyreg\u003c/em\u003e). To check the adequacy of the imputation models and assess convergence, both graphic and numeric diagnostic methods were employed, namely kernel density plots of the distributions of the observed and \u003cem\u003em\u003c/em\u003e\u0026thinsp;=\u0026thinsp;20 imputed datasets and the Kolmogorov-Smirnov test to assess departures from the assumptions made in the imputation model\u003csup\u003e\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eMain analysis\u003c/h2\u003e \u003cp\u003eAll statistical analyses were performed using R (version 4.3.1). First, univariate analyses were performed on all predictors to examine their distribution and properties. Next, two models were developed: a baseline model (clinic-based data only) and a full model (clinic-based, genetic, and register data). The same steps of statistical analysis were repeated for both models: first, separate bivariate linear regression models were fit to each predictor; next, all covariates were entered simultaneously into the multiple linear regression model. The purpose of conducting the analysis in two steps was dictated by the research questions. Given the complex nature of the studied phenomenon and the known associations between the covariates from existing literature on latent human traits, at least partial collinearity is inevitable. Thus, to evaluate independent associations between individual predictors and the outcome, assessment of simple regression coefficients is necessary to avoid misinterpreting small effect sizes, whereby a variable important in a bivariate model might be attenuated in a multiple model by other collinear predictors. For the second goal of the overall model appraisal, the accuracy of point estimates is not paramount. Therefore, variable selection was theory-driven, and all pre-selected variables were retained in the multiple regression model irrespective of their statistical significance. This was done to avoid the flaws related to the algorithm of stepwise variable selection whereby variables are added and removed based on their significance level\u003csup\u003e\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eA total of 45 predictors were used in the full model. All variables with zero variance were excluded from the analysis. Variables with very low variance (meeting both criteria: frequency of the second most common value of \u0026lt;\u0026thinsp;1% over the sample and percentage of unique values of \u0026lt;\u0026thinsp;10% of the total number of data points) (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;46) were excluded as independent predictors but retained within relevant composite variables. For a full list of included and excluded variables, see Supplementary Information. For prior diagnoses and medication, both composite and individual components were assessed through bivariate analyses for the sake of potentially discovering novel predictors. For example, using a composite variable that encompasses any prior psychotropic medication may be justified in the interest of model parsimony, but it may miss what specific medication drives the effect size of the association. However, only individual predictors were used in multiple regression analysis to avoid multicollinearity.\u003c/p\u003e \u003cp\u003eLastly, the baseline and the full model were evaluated based on their goodness of fit using adjusted coefficients of determination (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({R}_{adj}^{2}\\)\u003c/span\u003e\u003c/span\u003e). All additional covariates are likely to increase \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({R}^{2}\\)\u003c/span\u003e\u003c/span\u003e, regardless of their true explanatory power. In contrast, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({R}_{adj}^{2}\\)\u003c/span\u003e\u003c/span\u003e imposes a penalty for this inflation when comparing models with a varying number of predictors. Consequently, the assessment of the full model performance compared to the baseline model relied on the extra variance explained by additional predictors and was complemented with the appraisal of the Akaike information criterion (AIC) and root-mean-square error (RMSE) of the two models.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\n\u003ch2\u003eDescriptive statistics\u003c/h2\u003e\n\u003cp\u003eSample characteristics are presented in Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e (for more details, see Supplementary Information, S4 Table). Descriptive statistics for the observed, imputed, and complete dataset, as well as the distributional discrepancy between observed and imputed outcome values, are provided in Supplementary Information, S5 Table and Supplementary Information, Figure S1, respectively.\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n\u003ctable id=\"Tab1\" border=\"1\"\u003e\u003ccaption\u003e\n\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\n\u003cdiv class=\"CaptionContent\"\u003e\n\u003cp\u003eDescriptive statistics for the total sample and stratified by disorder\u003c/p\u003e\n\u003c/div\u003e\n\u003c/caption\u003e\n\u003cthead\u003e\n\u003ctr\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eVariable\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eTotal\u003c/p\u003e\n\u003cp\u003e(\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2668)\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eMDD\u003c/p\u003e\n\u003cp\u003e(\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1300)\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003ePD\u003c/p\u003e\n\u003cp\u003e(\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;727)\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eSAD\u003c/p\u003e\n\u003cp\u003e(\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;641)\u003c/p\u003e\n\u003c/th\u003e\n\u003c/tr\u003e\n\u003c/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePre-treatment symptom severity, mean (SD)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e47.6 (17.8)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e51.6 (15.7)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e40.2 (18.8)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e47.7 (17.9)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePost-treatment symptom severity, mean (SD)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e28.2 (18.8)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e30.3 (18.7)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e17.8 (16.1)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e35.3 (17.2)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eRelative symptom change, % (SD)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-41.3 (37.9)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-41.7 (34.3)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-53.4 (47.0)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-28.1 (26.5)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eAge, mean (SD)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e35.6 (11.4)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e37.6 (11.9)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e34.6 (10.8)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e32.7 (10.3)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eMale sex, \u003cem\u003en\u003c/em\u003e (%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1014 (38.0%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e443 (34.1%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e292 (40.2%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e279 (43.6%)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePsychiatric comorbidities, \u003cem\u003en\u003c/em\u003e (%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e862 (33.5%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e394 (31.4%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e265 (37.9%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e203 (32.9%)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eFamily history of psychopathology, \u003cem\u003en\u003c/em\u003e (%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1812 (70.6%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e888 (71%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e488 (69.7%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e436 (70.7%)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eSelf-reported married, \u003cem\u003en\u003c/em\u003e (%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1566 (58.8%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e727 (56.1%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e477 (65.7%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e362 (56.7%)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eSelf-reported having children, \u003cem\u003en\u003c/em\u003e (%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1102 (41.4%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e603 (46.4%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e302 (41.5%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e199 (31.0%)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eSelf-reported university education, \u003cem\u003en\u003c/em\u003e (%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1693 (63.6%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e890 (68.8%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e406 (55.9%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e397 (62.1%)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eRegister-based employed, \u003cem\u003en\u003c/em\u003e (%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2448 (92.0%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1205 (92.7%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e674 (92.8%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e571 (89.1%)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eAnnual income (SEK), mean (SD)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e241,713 (201,190)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e256,187 (202,155)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e244,258 (232,968)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e209,591 (150,742)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eFinancial benefits in the past year, \u003cem\u003en\u003c/em\u003e (%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e556 (20.9%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e302 (23.3%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e148 (20.4%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e106 (16.5%)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePrior psychiatric diagnosis, \u003cem\u003en\u003c/em\u003e (%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1793 (67.2%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e895 (68.8%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e518 (71.3%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e380 (59.3%)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePrior psychotropic medication, \u003cem\u003en\u003c/em\u003e (%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1665 (62.4%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e863 (66.4%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e472 (64.9%)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e330 (51.5%)\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003c/tbody\u003e\n\u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cem\u003eSD, standard deviation; SEK, Swedish krona; MDD, Major depressive disorder; PD, Panic disorder; SAD, Social anxiety disorder\u003c/em\u003e\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\n\u003ch2\u003ePredictor importance\u003c/h2\u003e\n\u003cp\u003eParameter estimates for clinic-based, genetic, and register-based predictors are displayed in Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n\u003ctable id=\"Tab2\" border=\"1\"\u003e\u003ccaption\u003e\n\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\n\u003cdiv class=\"CaptionContent\"\u003e\n\u003cp\u003eAssociation between clinic-based, genetic, and register-based predictors and treatment outcome. Results from bivariate and multiple regression analyses in the full model\u003c/p\u003e\n\u003c/div\u003e\n\u003c/caption\u003e\n\u003cthead\u003e\n\u003ctr\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003ePredictor\u003c/p\u003e\n\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003eBivariate\u003c/p\u003e\n\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003eMultiple\u003c/p\u003e\n\u003c/th\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eEstimate (95% CI)\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003e\u003cem\u003ep\u003c/em\u003e-value\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eEstimate (95% CI)\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003e\u003cem\u003ep\u003c/em\u003e-value\u003c/p\u003e\n\u003c/th\u003e\n\u003c/tr\u003e\n\u003c/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePre-treatment score\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.57 (0.53, 0.61)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.51 (0.47, 0.55)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eComorbidities\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e5.20 (3.54, 6.86)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.42 (-1.03, 1.88)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.569\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eFamily history\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e2.75 (0.92, 4.58)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.003**\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.19 (-1.41, 1.79)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.816\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eSex (ref: male)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.58 (-1.06, 2.23)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.488\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e-2.56 (-4.03, -1.09)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eUnmarried\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e4.62 (2.89, 6.34)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e1.67 (0.08, 3.26)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.040*\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eNo children\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e2.28 (0.72, 3.85)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.004**\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.72 (-0.84, 2.28)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.364\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eNo university\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e4.18 (2.52, 5.85)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e1.95 (0.49, 3.42)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.009**\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eUnemployed\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e7.45 (4.15, 10.75)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e2.98 (-0.01, 5.97)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.051\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eIncome (quintiles)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e-1.90 (-2.47, -1.34)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e-0.81 (-1.38, -0.25)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.005**\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eAny financial benefits\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e6.24 (4.23, 8.25)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e2.24 (0.37, 4.11)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.019*\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePRS MDD\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.71 (-0.19, 1.60)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.120\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.41 (-0.39, 1.21)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.310\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePRS ASD\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.58 (-0.37, 1.53)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.228\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.43 (-0.45, 1.32)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.335\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePRS ADHD\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.68 (-0.33, 1.69)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.182\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.14 (-0.71, 0.99)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.741\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePRS BPAD\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e-0.21 (-1.07, 0.66)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.640\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e-0.73 (-1.57, 0.12)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.091\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePRS Education\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e-0.30 (-1.24, 0.64)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.527\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.24 (-0.73, 1.20)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.630\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePRS IQ\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.37 (-0.59, 1.33)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.444\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.38 (-0.58, 1.34)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.434\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePRS SCZ\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.28 (-0.63, 1.19)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.544\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.46 (-0.45, 1.38)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.315\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eAny prior psychiatric diagnosis\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e5.54 (3.88, 7.20)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePrior F1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e3.16 (-0.07, 6.39)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.055\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e-1.76 (-4.63, 1.11)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.230\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePrior F3 (excl. BPAD)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e7.61 (5.89, 9.34)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e2.29 (0.56, 4.02)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.010*\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePrior F4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e3.29 (1.69, 4.90)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.93 (-0.61, 2.48)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.235\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePrior F5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e7.84 (5.05, 10.64)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e2.70 (0.11, 5.30)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.041*\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePrior F6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e15.70 (9.35, 22.06)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e5.55 (-0.11, 11.20)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.055\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eASD\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e24.98 (18.35, 31.61)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e10.35 (4.17, 16.53)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.001**\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eADHD\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e15.98 (12.24, 19.71)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e6.38 (2.91, 9.85)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eAny prior medication\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e4.52 (2.89, 6.16)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eAntidepressants\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e6.30 (4.66, 7.94)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eBupropion\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e14.16 (9.75, 18.58)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e6.04 (1.85, 10.24)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.005**\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eDuloxetine\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e10.84 (6.07, 15.61)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e-0.49 (-4.87, 3.90)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.827\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eFluoxetine\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e10.52 (6.97, 14.07)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e4.18 (0.91, 7.44)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.012*\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eVenlafaxine\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e9.44 (5.77, 13.10)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e2.78 (-0.62, 6.18)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.108\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eMirtazapine\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e8.31 (5.12, 11.50)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.48 (-2.62, 3.59)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.759\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eAmitriptyline\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e7.93 (2.89, 12.96)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.002**\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e4.04 (-0.27, 8.35)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.066\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eSertraline\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e6.51 (4.56, 8.47)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e2.70 (0.88, 4.52)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.004**\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eEscitalopram\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e6.42 (3.63, 9.21)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e1.77 (-0.72, 4.27)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.163\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eCitalopram\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e3.74 (1.60, 5.87)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e2.20 (0.22, 4.18)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.030*\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eClomipramine\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e1.71 (-5.96, 9.38)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.662\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e-2.16 (-8.73, 4.41)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.518\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eParoxetine\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e-1.49 (-5.97, 2.98)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.512\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e-0.09 (-3.95, 3.77)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.963\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eAnxiolytics\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e3.14 (1.40, 4.89)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eBuspirone\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e19.41 (11.63, 27.20)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e7.73 (0.61, 14.84)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.033*\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eAlprazolam\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e4.00 (-1.60, 9.59)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.161\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e1.70 (-3.23, 6.62)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.498\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eHydroxyzine\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e3.16 (1.26, 5.05)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.001**\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e-1.78 (-3.72, 0.15)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.070\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eDiazepam\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e2.59 (-1.36, 6.55)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.198\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e-1.15 (-4.60, 2.29)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.512\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eOxazepam\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e2.25 (-0.02, 4.53)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.052\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e-0.84 (-3.00, 1.32)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.444\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eHypnotics and sedatives\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e6.20 (4.40, 7.99)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eMelatonin\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e11.02 (5.74, 16.30)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e2.56 (-2.31, 7.42)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.301\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eZolpidem\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e6.23 (3.46, 9.01)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.91 (-1.73, 3.56)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.498\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eZopiclone\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e5.79 (3.40, 8.18)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.03 (-2.49, 2.43)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.984\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003ePropiomazine\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e5.18 (2.80, 7.56)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;.001***\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e-1.56 (-3.91, 0.79)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.193\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eAntipsychotics\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e5.98 (1.37, 10.59)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.011*\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e-3.63 (-7.79, 0.53)\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e.087\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003c/tbody\u003e\n\u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cem\u003ePRS, polygenic risk score; MDD, Major depressive disorder; ASD, Autism spectrum disorder; ADHD, Attention deficit hyperactivity disorder; BPAD, Bipolar affective disorder; IQ, intelligence quotient; SCZ, Schizophrenia; F1, Mental and behavioural disorders due to psychoactive substance use; F3, Mood [affective] disorders; F4, Anxiety, dissociative, stress-related, somatoform and other nonpsychotic mental disorders; F5, Behavioural syndromes associated with physiological disturbances and physical factors; F6, Disorders of adult personality and behaviour; CI, confidence interval\u003c/em\u003e\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\n\u003cp\u003e\u003cem\u003eSignificance levels: *p\u0026thinsp;\u0026lt;\u0026thinsp;.05, **p\u0026thinsp;\u0026lt;\u0026thinsp;.01, ***p\u0026thinsp;\u0026lt;\u0026thinsp;.001\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAs expected, pre-treatment symptom level was a strong predictor of post-treatment score severity (\u0026beta;\u0026thinsp;=\u0026thinsp;0.57, 95% CI [0.53, 0.61], \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001), remaining almost unattenuated in the adjusted model. No significant association between PRS and the outcome was observed. Socioeconomic predictors (no university education, unemployment, receipt of financial benefits, and lower income quintile) were strongly associated with higher post-treatment symptom severity and remained significant in the multiple regression model. Being childless was associated with a higher post-treatment score in the bivariate model, but the effect disappeared in the adjusted model. By contrast, female sex showed association with lower post-treatment score in the adjusted model only.\u003c/p\u003e\n\u003cp\u003eHaving a history of psychiatric diagnoses predicted higher post-treatment symptom severity. In addition to previous records of depressive and anxiety disorders, eating and personality disorders were strongly associated with poorer treatment outcome. A particularly large effect sizes were found for comorbid ASD (\u0026beta;\u0026thinsp;=\u0026thinsp;24.98, 95% CI [18.35, 31.61], \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001) and ADHD (\u0026beta;\u0026thinsp;=\u0026thinsp;15.98, 95% CI [12.24, 19.71], \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001), the observed relationship persisted in the adjusted model.\u003c/p\u003e\n\u003cp\u003eThe other significant predictor was prior use of almost all psychotropic medications. In addition to the substantial effect of most antidepressants and some anxiolytics (particularly buspirone), it is worth emphasizing quite a strong association with the prior use of hypnotics and sedatives (with a notably large effect of melatonin) as well as antipsychotics.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\n\u003ch2\u003eModel performance\u003c/h2\u003e\n\u003cp\u003eEstimates of comparative performance of the models are presented in Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e and visualized in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e. \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({R}_{adj}^{2}\\)\u003c/span\u003e\u003c/span\u003e of the baseline model was .27, thus explaining 27% of the variance in post-treatment symptom severity, which is almost entirely driven by pre-treatment symptom severity. The full model yielded \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({R}_{adj}^{2}\\)\u003c/span\u003e\u003c/span\u003e of .34. Addition of PRS did not contribute to the model performance, increasing \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({R}_{adj}^{2}\\)\u003c/span\u003e\u003c/span\u003e of the full model by only 0.2 percentage points above clinic-based and register-based predictors. The full model was also superior when comparing the two models\u0026rsquo; AIC and RMSE. These results suggest that including rich register data provided some additional explanatory power beyond the clinic-based predictors used in the baseline model, accounting for a further 7% of the variance.\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n\u003ctable id=\"Tab3\" border=\"1\"\u003e\u003ccaption\u003e\n\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\n\u003cdiv class=\"CaptionContent\"\u003e\n\u003cp\u003eAdjusted R-squared (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\varvec{R}}_{\\varvec{a}\\varvec{d}\\varvec{j}}^{2}\\)\u003c/span\u003e\u003c/span\u003e), Akaike information criterion (\u0026Delta;AIC), and Root mean square error (RMSE) values of baseline and full models\u003c/p\u003e\n\u003c/div\u003e\n\u003c/caption\u003e\n\u003cthead\u003e\n\u003ctr\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eModel\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({R}_{adj}^{2}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003e\u0026Delta;AIC\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eRMSE\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003e# of predictors\u003c/p\u003e\n\u003c/th\u003e\n\u003c/tr\u003e\n\u003c/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eBaseline model\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.27\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e22613.5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e15.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e6\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eFull model\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.34\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e22411.2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e14.3\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e45\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003c/tbody\u003e\n\u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\({R}_{adj}^{2}\\)\u003c/span\u003e \u003c/span\u003e, \u003cem\u003eAdjusted R-squared; \u0026Delta;AIC, Akaike information criterion; RMSE, Root mean square error\u003c/em\u003e\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe first goal of this study was to identify predictors of post-treatment symptom severity in patients with mild to moderate MDD, PD, and SAD treated with highly standardized ICBT protocols and using a large and diverse pool of clinical, genetic, and register predictors. Our findings support the previously established importance of pre-treatment symptom severity, psychiatric comorbidities, family history of psychopathology, and socioeconomic status. In line with existing research on the link between marital status and mental health, being single, divorced or widowed was predictive of a poorer treatment outcome. Furthermore, thanks to unique data sources available in this study, we were able to assess a variety of predictors that have not been investigated before: the role of prior and concurrent psychiatric diagnoses and specific medications is a novel addition of this study, as are socioeconomic predictors such as income and receipt of financial benefits.\u003c/p\u003e \u003cp\u003eIdentification of comorbid ASD and ADHD as strong hindering factors bears potential clinical relevance. The need to adjust psychotherapeutic interventions for depression and anxiety to the patient\u0026rsquo;s ASD symptomatology has long been acknowledged. Socio-communication impairment, difficulties with introspection, and limited cognitive flexibility have been suggested as impeding treatment effectiveness\u003csup\u003e\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e\u003c/sup\u003e. Consequently, multiple CBT adaptations for patients with comorbid ASD have been developed\u003csup\u003e\u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e\u003c/sup\u003e. The discovery that in this patient group, the average post-treatment score was nearly twice as high as for patients without ASD potentially highlights absence of relevant modifications in the studied ICBT treatment. In patients with comorbid ADHD, the findings may reflect a similar lack of treatment accommodation. Perhaps unsurprisingly, maintaining attention and organizing oneself over a 12-week treatment period, which involves extensive homework and the absence of immediate therapist guidance that is characteristic of ICBT, may pose a challenge for this patient group and calls for adaptive treatment strategies to avoid treatment failure\u003csup\u003e\u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e74\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe application of PRS in psychiatry is promising as it is a constant throughout lifetime and can be used for long-term prediction. Yet, their limitation is that on a population level, PRS tend to follow a Gaussian distribution with significant overlap between cases and controls, and they only explain a small fraction of genetic variation, which in turn explains around 50% of phenotypic variation. Failure to detect the association between PRS and post-treatment symptom severity in our study is perhaps not surprising. We used PRS related to psychopathology and personality, and, while genetic liability to these traits is likely to be shared with that of treatment outcome, it does not necessarily follow that the same PRS will be helpful in predicting it. Thus, a specific PRS constructed from GWAS of treatment response, which is still missing, could potentially yield a stronger effect.\u003c/p\u003e \u003cp\u003eThe second goal of the study was to assess whether employing a wider range of predictors would be advantageous in identifying patients who will have higher post-treatment symptom severity. We found that the full model with multimodal predictors had a superior performance in explaining the variance in post-treatment severity compared to the baseline model. However, the improvement in explained variance was modest. One potential interpretation is that even though the evaluated predictors encompass a distinct range of properties, they might, to a certain degree, be considered as capturing a shared underlying construct and providing little independent information. For example, genetic predictors will inevitably be correlated due to vertical pleiotropy, whereby the same genetic variant affects multiple different traits, e.g., a relevant SNP influences intelligence, which serves as a mediator influencing educational attainment, SES, and, in turn, post-treatment symptom severity. While the collinearity of independent variables does not negatively affect the explanatory power of the whole model, it does not improve it either, since many variables explain an overlapping proportion of variance in the outcome. Further, the full model would likely benefit from the inclusion of additional strong predictors. For example, in the exploratory complete-case analysis, duration of symptoms was strongly associated with the outcome in its effect size and variance explained but could not be included as it surpassed the predefined threshold for the allowed amount of missingness. Moreover, human complex traits are dominated by stochasticity and are often attributed to idiosyncratic factors which remain largely unmeasured. Another possibility is that pooling the three disorders makes the studied phenotype more heterogeneous and possibly dilutes the findings. Also, while the current sample size exceeds those in the earlier studies, it may still not be sufficiently powered. Finally, the relationship between predictors and the outcome may be non-linear and contain underlying interaction effects, making classical ordinary least squares modelling unsuitable.\u003c/p\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eStrengths and limitations\u003c/h2\u003e \u003cp\u003eThe main strength of this study is its relatively large sample size and diversity of evaluated predictors. Moreover, register-based data have almost no missing values and are highly reliable. Another strength is a homogenous group of patients that completed a highly protocolized treatment, which allows for meaningful comparisons of the outcome. Deriving the data from routine care has additional benefits since most studies of treatment outcome constitute a primary or secondary analysis of RCT data, where user characteristics are subject to stringent inclusion criteria, thus introducing selection bias and limiting the generalisability\u003csup\u003e\u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e\u003c/sup\u003e. However, a potential limitation in the applicability of the findings to other populations needs to be mentioned. All the study participants lived in Sweden, were more educated than the general population, were fluent in Swedish, and exhibited mild to moderate symptom severity. In addition, there is a significant self-selection bias, with most patients self-referring to the Internet psychiatry clinic.\u003c/p\u003e \u003cp\u003eFinally, it must be emphasized that the findings of this study cannot be viewed as supporting the causal nature of the relationship between any of the predictors and the outcome. While the terms \u003cem\u003etreatment outcome\u003c/em\u003e and \u003cem\u003etreatment response\u003c/em\u003e do semantically assume, at least partially, that the post-treatment symptom measure is conditional upon and a direct consequence of the intervention, no such claims can be made given the non-experimental design of the study. However, causal language is widely and loosely applied across the literature and, in the interest of brevity and recognition, is also sometimes used in this paper, more so when referring to the previous findings rather than the current study, where we try to adapt the more observational term \u003cem\u003epost-treatment symptom severity\u003c/em\u003e. This operationalization of the response variable was chosen as it is deemed by the authors as the most appropriate measure to be predicted at the baseline. Since the goal of the future predictive model is not to assess treatment effectiveness, relative symptom change was not of interest here. Moreover, even when the percentage change is substantial and passes some predefined threshold, the patient may still be symptomatic beyond what is intended. Choosing remission as the outcome was rejected due to its somewhat arbitrary cut-off, loss of information, and diminished statistical power that ensues from dichotomization.\u003c/p\u003e \u003c/div\u003e"},{"header":"Conclusions","content":"\u003cp\u003eThe current study adds to the growing literature on predictors of treatment outcome by supporting established ones and suggesting some novel predictors. It also proposes that a statistical model based on a few relatively easily obtainable measures explains a comparable proportion of the variance in the outcome as a model containing a much broader array of multimodal data. As the next step, a machine learning model should be developed to address non-linear associations and higher-order interactions between input features to perform predictions at individual patient level. Additional future endeavours may include leveraging all available SNP information beyond what meets a strict GWAS significance threshold and appraising an even larger and more diverse pool of register-based predictors.\u003c/p\u003e "},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eCode availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eR code can be provided upon request to the corresponding author.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was funded by the S\u0026ouml;derstr\u0026ouml;m-K\u0026ouml;nig Foundation, FORTE, the Swedish Research Council, and the Centre for innovative medicine. We gratefully acknowledge the contribution of the psychiatric research nurse Monica Hellberg to the collection of blood samples and of Bjorn Roelstraete to the dataset preparation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOK, JW, and CR designed the study; OK conducted data analysis and wrote the manuscript; JJC and MH conducted pre-processing of genetic data and constructed PRS; all authors contributed to and approved the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no conflict of interest.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eKessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. \u003cem\u003eArch Gen Psychiatry\u003c/em\u003e 2005; \u003cstrong\u003e62\u003c/strong\u003e: 593\u0026ndash;602.\u003c/li\u003e\n\u003cli\u003eWorld Health Organization. Depression and Other Common Mental Disorders: Global Health Estimates. World Health Organization, 2017. https://www.jstor.org/stable/resrep28026.1 (accessed 7 Mar 2024).\u003c/li\u003e\n\u003cli\u003eAmerican Psychological Association. Clinical Practice Guideline for the Treatment of Depression Across Three Age Cohorts. https://www.apa.org. https://www.apa.org/depression-guideline (accessed 4 Oct 2023).\u003c/li\u003e\n\u003cli\u003ePractice Guideline for the Treatment of Patients With Panic Disorder. In: \u003cem\u003eAPA Practice Guidelines for the Treatment of Psychiatric Disorders: Comprehensive Guidelines and Guideline Watches\u003c/em\u003e. American Psychiatric Association: Arlington, VA, 2006 doi:10.1176/appi.books.9780890423363.51396.\u003c/li\u003e\n\u003cli\u003eCarlbring P, Andersson G, Cuijpers P, Riper H, Hedman-Lagerl\u0026ouml;f E. Internet-based vs. face-to-face cognitive behavior therapy for psychiatric and somatic disorders: an updated systematic review and meta-analysis. \u003cem\u003eCogn Behav Ther\u003c/em\u003e 2018; \u003cstrong\u003e47\u003c/strong\u003e: 1\u0026ndash;18.\u003c/li\u003e\n\u003cli\u003eHedman E, Lj\u0026oacute;tsson B, Lindefors N. Cognitive behavior therapy via the Internet: a systematic review of applications, clinical efficacy and cost\u0026ndash;effectiveness. \u003cem\u003eExpert Rev Pharmacoecon Outcomes Res\u003c/em\u003e 2012; \u003cstrong\u003e12\u003c/strong\u003e: 745\u0026ndash;764.\u003c/li\u003e\n\u003cli\u003eKiropoulos LA, Klein B, Austin DW, Gilson K, Pier C, Mitchell J \u003cem\u003eet al.\u003c/em\u003e Is internet-based CBT for panic disorder and agoraphobia as effective as face-to-face CBT? \u003cem\u003eJ Anxiety Disord\u003c/em\u003e 2008; \u003cstrong\u003e22\u003c/strong\u003e: 1273\u0026ndash;1284.\u003c/li\u003e\n\u003cli\u003eHedman E, Andersson G, Lj\u0026oacute;tsson B, Andersson E, R\u0026uuml;ck C, M\u0026ouml;rtberg E \u003cem\u003eet al.\u003c/em\u003e Internet-Based Cognitive Behavior Therapy vs. Cognitive Behavioral Group Therapy for Social Anxiety Disorder: A Randomized Controlled Non-inferiority Trial. \u003cem\u003ePLoS ONE\u003c/em\u003e 2011; \u003cstrong\u003e6\u003c/strong\u003e: e18001.\u003c/li\u003e\n\u003cli\u003eAndrews G, Davies M, Titov N. Effectiveness Randomized Controlled Trial of Face to Face Versus Internet Cognitive Behaviour Therapy for Social Phobia. \u003cem\u003eAust N Z J Psychiatry\u003c/em\u003e 2011; \u003cstrong\u003e45\u003c/strong\u003e: 337\u0026ndash;340.\u003c/li\u003e\n\u003cli\u003eAndrews G, Cuijpers P, Craske MG, McEvoy P, Titov N. Computer Therapy for the Anxiety and Depressive Disorders Is Effective, Acceptable and Practical Health Care: A Meta-Analysis. \u003cem\u003ePLoS ONE\u003c/em\u003e 2010; \u003cstrong\u003e5\u003c/strong\u003e: e13196.\u003c/li\u003e\n\u003cli\u003eWarmerdam L, Smit F, Van Straten A, Riper H, Cuijpers P. Cost-Utility and Cost-Effectiveness of Internet-Based Treatment for Adults With Depressive Symptoms: Randomized Trial. \u003cem\u003eJ Med Internet Res\u003c/em\u003e 2010; \u003cstrong\u003e12\u003c/strong\u003e: e53.\u003c/li\u003e\n\u003cli\u003eKraepelien M, Mattsson S, Hedman-Lagerl\u0026ouml;f E, Petersson IF, Forsell Y, Lindefors N \u003cem\u003eet al.\u003c/em\u003e Cost-effectiveness of internet-based cognitive\u0026ndash;behavioural therapy and physical exercise for depression. \u003cem\u003eBJPsych Open\u003c/em\u003e 2018; \u003cstrong\u003e4\u003c/strong\u003e: 265\u0026ndash;273.\u003c/li\u003e\n\u003cli\u003eRozental A, Andersson G, Carlbring P. In the Absence of Effects: An Individual Patient Data Meta-Analysis of Non-response and Its Predictors in Internet-Based Cognitive Behavior Therapy. \u003cem\u003eFront Psychol\u003c/em\u003e 2019; \u003cstrong\u003e10\u003c/strong\u003e: 589.\u003c/li\u003e\n\u003cli\u003eAndersson G, Carlbring P, Rozental A. Response and Remission Rates in Internet-Based Cognitive Behavior Therapy: An Individual Patient Data Meta-Analysis. \u003cem\u003eFront Psychiatry\u003c/em\u003e 2019; \u003cstrong\u003e10\u003c/strong\u003e: 749.\u003c/li\u003e\n\u003cli\u003eLoerinc AG, Meuret AE, Twohig MP, Rosenfield D, Bluett EJ, Craske MG. Response rates for CBT for anxiety disorders: Need for standardized criteria. \u003cem\u003eClin Psychol Rev\u003c/em\u003e 2015; \u003cstrong\u003e42\u003c/strong\u003e: 72\u0026ndash;82.\u003c/li\u003e\n\u003cli\u003eAndersson G, Bergstr\u0026ouml;m J, Holl\u0026auml;ndare F, Ekselius L, Carlbring P. Delivering Cognitive Behavioural Therapy for Mild to Moderate Depression via the Internet: Predicting Outcome at 6-Month Follow-Up. \u003cem\u003eVerhaltenstherapie\u003c/em\u003e 2004; \u003cstrong\u003e14\u003c/strong\u003e: 185\u0026ndash;189.\u003c/li\u003e\n\u003cli\u003eHamilton KE, Dobson KS. Cognitive therapy of depression. \u003cem\u003eClin Psychol Rev\u003c/em\u003e 2002; \u003cstrong\u003e22\u003c/strong\u003e: 875\u0026ndash;893.\u003c/li\u003e\n\u003cli\u003eJarrett RB, Eaves GG, Grannemann BD, Rush AJ. Clinical, cognitive, and demographic predictors of response to cognitive therapy for depression: A preliminary report. \u003cem\u003ePsychiatry Res\u003c/em\u003e 1991; \u003cstrong\u003e37\u003c/strong\u003e: 245\u0026ndash;260.\u003c/li\u003e\n\u003cli\u003eBlack DW, Wesner RB, Gabel J, Bowers W, Monahan P. Predictors of short-term treatment response in 66 patients with panic disorder. \u003cem\u003eJ Affect Disord\u003c/em\u003e 1994; \u003cstrong\u003e30\u003c/strong\u003e: 233\u0026ndash;241.\u003c/li\u003e\n\u003cli\u003eDe Graaf LE, Hollon SD, Huibers MJH. Predicting outcome in computerized cognitive behavioral therapy for depression in primary care: A randomized trial. \u003cem\u003eJ Consult Clin Psychol\u003c/em\u003e 2010; \u003cstrong\u003e78\u003c/strong\u003e: 184\u0026ndash;189.\u003c/li\u003e\n\u003cli\u003eEl Alaoui S, Hedman E, Lj\u0026oacute;tsson B, Lindefors N. Long-term effectiveness and outcome predictors of therapist-guided internet-based cognitive\u0026ndash;behavioural therapy for social anxiety disorder in routine psychiatric care. \u003cem\u003eBMJ Open\u003c/em\u003e 2015; \u003cstrong\u003e5\u003c/strong\u003e: e007902.\u003c/li\u003e\n\u003cli\u003eEdmonds M, Hadjistavropoulos HD, Schneider LH, Dear BF, Titov N. Who benefits most from therapist-assisted internet-delivered cognitive behaviour therapy in clinical practice? Predictors of symptom change and dropout. \u003cem\u003eJ Anxiety Disord\u003c/em\u003e 2018; \u003cstrong\u003e54\u003c/strong\u003e: 24\u0026ndash;32.\u003c/li\u003e\n\u003cli\u003eButton KS, Wiles NJ, Lewis G, Peters TJ, Kessler D. Factors associated with differential response to online cognitive behavioural therapy. \u003cem\u003eSoc Psychiatry Psychiatr Epidemiol\u003c/em\u003e 2012; \u003cstrong\u003e47\u003c/strong\u003e: 827\u0026ndash;833.\u003c/li\u003e\n\u003cli\u003eHadjistavropoulos HD, Pugh NE, Hesser H, Andersson G. Predicting Response to Therapist-Assisted Internet-Delivered Cognitive Behavior Therapy for Depression or Anxiety Within an Open Dissemination Trial. \u003cem\u003eBehav Ther\u003c/em\u003e 2016; \u003cstrong\u003e47\u003c/strong\u003e: 155\u0026ndash;165.\u003c/li\u003e\n\u003cli\u003eNordgreen T, Havik OE, \u0026Ouml;st LG, Furmark T, Carlbring P, Andersson G. Outcome predictors in guided and unguided self-help for social anxiety disorder. \u003cem\u003eBehav Res Ther\u003c/em\u003e 2012; \u003cstrong\u003e50\u003c/strong\u003e: 13\u0026ndash;21.\u003c/li\u003e\n\u003cli\u003eCopay AG, Subach BR, Glassman SD, Polly DW, Schuler TC. Understanding the minimum clinically important difference: a review of concepts and methods. \u003cem\u003eSpine J\u003c/em\u003e 2007; \u003cstrong\u003e7\u003c/strong\u003e: 541\u0026ndash;546.\u003c/li\u003e\n\u003cli\u003eMululo SCC, De Menezes GB, Vigne P, Fontenelle LF. A review on predictors of treatment outcome in social anxiety disorder. \u003cem\u003eRev Bras Psiquiatr\u003c/em\u003e 2012; \u003cstrong\u003e34\u003c/strong\u003e: 92\u0026ndash;100.\u003c/li\u003e\n\u003cli\u003eHedman E, Andersson E, Lj\u0026oacute;tsson B, Andersson G, Andersson E, Schalling M \u003cem\u003eet al.\u003c/em\u003e Clinical and genetic outcome determinants of Internet- and group-based cognitive behavior therapy for social anxiety disorder: Predictors and moderators of CBT for SAD 1. \u003cem\u003eActa Psychiatr Scand\u003c/em\u003e 2012; \u003cstrong\u003e126\u003c/strong\u003e: 126\u0026ndash;136.\u003c/li\u003e\n\u003cli\u003eChambless DL, Tran GQ, Glass CR. Predictors of Response to Cognitive-Behavioral Group Therapy for Social Phobia. \u003cem\u003eJ Anxiety Disord\u003c/em\u003e 1997; \u003cstrong\u003e11\u003c/strong\u003e: 221\u0026ndash;240.\u003c/li\u003e\n\u003cli\u003ePrasko J, Ociskova M, Latalova K, Kamaradova D, Grambal A. Psychological factors and treatment effectiveness in resistant anxiety disorders in highly comorbid inpatients. \u003cem\u003eNeuropsychiatr Dis Treat\u003c/em\u003e 2016; \u003cstrong\u003eVolume 12\u003c/strong\u003e: 1539\u0026ndash;1551.\u003c/li\u003e\n\u003cli\u003eEl Alaoui S, Lj\u0026oacute;tsson B, Hedman E, Svanborg C, Kaldo V, Lindefors N. Predicting Outcome in Internet-Based Cognitive Behaviour Therapy for Major Depression: A Large Cohort Study of Adult Patients in Routine Psychiatric Care. \u003cem\u003ePLOS ONE\u003c/em\u003e 2016; \u003cstrong\u003e11\u003c/strong\u003e: e0161191.\u003c/li\u003e\n\u003cli\u003eSchat A, Van Noorden MS, Noom MJ, Giltay EJ, Van Der Wee NJA, Vermeiren RRJM \u003cem\u003eet al.\u003c/em\u003e Predictors of outcome in outpatients with anxiety disorders: The Leiden routine outcome monitoring study. \u003cem\u003eJ Psychiatr Res\u003c/em\u003e 2013; \u003cstrong\u003e47\u003c/strong\u003e: 1876\u0026ndash;1885.\u003c/li\u003e\n\u003cli\u003eKolek A, Prasko J, Ociskova M, Holubova M, Vanek J, Grambal A \u003cem\u003eet al.\u003c/em\u003e Severity of panic disorder, adverse events in childhood, dissociation, self-stigma and comorbid personality disorders Part 2: Therapeutic effectiveness of a combined cognitive behavioural therapy and pharmacotherapy in treatment-resistant inpatients. \u003cem\u003eNeuro Endocrinol Lett\u003c/em\u003e 2019; \u003cstrong\u003e40\u003c/strong\u003e: 271\u0026ndash;283.\u003c/li\u003e\n\u003cli\u003eKamaradova D PJ. Therapeutic response to complex cognitive-behavioral and pharmacological treatment in patients with social phobia. \u003c/li\u003e\n\u003cli\u003eMausbach BT, Moore R, Roesch S, Cardenas V, Patterson TL. The Relationship Between Homework Compliance and Therapy Outcomes: An Updated Meta-Analysis. \u003cem\u003eCogn Ther Res\u003c/em\u003e 2010; \u003cstrong\u003e34\u003c/strong\u003e: 429\u0026ndash;438.\u003c/li\u003e\n\u003cli\u003eBur OT, Bielinski LL, Krauss S, H\u0026auml;fliger A, Guggisberg J, Krieger T \u003cem\u003eet al.\u003c/em\u003e Working alliance and adherence mediate the effect of guidance in a web-based program for participants with mild to moderate depressive symptoms: A secondary mediation analysis. \u003cem\u003eInternet Interv\u003c/em\u003e 2022; \u003cstrong\u003e30\u003c/strong\u003e: 100593.\u003c/li\u003e\n\u003cli\u003eEley TC. The future of therapygenetics: where will studies predicting psychological treatment response from genomic markers lead?: The Cutting Edge: The Future of Therapygenetics. \u003cem\u003eDepress Anxiety\u003c/em\u003e 2014; \u003cstrong\u003e31\u003c/strong\u003e: 617\u0026ndash;620.\u003c/li\u003e\n\u003cli\u003eFusar-Poli L, Rutten BPF, Van Os J, Aguglia E, Guloksuz S. Polygenic risk scores for predicting outcomes and treatment response in psychiatry: hope or hype? \u003cem\u003eInt Rev Psychiatry\u003c/em\u003e 2022; \u003cstrong\u003e34\u003c/strong\u003e: 663\u0026ndash;675.\u003c/li\u003e\n\u003cli\u003eMeerman JJ, Ter Hark SE, Janzing JGE, Coenen MJH. The Potential of Polygenic Risk Scores to Predict Antidepressant Treatment Response in Major Depression: A Systematic Review. \u003cem\u003eJ Affect Disord\u003c/em\u003e 2022; \u003cstrong\u003e304\u003c/strong\u003e: 1\u0026ndash;11.\u003c/li\u003e\n\u003cli\u003eWard J, Graham N, Strawbridge RJ, Ferguson A, Jenkins G, Chen W \u003cem\u003eet al.\u003c/em\u003e Polygenic risk scores for major depressive disorder and neuroticism as predictors of antidepressant response: Meta-analysis of three treatment cohorts. \u003cem\u003ePLOS ONE\u003c/em\u003e 2018; \u003cstrong\u003e13\u003c/strong\u003e: e0203896.\u003c/li\u003e\n\u003cli\u003eAndersson E, Crowley JJ, Lindefors N, Lj\u0026oacute;tsson B, Hedman-Lagerl\u0026ouml;f E, Boberg J \u003cem\u003eet al.\u003c/em\u003e Genetics of response to cognitive behavior therapy in adults with major depression: a preliminary report. \u003cem\u003eMol Psychiatry\u003c/em\u003e 2019; \u003cstrong\u003e24\u003c/strong\u003e: 484\u0026ndash;490.\u003c/li\u003e\n\u003cli\u003eWallert J, Boberg J, Kaldo V, Mataix-Cols D, Flygare O, Crowley JJ \u003cem\u003eet al.\u003c/em\u003e Predicting remission after internet-delivered psychotherapy in patients with depression using machine learning and multi-modal data. \u003cem\u003eTransl Psychiatry\u003c/em\u003e 2022; \u003cstrong\u003e12\u003c/strong\u003e: 357.\u003c/li\u003e\n\u003cli\u003eRayner C, Coleman JRI, Purves KL, Hodsoll J, Goldsmith K, Alpers GW \u003cem\u003eet al.\u003c/em\u003e A genome-wide association meta-analysis of prognostic outcomes following cognitive behavioural therapy in individuals with anxiety and depressive disorders. \u003cem\u003eTransl Psychiatry\u003c/em\u003e 2019; \u003cstrong\u003e9\u003c/strong\u003e: 150.\u003c/li\u003e\n\u003cli\u003eColeman JRI, Lester KJ, Keers R, Roberts S, Curtis C, Arendt K \u003cem\u003eet al.\u003c/em\u003e Genome-wide association study of response to cognitive\u0026ndash;behavioural therapy in children with anxiety disorders. \u003cem\u003eBr J Psychiatry\u003c/em\u003e 2016; \u003cstrong\u003e209\u003c/strong\u003e: 236\u0026ndash;243.\u003c/li\u003e\n\u003cli\u003eThe MULTI-PSYCH cohort | Karolinska Institutet. https://ki.se/en/cns/the-multi-psych-cohort (accessed 4 Oct2023).\u003c/li\u003e\n\u003cli\u003eTitov N, Dear B, Nielssen O, Staples L, Hadjistavropoulos H, Nugent M \u003cem\u003eet al.\u003c/em\u003e ICBT in routine care: A descriptive analysis of successful clinics in five countries. \u003cem\u003eInternet Interv\u003c/em\u003e 2018; \u003cstrong\u003e13\u003c/strong\u003e: 108\u0026ndash;115.\u003c/li\u003e\n\u003cli\u003eFantino B, Moore N. The self-reported Montgomery-\u0026Aring;sberg depression rating scale is a useful evaluative tool in major depressive disorder. \u003cem\u003eBMC Psychiatry\u003c/em\u003e 2009; \u003cstrong\u003e9\u003c/strong\u003e: 26.\u003c/li\u003e\n\u003cli\u003eHouck PR, Spiegel DA, Shear MK, Rucci P. Reliability of the self-report version of the panic disorder severity scale. \u003cem\u003eDepress Anxiety\u003c/em\u003e 2002; \u003cstrong\u003e15\u003c/strong\u003e: 183\u0026ndash;185.\u003c/li\u003e\n\u003cli\u003eBaker SL, Heinrichs N, Kim H-J, Hofmann SG. The Liebowitz social anxiety scale as a self-report instrument: a preliminary psychometric analysis. \u003cem\u003eBehav Res Ther\u003c/em\u003e 2002; \u003cstrong\u003e40\u003c/strong\u003e: 701\u0026ndash;715.\u003c/li\u003e\n\u003cli\u003eLahey BB, Krueger RF, Rathouz PJ, Waldman ID, Zald DH. A hierarchical causal taxonomy of psychopathology across the life span. \u003cem\u003ePsychol Bull\u003c/em\u003e 2017; \u003cstrong\u003e143\u003c/strong\u003e: 142\u0026ndash;186.\u003c/li\u003e\n\u003cli\u003ePettersson E, Larsson H, D\u0026rsquo;Onofrio BM, B\u0026ouml;lte S, Lichtenstein P. The general factor of psychopathology: a comparison with the general factor of intelligence with respect to magnitude and predictive validity. \u003cem\u003eWorld Psychiatry\u003c/em\u003e 2020; \u003cstrong\u003e19\u003c/strong\u003e: 206\u0026ndash;213.\u003c/li\u003e\n\u003cli\u003eKotov R, Krueger RF, Watson D, Achenbach TM, Althoff RR, Bagby RM \u003cem\u003eet al.\u003c/em\u003e The Hierarchical Taxonomy of Psychopathology (HiTOP): A dimensional alternative to traditional nosologies. \u003cem\u003eJ Abnorm Psychol\u003c/em\u003e 2017; \u003cstrong\u003e126\u003c/strong\u003e: 454\u0026ndash;477.\u003c/li\u003e\n\u003cli\u003eZantvoort K, Hentati Isacsson N, Funk B, Kaldo V. Data set size vs homogeneity \u0026ndash; A Machine Learning study on pooling intervention data in E-Mental Health dropout predictions. 2024 [submitted for publication]\u003c/li\u003e\n\u003cli\u003eMataix-Cols D, Fern\u0026aacute;ndez De La Cruz L, Monzani B, Rosenfield D, Andersson E, P\u0026eacute;rez-Vigil A \u003cem\u003eet al.\u003c/em\u003e D-Cycloserine Augmentation of Exposure-Based Cognitive Behavior Therapy for Anxiety, Obsessive-Compulsive, and Posttraumatic Stress Disorders: A Systematic Review and Meta-analysis of Individual Participant Data. \u003cem\u003eJAMA Psychiatry\u003c/em\u003e 2017; \u003cstrong\u003e74\u003c/strong\u003e: 501.\u003c/li\u003e\n\u003cli\u003eLam M, Awasthi S, Watson HJ, Goldstein J, Panagiotaropoulou G, Trubetskoy V \u003cem\u003eet al.\u003c/em\u003e RICOPILI: Rapid Imputation for COnsortias PIpeLIne. \u003cem\u003eBioinformatics\u003c/em\u003e 2020; \u003cstrong\u003e36\u003c/strong\u003e: 930\u0026ndash;933.\u003c/li\u003e\n\u003cli\u003eBoberg J, Kaldo V, Mataix-Cols D, Crowley JJ, Roelstraete B, Halvorsen M \u003cem\u003eet al.\u003c/em\u003e Swedish multimodal cohort of patients with anxiety or depression treated with internet-delivered psychotherapy (MULTI-PSYCH). \u003cem\u003eBMJ Open\u003c/em\u003e 2023; \u003cstrong\u003e13\u003c/strong\u003e: e069427.\u003c/li\u003e\n\u003cli\u003eLudvigsson JF, Almqvist C, Bonamy A-KE, Ljung R, Micha\u0026euml;lsson K, Neovius M \u003cem\u003eet al.\u003c/em\u003e Registers of the Swedish total population and their use in medical research. \u003cem\u003eEur J Epidemiol\u003c/em\u003e 2016; \u003cstrong\u003e31\u003c/strong\u003e: 125\u0026ndash;136.\u003c/li\u003e\n\u003cli\u003eStat. SCB. Labour statistics based on administrative sources. https://www.scb.se/en/finding-statistics/statistics-by-subject-area/labour-market/employment-and-working-hours/labour-statistics-based-on-administrative-sources/ (accessed 7 Mar2024).\u003c/li\u003e\n\u003cli\u003eStatistics Sweden. Total population register. https://www.scb.se/contentassets/8f66bcf5abc34d0b98afa4fcbfc0e060/rtb-bar-2016-eng.pdf.\u003c/li\u003e\n\u003cli\u003eSocialstyrelsen. National Patient Register. 2023.https://www.socialstyrelsen.se/en/statistics-and-data/registers/national-patient-register/ (accessed 7 Mar2024).\u003c/li\u003e\n\u003cli\u003eVAL-databaserna \u0026ndash; Region Stockholm. https://www.folkhalsokollen.se/datakallor/val-databaserna/ (accessed 7 Mar2024).\u003c/li\u003e\n\u003cli\u003eSocialstyrelsen. National Prescribed Drug Register. 2024.https://www.socialstyrelsen.se/en/statistics-and-data/registers/national-prescribed-drug-register/ (accessed 7 Mar2024).\u003c/li\u003e\n\u003cli\u003eRubin DB. Inference and missing data. \u003cem\u003eBiometrika\u003c/em\u003e 1976; \u003cstrong\u003e63\u003c/strong\u003e: 581\u0026ndash;592.\u003c/li\u003e\n\u003cli\u003eBell ML, Fairclough DL. Practical and statistical issues in missing data for longitudinal patient-reported outcomes. \u003cem\u003eStat Methods Med Res\u003c/em\u003e 2014; \u003cstrong\u003e23\u003c/strong\u003e: 440\u0026ndash;459.\u003c/li\u003e\n\u003cli\u003eMoons KGM, Donders RART, Stijnen T, Harrell FE. Using the outcome for imputation of missing predictor values was preferred. \u003cem\u003eJ Clin Epidemiol\u003c/em\u003e 2006; \u003cstrong\u003e59\u003c/strong\u003e: 1092\u0026ndash;1101.\u003c/li\u003e\n\u003cli\u003eVan Buuren S. \u003cem\u003eFlexible Imputation of Missing Data, Second Edition\u003c/em\u003e. 2nd ed. Chapman and Hall/CRC: Second edition. | Boca Raton, Florida : CRC Press, [2019] |, 2018 doi:10.1201/9780429492259.\u003c/li\u003e\n\u003cli\u003eSterne JAC, White IR, Carlin JB, Spratt M, Royston P, Kenward MG \u003cem\u003eet al.\u003c/em\u003e Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. \u003cem\u003eBMJ\u003c/em\u003e 2009; \u003cstrong\u003e338\u003c/strong\u003e: b2393\u0026ndash;b2393.\u003c/li\u003e\n\u003cli\u003eAbayomi K, Gelman A, Levy M. Diagnostics for Multivariate Imputations. \u003cem\u003eJ R Stat Soc Ser C Appl Stat\u003c/em\u003e 2008; \u003cstrong\u003e57\u003c/strong\u003e: 273\u0026ndash;291.\u003c/li\u003e\n\u003cli\u003eHern\u0026aacute;n MA, Hern\u0026aacute;ndez-D\u0026iacute;az S, Robins JM. A Structural Approach to Selection Bias: \u003cem\u003eEpidemiology\u003c/em\u003e 2004; \u003cstrong\u003e15\u003c/strong\u003e: 615\u0026ndash;625.\u003c/li\u003e\n\u003cli\u003eBuuren SV, Groothuis-Oudshoorn K. \u003cstrong\u003emice\u003c/strong\u003e : Multivariate Imputation by Chained Equations in \u003cem\u003eR\u003c/em\u003e. \u003cem\u003eJ Stat Softw\u003c/em\u003e 2011; \u003cstrong\u003e45\u003c/strong\u003e. doi:10.18637/jss.v045.i03.\u003c/li\u003e\n\u003cli\u003eHarrell FE. \u003cem\u003eRegression Modeling Strategies: With Applications to Linear Models, Logistic Regression, and Survival Analysis\u003c/em\u003e. Springer New York: New York, NY, 2001 doi:10.1007/978-1-4757-3462-1.\u003c/li\u003e\n\u003cli\u003eSpain D, Happ\u0026eacute; F. How to Optimise Cognitive Behaviour Therapy (CBT) for People with Autism Spectrum Disorders (ASD): A Delphi Study. \u003cem\u003eJ Ration-Emotive Cogn-Behav Ther\u003c/em\u003e 2020; \u003cstrong\u003e38\u003c/strong\u003e: 184\u0026ndash;208.\u003c/li\u003e\n\u003cli\u003eMenezes M, Harkins C, Robinson MF, Mazurek MO. Treatment of Depression in Individuals with Autism Spectrum Disorder: A Systematic Review. \u003cem\u003eRes Autism Spectr Disord\u003c/em\u003e 2020; \u003cstrong\u003e78\u003c/strong\u003e: 101639.\u003c/li\u003e\n\u003cli\u003eEddy LD, Knouse LE, Safren SA. Adapting CBT for Treating Anxiety Disorders and Depression in Adults with ADHD. In: Todd G, Branch R (eds). \u003cem\u003eEvidence-Based Treatment for Anxiety Disorders and Depression: A Cognitive Behavioral Therapy Compendium\u003c/em\u003e. Cambridge University Press: Cambridge, 2022, pp 553\u0026ndash;572.\u003c/li\u003e\n\u003cli\u003eTaylor S, Abramowitz JS, McKay D. Non-adherence and non-response in the treatment of anxiety disorders. \u003cem\u003eJ Anxiety Disord\u003c/em\u003e 2012; \u003cstrong\u003e26\u003c/strong\u003e: 583\u0026ndash;589.\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Supplementary Information","content":"\u003cp\u003eSupplementary Figure and Supplementary Tables are not available with this version.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-4075444/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4075444/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eInternet-delivered cognitive behavioural therapy (ICBT) is an effective and accessible treatment for mild to moderate depression and anxiety disorders. However, up to 50% of patients do not experience sufficient symptom relief. Identifying patient characteristics predictive of higher post-treatment symptom severity is crucial for devising personalized interventions to avoid treatment failures and reduce healthcare costs. Using the new Swedish multimodal database MULTI-PSYCH, we expand upon established predictors of treatment outcome and assess the added benefit of utilizing polygenic risk scores (PRS) and nationwide register data in a combined sample of 2668 patients treated with ICBT for major depressive disorder (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1300), panic disorder (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;727), and social anxiety disorder (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;641). We present two linear regression models: a baseline model using six well-established predictors and a full model incorporating six clinic-based, 32 register-based predictors, and PRS for seven psychiatric disorders and traits. First, we assessed predictor importance through bivariate associations and then compared the models based on the proportion of variance explained in post-treatment scores. Our analysis identified several novel predictors of higher post-treatment severity, including comorbid ASD and ADHD, receipt of financial benefits, and prior use of some psychotropic medications. The baseline model explained 27% of the variance in post-treatment symptom scores, while the full model offered a modest improvement, explaining 34%. Developing a machine learning model that can capture complex non-linear associations and interactions between high-quality multimodal input features is a viable next step to improve prediction of symptom severity post ICBT.\u003c/p\u003e","manuscriptTitle":"Clinical, genetic, and sociodemographic predictors of symptom severity after internet-delivered cognitive behavioural therapy for depression and anxiety","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-03-29 18:23:54","doi":"10.21203/rs.3.rs-4075444/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a3cb487d-4272-4785-b2cc-1482d7a5f4cf","owner":[],"postedDate":"March 29th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":29352445,"name":"Health sciences/Diseases/Psychiatric disorders/Depression"},{"id":29352446,"name":"Health sciences/Diseases/Psychiatric disorders/ADHD"},{"id":29352447,"name":"Health sciences/Diseases/Psychiatric disorders/Autism spectrum disorders"}],"tags":[],"updatedAt":"2025-02-04T07:38:39+00:00","versionOfRecord":[],"versionCreatedAt":"2024-03-29 18:23:54","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4075444","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4075444","identity":"rs-4075444","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}