Cohort Profile: Investigating Antidepressant Response within Generation Scotland

This study describes the recruitment and initial characterization of 1,180 Generation Scotland participants with a history of antidepressant treatment, who completed a detailed questionnaire to establish clinically meaningful phenotypes of treatment response that will be validated against electronic health records and linked with biological samples to investigate the mechanisms underlying variability in antidepressant efficacy.

The Gist

The Big Picture: Why Are We Doing This?

Imagine depression as a giant, tangled knot of yarn. For millions of people, this knot is tight and painful. Doctors have a box of tools called antidepressants (like SSRIs, SNRIs, etc.) to try to untie it.

The problem? The tools don't work for everyone in the same way.

• For some people, a specific tool unties the knot instantly.
• For others, the tool just makes the knot tighter or causes the yarn to snap (side effects).
• Currently, doctors often have to play a game of "trial and error." They guess which tool might work, try it for a few months, and if it fails, they swap it for another. This process can take years and leave patients feeling hopeless.

This study is part of a massive project called AMBER (Antidepressant Medications: Biology, Exposure & Response). The goal is to stop guessing and start knowing why a specific tool works for one person but not another.

The Setting: "Generation Scotland"

Think of Generation Scotland as a giant, well-organized library of people's lives. It holds data on over 40,000 volunteers who have already shared their DNA, health records, and lifestyle habits. It's like having a massive map of the population's genetic and health landscape.

The researchers wanted to use this library to find people who had tried antidepressants and ask them: "How did it go?"

The Experiment: The "Life Story" Questionnaire

Between July and November 2025, the researchers sent a digital invitation to about 15,000 people in this library. They asked: "If you've ever taken antidepressants for depression, please tell us your story."

Who answered?
About 1,180 people filled out a very detailed questionnaire.

• The Crowd: Most were women (78%), mostly white (98%), and generally older (average age 57) than the average Scottish person.
• The Story: They didn't just say "it worked" or "it didn't." They answered deep questions about:
  • Which specific drugs they tried.
  • How long they took them.
  • How much the drugs helped specific symptoms (like sleep, anxiety, or lack of energy).
  • What side effects they suffered (like weight gain or nausea).

The Findings: What Did the Stories Reveal?

The researchers analyzed these stories and found some fascinating patterns:

1. The "One-Size-Fits-All" Myth is Dead
The symptoms people felt were incredibly diverse. While everyone felt sad, some felt tired and heavy (like carrying a backpack of rocks), while others felt anxious and jittery (like a hummingbird).

• Analogy: It's like trying to fix a car. Some cars have flat tires (anxiety), others have engine trouble (low energy). Using the same wrench on both doesn't make sense.

2. The "Super-Responders" vs. The "Stuck"
Using strict rules, the team categorized the participants:

• The Responders (24%): These people took an SSRI (a common antidepressant), stuck with it for at least 6 weeks, and felt a lot better. They are the "success stories."
• The Non-Responders (1.5%): These people tried at least two different SSRIs for a long time, but none of them worked. They are the "stuck" group.
• The "Unsure" (75%): Most people fell in the middle. They might have had mixed results or didn't fit the strict rules. The researchers intentionally made the rules strict to find the clearest examples of "success" and "failure" for their next steps.

3. The "SNRI" Discovery (The Plot Twist)
Here is the most exciting part. The researchers found a tiny group of people who failed with the standard "SSRI" tools but found relief with a different tool called an SNRI.

• Who were they? These people had a very specific "flavor" of depression: extreme fatigue, suicidal thoughts, headaches, and feeling detached from reality.
• The Lesson: It suggests that for people with this specific type of heavy, tired depression, the standard first-line drug might be the wrong key. They might need the SNRI key right from the start.

What Happens Next? (The "Lab" Phase)

The questionnaire was just the first step. Now, the researchers are moving to the biological detective work.

1. Double-Checking: They will link the questionnaire answers to the participants' actual medical records (like a bank statement vs. a diary) to make sure the stories are accurate.
2. The DNA & Cell Hunt: They will invite 25 "Super-Responders" and 25 "Non-Responders" to give blood and saliva samples.
3. The Cell Lab: They will grow tiny "mini-brains" (cell lines) from these people's blood cells in a petri dish.
4. The Test: They will expose these cells to antidepressants in the lab.
   • The Goal: To see if the cells from the "Super-Responders" light up differently than the "Non-Responders" when hit with the drug. This will reveal the biological "fingerprint" of why the drug works for some and fails for others.

Why Does This Matter?

Imagine if, instead of guessing which antidepressant to try, a doctor could take a quick blood test and say: "Your brain chemistry looks like Group A, so Drug X will likely work for you, and Drug Y will probably give you side effects."

This study is building the foundation for that future. By combining real-life stories (from the questionnaire) with hard science (DNA and cell labs), they hope to turn the "trial and error" game into a precision prescription, saving patients years of suffering and helping them find the right key to untie their knot much faster.

Technical Summary

1. Problem Statement

Major Depressive Disorder (MDD) is a leading cause of global disability, yet pharmacological treatment response is highly heterogeneous. Approximately one-third of patients achieve full remission with their first antidepressant trial, while another third progress to Treatment-Resistant Depression (TRD), defined as a lack of response to two or more adequate trials. Current prescribing relies heavily on a "trial-and-error" approach due to a lack of understanding regarding the biological mechanisms driving variability in treatment outcomes.

Key challenges hindering progress include:

• Phenotypic Heterogeneity: MDD presents with diverse symptom clusters (somatic, cognitive, mood, anxiety) that are not fully captured by standard diagnostic tools like the PHQ-9.
• Data Gaps: There is a scarcity of large-scale studies integrating patient-reported outcomes (PROs), longitudinal electronic health records (EHRs), and genomic data to define precise response phenotypes.
• Biological Mechanisms: The genetic and epigenetic factors underlying why specific individuals respond to Selective Serotonin Reuptake Inhibitors (SSRIs) versus other classes (e.g., SNRIs) remain poorly understood.

2. Methodology

The study is part of the AMBER (Antidepressant Medications: Biology, Exposure & Response) research programme, utilizing the Generation Scotland (GS) cohort, a large, consented, population-based resource with genomic, clinical, and EHR linkage capabilities.

• Study Design: A cross-sectional survey deployed to a subset of the GS cohort between July and November 2025.
• Participants:
  • Target Population: 15,117 GS participants (aged ≥18) with consent for recontact.
  • Analytic Sample: 1,180 participants who completed the questionnaire and reported a history of antidepressant use for mood-related symptoms.
• Instrument Development:
  • A bespoke questionnaire was adapted from the Australian Genetics of Depression Study (AGDS:Cell-o).
  • Co-production: The instrument was refined with the AMBER Lived Experience Advisory Panel (LEAP) to ensure clarity, sensitivity, and relevance to real-world patient experiences.
  • Content: The survey covered depressive symptom domains (mood, anxiety, cognitive, sleep, behavior, physical), medication history (specific drugs, duration, discontinuation reasons), perceived efficacy, side effects, and time-to-response.
• Response Classification Framework:
  • A rule-based algorithm was developed to categorize participants into Responders, Non-Responders, or Unclassified based on self-reported data.
  • Responders: Taken ≥1 antidepressant for ≥6 weeks, rated as "Yes, a lot" effective, and did not discontinue due to inefficacy.
  • Non-Responders: Failed ≥2 adequate trials (≥6 weeks) rated as "No" effective, with discontinuations due to inefficacy (side-effect-only discontinuations excluded).
  • Focus: The primary classification focused on SSRIs to facilitate subsequent biological sampling.
• Future Integration: The study outlines a pipeline to link questionnaire data with NHS EHRs for validation, followed by the collection of biological samples (blood/saliva) from extreme phenotypes (25 responders, 25 non-responders) for DNA methylation profiling and the generation of patient-derived cell lines (LCLs/PBMCs) for in vitro SSRI exposure studies.

3. Key Contributions

• Novel Phenotyping: Established a stringent, multi-dimensional classification system for antidepressant response that moves beyond binary "remission" to capture specific drug-class responses (SSRI vs. SNRI) and symptom-specific improvements.
• Patient-Centric Data: Successfully captured rich, self-reported data on symptom heterogeneity and treatment journeys that are often missing from EHRs alone.
• Identification of Subtypes: Identified a distinct subgroup of patients who failed multiple SSRIs but responded to SNRIs, characterized by a specific "somatic/anergic" symptom profile (e.g., high suicidality, fatigue, headaches).
• Infrastructure for Precision Psychiatry: Created a foundational dataset linking PROs, genomic data, and future EHR/molecular data, serving as a resource for future genetic and epigenetic discovery.

4. Key Results

• Demographics: The analytic sample (N=1,180) was predominantly female (78.1%), White (98.6%), and older (median age 57) than the wider GS cohort.
• Symptom Profile:
  • The most endorsed symptoms were lack of energy (95.1%), waking unrefreshed/fatigue (93.0%), and reduced social activity (90.8%).
  • Symptoms spanned six domains, highlighting the multi-dimensional nature of depression in this cohort.
• Medication Usage:
  • Polypharmacy: 31.1% of participants had tried three or more different antidepressants.
  • Drug Classes: SSRIs were the most common (89.1%), followed by Tricyclics (28.1%), SNRIs (21.1%), and Atypicals (17.7%).
  • Top Agents: Fluoxetine (47.1%), Sertraline (42.9%), and Citalopram (41.7%).
• Efficacy and Side Effects:
  • Efficacy: Sertraline had the highest proportion of "Yes, a lot" ratings (36.9%). SSRIs were generally rated most effective for mood and anxiety; SNRIs showed higher perceived impact on somatic symptoms.
  • Side Effects: Sleep disturbances (33.7%), weight changes (29.7%), and nausea (28.4%) were most common. Venlafaxine had the highest rate of side effect endorsement (43.7%).
• Response Classification:
  • Responders: 23.8% (281 participants) met strict SSRI responder criteria.
  • Non-Responders: 1.5% (18 participants) met strict SSRI non-responder criteria.
  • Unclassified: 74.7% (881 participants) did not meet the strict criteria for either extreme, reflecting the study's prioritization of phenotype certainty over sample size.
• SNRI Subgroup Discovery: A small group (N=9) of patients who failed SSRIs but responded to SNRIs (Venlafaxine/Duloxetine) exhibited a distinct phenotype: 100% reported suicidal thoughts, headaches/migraines, and anergia. This suggests a potential noradrenergic subtype of depression.

5. Significance and Future Directions

• Validation and Refinement: The current self-reported classifications will be validated against NHS EHRs to confirm diagnosis, prescription duration, and switching patterns, enhancing clinical validity.
• Biological Mechanisms: The study will proceed to collect biological samples from the extreme phenotypes (25 responders/25 non-responders). These will undergo:
  • DNA Methylation Profiling: To identify epigenetic signatures associated with response.
  • Cell Line Generation: Creation of lymphoblastoid cell lines (LCLs) and iPSCs to test gene expression changes in vitro upon SSRI exposure.
• Clinical Impact: By identifying distinct symptom clusters (e.g., the SNRI-responsive somatic/anergic subtype) and linking them to biological markers, the AMBER project aims to move psychiatry toward precision medicine, reducing trial-and-error prescribing and improving remission rates.
• Limitations: The study acknowledges potential recall bias in self-reported data and a demographic skew (older, female, White) that may limit generalizability, though the co-produced instrument and rigorous validation steps aim to mitigate these issues.

In conclusion, this paper establishes a robust, patient-centered framework for investigating antidepressant response, bridging the gap between subjective patient experience and objective biological mechanisms to advance the field of precision psychiatry.