Michelle Kamp
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Michelle Kamp has a total of 91 co-authors. Here are the top co-authors by number of shared publications:
- Cathryn M. Lewis
Social, Genetic & Developmental Psychiatry Centre, Institute of Psychiatry, Psychology and Neuroscience, King’s College London, London, United Kingdom
6 shared publications - Chris Wai Hang Lo
Social, Genetic & Developmental Psychiatry Centre, Institute of Psychiatry, Psychology and Neuroscience, King’s College London, London, United Kingdom
6 shared publications - Oliver Pain
Department of Basic and Clinical Neuroscience, Institute of Psychiatry, Psychology and Neuroscience, King’s College London, London, UK
5 shared publications - Alexandra Claire Gillett
Social, Genetic and Developmental Psychiatry Centre, Institute of Psychiatry, Psychology and Neuroscience, King’s College London
4 shared publications - Michèle Ramsay
Sydney Brenner Institute for Molecular Bioscience, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa
4 shared publications AMBER Research Team
3 shared publications
Latest Publications
- Open Access
Genetic architecture and polygenic prediction of self-reported response to selective serotonin reuptake inhibitors in individuals with depression: a genome-wide meta-analysis
Selective serotonin reuptake inhibitors (SSRIs) are first-line treatments for depression, but response is highly variable, and no robust genetic predictors have been identified. We conducted a genome-wide association meta-analysis of self-reported SSRI response in 62,879 individuals of European ancestry with depression (72.6% responders) across six cohorts (seven datasets), followed by gene-set, transcriptomic, proteomic, drug-target and polygenic score (PGS) analyses. SNP-based heritability was modest but significant (h² = 3.1%, SE = 1.4%) with no genome-wide significant variants detected. Gene-set analyses implicated gamma-aminobutyric acid (GABA)ergic and glutamatergic neurotransmission, and a gene involved in glutamate transport ( SLC17A6 ) showed consistent sub-threshold evidence across pathway, transcriptomic and proteomic analyses. Drug-target enrichment for hypnotics, anxiolytics and general anesthetics further supported the GABAergic signal. A leave-one-out PGS from the discovery meta-analysis was associated with self-reported SSRI response in independent cohorts (meta-analysis OR=1.05, 95% CI 1.02–1.08, p=0.002). Higher PGS for depression, attention-deficit/hyperactivity disorder (ADHD), anxiety, post-traumatic stress disorder (PTSD), autism, addiction and schizophrenia were associated with lower odds of reported SSRI response, with ADHD, PTSD, autism and addiction remaining after adjustment for depression PGS. In contrast, educational attainment PGS was associated with better SSRI response. Self-reported SSRI response captures genetic insights relevant to antidepressant response, implicating GABAergic and glutamatergic neurotransmission. Individuals with higher genetic loading for psychiatric conditions report poorer response, and higher genetic loading for educational attainment PGS a better reported response. Although larger studies are needed to identify specific variants, self-report offers a practical, patient-centered and scalable route to collect the larger samples required.
Open AccessGenetic architecture and polygenic prediction of self-reported response to selective serotonin reuptake inhibitors in individuals with depression: a genome-wide meta-analysis
Selective serotonin reuptake inhibitors (SSRIs) are first-line treatments for depression, but response is highly variable, and no robust genetic predictors have been identified. We conducted a genome-wide association meta-analysis of self-reported SSRI response in 62,879 individuals of European ancestry with depression (72.6% responders) across six cohorts (seven datasets), followed by gene-set, transcriptomic, proteomic, drug-target and polygenic score (PGS) analyses. SNP-based heritability was modest but significant (h² = 3.1%, SE = 1.4%) with no genome-wide significant variants detected. Gene-set analyses implicated gamma-aminobutyric acid (GABA)ergic and glutamatergic neurotransmission, and a gene involved in glutamate transport ( SLC17A6 ) showed consistent sub-threshold evidence across pathway, transcriptomic and proteomic analyses. Drug-target enrichment for hypnotics, anxiolytics and general anesthetics further supported the GABAergic signal. A leave-one-out PGS from the discovery meta-analysis was associated with self-reported SSRI response in independent cohorts (meta-analysis OR=1.05, 95% CI 1.02–1.08, p=0.002). Higher PGS for depression, attention-deficit/hyperactivity disorder (ADHD), anxiety, post-traumatic stress disorder (PTSD), autism, addiction and schizophrenia were associated with lower odds of reported SSRI response, with ADHD, PTSD, autism and addiction remaining after adjustment for depression PGS. In contrast, educational attainment PGS was associated with better SSRI response. Self-reported SSRI response captures genetic insights relevant to antidepressant response, implicating GABAergic and glutamatergic neurotransmission. Individuals with higher genetic loading for psychiatric conditions report poorer response, and higher genetic loading for educational attainment PGS a better reported response. Although larger studies are needed to identify specific variants, self-report offers a practical, patient-centered and scalable route to collect the larger samples required.
- Open Access
Clinical, sociodemographic, and genetic predictors of depressive episode duration in the UK Biobank
Background: The course of major depressive disorder is heterogeneous, with UK Biobank (UKB) participants reporting episode durations ranging from 24 months. Here, we identify predictors of episode duration, characterise its genetic architecture, and examine links to treatment seeking and response. Methods: In UKB participants meeting criteria for major depressive disorder, we examined clinical, sociodemographic, and genetic predictors of short (0-3 months) and long (>24 months) episode duration, fitted in predictor-specific, domain-level, and combined models. We also conducted genome-wide association studies in European-ancestry participants (n = 40,858) and estimated common-variant heritability. Results: Clinical features were most informative: higher childhood trauma scores, a stressful trigger, and recurrence showed the most consistent associations with short and long durations across models (OR combined : short = 0.75-0.95; long = 1.13-1.45; all p≤0.02). Higher neuroticism scores were also associated with both durations (OR combined : short = 0.977; long = 1.053; p<0.001). Polygenic risk for depression was associated with episode duration, though its independent contribution was modest. Long episodes were more predictable than short in validation analyses (AUC = 0.705 vs 0.601) and were associated with greater treatment engagement but lower perceived benefit; SNP-based heritability was nominally significant. Conclusions: Clinical features captured most of the predictable variance in episode duration, with the same predictors largely operating in opposite directions for short and long episodes, consistent with a continuum of chronicity. Those at risk for long episodes emerge as a priority for early identification and intervention.
Open AccessClinical, sociodemographic, and genetic predictors of depressive episode duration in the UK Biobank
Background: The course of major depressive disorder is heterogeneous, with UK Biobank (UKB) participants reporting episode durations ranging from 24 months. Here, we identify predictors of episode duration, characterise its genetic architecture, and examine links to treatment seeking and response. Methods: In UKB participants meeting criteria for major depressive disorder, we examined clinical, sociodemographic, and genetic predictors of short (0-3 months) and long (>24 months) episode duration, fitted in predictor-specific, domain-level, and combined models. We also conducted genome-wide association studies in European-ancestry participants (n = 40,858) and estimated common-variant heritability. Results: Clinical features were most informative: higher childhood trauma scores, a stressful trigger, and recurrence showed the most consistent associations with short and long durations across models (OR combined : short = 0.75-0.95; long = 1.13-1.45; all p≤0.02). Higher neuroticism scores were also associated with both durations (OR combined : short = 0.977; long = 1.053; p<0.001). Polygenic risk for depression was associated with episode duration, though its independent contribution was modest. Long episodes were more predictable than short in validation analyses (AUC = 0.705 vs 0.601) and were associated with greater treatment engagement but lower perceived benefit; SNP-based heritability was nominally significant. Conclusions: Clinical features captured most of the predictable variance in episode duration, with the same predictors largely operating in opposite directions for short and long episodes, consistent with a continuum of chronicity. Those at risk for long episodes emerge as a priority for early identification and intervention.
- Open Access
- 19 May 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
A Stratification Method for Identifying Subgroups at High-Risk for Type 2 Diabetes in sub-Saharan Africa
Type 2 diabetes is rising globally, with sub-Saharan Africa facing the steepest projected increase. In low-resource settings, limited screening contributes to high rates of undiagnosed disease, delaying intervention. Traditional risk models often rely on single-factor thresholds or population averages, which can miss combinations of demographic, anthropometric, and lifestyle factors relevant in African contexts. However, emerging evidence suggests that interactions between multiple risk factors provide more accurate, region-specific characterizations of risk. We applied a multidimensional subgroup discovery algorithm to cross-sectional data from three African populations and identified combinations of risk factor cutoffs that define high-risk subgroups. The most consistent profile, such as waist-to-hip ratio greater than 0.9, physical activity less than or equal to 2448 metabolic equivalent-minutes per week, and family history, shows elevated risk across regions and outperforms guideline-based definitions. Here we show that core anthropometric and familial factors are stable across populations, while others are region-specific, improving predictive performance for screening.
- 19 May 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Open AccessA Stratification Method for Identifying Subgroups at High-Risk for Type 2 Diabetes in sub-Saharan Africa
Type 2 diabetes is rising globally, with sub-Saharan Africa facing the steepest projected increase. In low-resource settings, limited screening contributes to high rates of undiagnosed disease, delaying intervention. Traditional risk models often rely on single-factor thresholds or population averages, which can miss combinations of demographic, anthropometric, and lifestyle factors relevant in African contexts. However, emerging evidence suggests that interactions between multiple risk factors provide more accurate, region-specific characterizations of risk. We applied a multidimensional subgroup discovery algorithm to cross-sectional data from three African populations and identified combinations of risk factor cutoffs that define high-risk subgroups. The most consistent profile, such as waist-to-hip ratio greater than 0.9, physical activity less than or equal to 2448 metabolic equivalent-minutes per week, and family history, shows elevated risk across regions and outperforms guideline-based definitions. Here we show that core anthropometric and familial factors are stable across populations, while others are region-specific, improving predictive performance for screening.
- Open Access
T-Rx: A toolbox for reproducible processing of prescriptions (Rx) from electronic health records
Linkage between population-wide biobanks and electronic health records (EHRs) opens new opportunities to study the genetic and epidemiological underpinnings of treatment outcomes, a key step forward in delivering precision medicine. However, challenges include complexities in data extraction for longitudinal analyses, and the absence of reproducible phenotyping algorithms. Here, we present T-Rx, an open-source R package to streamline the processing of prescription and dispensing records, to enable reproducible and scalable analysis across EHR databases. T-Rx consists of three modules to derive treatment-related phenotypes from uncleaned prescription records: (1) Extraction and Imputation Module for extracting and imputing prescription details; (2) Exposure Ascertainment Module for converting prescriptions to longitudinal exposure periods; and (3) Phenotyping Module for creating reproducible proxy phenotypes that capture treatment and response patterns. We tested the utility of T-Rx in UK Biobank primary care records, with strength and quantity information extracted and imputed for 2,721,921 antidepressant and 430,705 antipsychotic prescriptions. The extraction functions were validated using oral hypoglycemic agent prescriptions in Clinical Practice Research Datalink Aurum, showing comparable performances to extraction using the NHS Dictionary of Medicines and Devices (dm+d) codes. The Exposure Ascertainment Module of T-Rx converts discrete prescription or dispensing events into longitudinal exposure periods in one-line R commands, with customizable parameters to account for real-world treatment complexities. The Phenotyping Module takes prescriptions as direct user input and returns analysis -ready data frames. Current phenotyping algorithms include antidepressant switching and treatment-resistant depression. The phenotyping functions also allow flexible parameter choices, such as treatment episode windows, definitions for switching and quality control criteria. Researchers can contribute phenotyping algorithms to T-Rx for reproducible use. T-Rx improves the accessibility of prescription information in biobanks, and analysis of dosage- and treatment-patterns across therapeutic areas. T-Rx contributes to open science through harmonized phenotypic definitions and reproducible analyses of proxy treatment outcomes.
Open AccessT-Rx: A toolbox for reproducible processing of prescriptions (Rx) from electronic health records
Linkage between population-wide biobanks and electronic health records (EHRs) opens new opportunities to study the genetic and epidemiological underpinnings of treatment outcomes, a key step forward in delivering precision medicine. However, challenges include complexities in data extraction for longitudinal analyses, and the absence of reproducible phenotyping algorithms. Here, we present T-Rx, an open-source R package to streamline the processing of prescription and dispensing records, to enable reproducible and scalable analysis across EHR databases. T-Rx consists of three modules to derive treatment-related phenotypes from uncleaned prescription records: (1) Extraction and Imputation Module for extracting and imputing prescription details; (2) Exposure Ascertainment Module for converting prescriptions to longitudinal exposure periods; and (3) Phenotyping Module for creating reproducible proxy phenotypes that capture treatment and response patterns. We tested the utility of T-Rx in UK Biobank primary care records, with strength and quantity information extracted and imputed for 2,721,921 antidepressant and 430,705 antipsychotic prescriptions. The extraction functions were validated using oral hypoglycemic agent prescriptions in Clinical Practice Research Datalink Aurum, showing comparable performances to extraction using the NHS Dictionary of Medicines and Devices (dm+d) codes. The Exposure Ascertainment Module of T-Rx converts discrete prescription or dispensing events into longitudinal exposure periods in one-line R commands, with customizable parameters to account for real-world treatment complexities. The Phenotyping Module takes prescriptions as direct user input and returns analysis -ready data frames. Current phenotyping algorithms include antidepressant switching and treatment-resistant depression. The phenotyping functions also allow flexible parameter choices, such as treatment episode windows, definitions for switching and quality control criteria. Researchers can contribute phenotyping algorithms to T-Rx for reproducible use. T-Rx improves the accessibility of prescription information in biobanks, and analysis of dosage- and treatment-patterns across therapeutic areas. T-Rx contributes to open science through harmonized phenotypic definitions and reproducible analyses of proxy treatment outcomes.
- Open Access
- 10 July 2025
- Published by BMJ in BMJ Health & Care Informatics
- Vol. 32 (1) , e101294
Leveraging data science to understand and address multimorbidity in sub-Saharan Africa: the MADIVA protocol
Introduction: Multimorbidity (MM), defined as two or more chronic diseases in an individual, is linked to adverse outcomes. MM is increasing in sub-Saharan Africa due to rapidly advancing epidemiological and social transitions. The Multimorbidity in Africa: Digital Innovation, Visualisation and Application Research Hub (MADIVA) aims to address MM by developing data science solutions informed by stakeholder engagement. Methods and analysis: MADIVA uses complex, individual-level datasets from research centres in rural Bushbuckridge, South Africa and urban Nairobi, Kenya. These datasets will be harmonised, linked and curated, and then used to develop MM risk prediction models, novel data science methods and interactive dashboards for research and clinical use. Pilot projects and mentorship programmes will support data science capacity development. Ethics and dissemination: Ethics approval has been granted. Dissemination will occur through scientific meetings and publications. MADIVA is committed to making data FAIR: findable, accessible, interoperable and reusable.
- 10 July 2025
- Published by BMJ in BMJ Health & Care Informatics
- Vol. 32 (1) , e101294
Open AccessLeveraging data science to understand and address multimorbidity in sub-Saharan Africa: the MADIVA protocol
Introduction: Multimorbidity (MM), defined as two or more chronic diseases in an individual, is linked to adverse outcomes. MM is increasing in sub-Saharan Africa due to rapidly advancing epidemiological and social transitions. The Multimorbidity in Africa: Digital Innovation, Visualisation and Application Research Hub (MADIVA) aims to address MM by developing data science solutions informed by stakeholder engagement. Methods and analysis: MADIVA uses complex, individual-level datasets from research centres in rural Bushbuckridge, South Africa and urban Nairobi, Kenya. These datasets will be harmonised, linked and curated, and then used to develop MM risk prediction models, novel data science methods and interactive dashboards for research and clinical use. Pilot projects and mentorship programmes will support data science capacity development. Ethics and dissemination: Ethics approval has been granted. Dissemination will occur through scientific meetings and publications. MADIVA is committed to making data FAIR: findable, accessible, interoperable and reusable.
- Open Access
- 12 March 2025
- Published by Cambridge University Press (CUP) in Psychological Medicine
- Vol. 55, e80
Sociodemographic, clinical, and genetic factors associated with self-reported antidepressant response outcomes in the UK Biobank
Background: In major depressive disorder (MDD), only ~35% achieve remission after first-line antidepressant therapy. Using UK Biobank data, we identify sociodemographic, clinical, and genetic predictors of antidepressant response through self-reported outcomes, aiming to inform personalized treatment strategies. Methods: In UK Biobank Mental Health Questionnaire 2, participants with MDD reported whether specific antidepressants helped them. We tested whether retrospective lifetime response to four selective serotonin reuptake inhibitors (SSRIs) (N = 19,516) – citalopram (N = 8335), fluoxetine (N = 8476), paroxetine (N = 2297) and sertraline (N = 5883) – was associated with sociodemographic (e.g. age, gender) and clinical factors (e.g. episode duration). Genetic analyses evaluated the association between CYP2C19 variation and self-reported response, while polygenic score (PGS) analysis assessed whether genetic predisposition to psychiatric disorders and antidepressant response predicted self-reported SSRI outcomes. Results: 71%–77% of participants reported positive responses to SSRIs. Non-response was significantly associated with alcohol and illicit drug use (OR = 1.59, p = 2.23 × 10−20), male gender (OR = 1.25, p = 8.29 × 10−08), and lower-income (OR = 1.35, p = 4.22 × 10−07). The worst episode lasting over 2 years (OR = 1.93, p = 3.87 × 10−16) and no mood improvement from positive events (OR = 1.35, p = 2.37 × 10−07) were also associated with non-response. CYP2C19 poor metabolizers had nominally higher non-response rates (OR = 1.31, p = 1.77 × 10−02). Higher PGS for depression (OR = 1.08, p = 3.37 × 10−05) predicted negative SSRI outcomes after multiple testing corrections. Conclusions: Self-reported antidepressant response in the UK Biobank is influenced by sociodemographic, clinical, and genetic factors, mirroring clinical response measures. While positive outcomes are more frequent than remission reported in clinical trials, these self-reports replicate known treatment associations, suggesting they capture meaningful aspects of antidepressant effectiveness from the patient’s perspective.
- 12 March 2025
- Published by Cambridge University Press (CUP) in Psychological Medicine
- Vol. 55, e80
Open AccessSociodemographic, clinical, and genetic factors associated with self-reported antidepressant response outcomes in the UK Biobank
Background: In major depressive disorder (MDD), only ~35% achieve remission after first-line antidepressant therapy. Using UK Biobank data, we identify sociodemographic, clinical, and genetic predictors of antidepressant response through self-reported outcomes, aiming to inform personalized treatment strategies. Methods: In UK Biobank Mental Health Questionnaire 2, participants with MDD reported whether specific antidepressants helped them. We tested whether retrospective lifetime response to four selective serotonin reuptake inhibitors (SSRIs) (N = 19,516) – citalopram (N = 8335), fluoxetine (N = 8476), paroxetine (N = 2297) and sertraline (N = 5883) – was associated with sociodemographic (e.g. age, gender) and clinical factors (e.g. episode duration). Genetic analyses evaluated the association between CYP2C19 variation and self-reported response, while polygenic score (PGS) analysis assessed whether genetic predisposition to psychiatric disorders and antidepressant response predicted self-reported SSRI outcomes. Results: 71%–77% of participants reported positive responses to SSRIs. Non-response was significantly associated with alcohol and illicit drug use (OR = 1.59, p = 2.23 × 10−20), male gender (OR = 1.25, p = 8.29 × 10−08), and lower-income (OR = 1.35, p = 4.22 × 10−07). The worst episode lasting over 2 years (OR = 1.93, p = 3.87 × 10−16) and no mood improvement from positive events (OR = 1.35, p = 2.37 × 10−07) were also associated with non-response. CYP2C19 poor metabolizers had nominally higher non-response rates (OR = 1.31, p = 1.77 × 10−02). Higher PGS for depression (OR = 1.08, p = 3.37 × 10−05) predicted negative SSRI outcomes after multiple testing corrections. Conclusions: Self-reported antidepressant response in the UK Biobank is influenced by sociodemographic, clinical, and genetic factors, mirroring clinical response measures. While positive outcomes are more frequent than remission reported in clinical trials, these self-reports replicate known treatment associations, suggesting they capture meaningful aspects of antidepressant effectiveness from the patient’s perspective.
- Open Access
Antidepressant switching as a proxy phenotype for drug non-response: investigating clinical, demographic and genetic characteristics
Background: Selective serotonin reuptake inhibitors (SSRIs) are a first-line pharmacological therapy in major depressive disorder (MDD), but treatment response rates are low. Clinical trials lack the power to study the genetic contribution to SSRI response. Real-world evidence from electronic health records provides larger sample sizes, but novel response definitions are needed to accurately define SSRI non-responders. Methods: In UK Biobank (UKB) and Generation Scotland, SSRI switching was defined using a ≤ 90-day gap between prescriptions for an SSRI and another antidepressant in primary care. Non-switchers were participants with ≥ 3 consecutive prescriptions for an SSRI. In UKB, clinical, demographic and polygenic score (PGS) associations with switching were determined, and the common-variant heritability was estimated. Results: In UKB, 5,133 (13.2%) SSRI switchers and 33,680 non-switchers were defined. The mean time to switch was 28 days (IQR: 17-49). Switching patterns were consistent across UKB and Generation Scotland (n = 498 switchers). Higher annual income and educational levels (OR [95% CI] for university degree: 0.73 [0.67-0.79], compared to no qualifications) were associated with lower levels of switching. PGS for non-remission, based on clinical studies, were associated with increased risk of switching (OR: 1.07 [1.02-1.12], p=0.007). MDD PGS and family history of depression were not significantly associated with switching. Using GCTB, the heritability of SSRI switching was approximately 4% (SE: 0.016) on the observed scale. Conclusion: This study identified SSRI switching as a proxy of non-response, scalable across biobanks with EHR, capturing demographic and genetics of treatment non-response, and independent of MDD genetics.
Open AccessAntidepressant switching as a proxy phenotype for drug non-response: investigating clinical, demographic and genetic characteristics
Background: Selective serotonin reuptake inhibitors (SSRIs) are a first-line pharmacological therapy in major depressive disorder (MDD), but treatment response rates are low. Clinical trials lack the power to study the genetic contribution to SSRI response. Real-world evidence from electronic health records provides larger sample sizes, but novel response definitions are needed to accurately define SSRI non-responders. Methods: In UK Biobank (UKB) and Generation Scotland, SSRI switching was defined using a ≤ 90-day gap between prescriptions for an SSRI and another antidepressant in primary care. Non-switchers were participants with ≥ 3 consecutive prescriptions for an SSRI. In UKB, clinical, demographic and polygenic score (PGS) associations with switching were determined, and the common-variant heritability was estimated. Results: In UKB, 5,133 (13.2%) SSRI switchers and 33,680 non-switchers were defined. The mean time to switch was 28 days (IQR: 17-49). Switching patterns were consistent across UKB and Generation Scotland (n = 498 switchers). Higher annual income and educational levels (OR [95% CI] for university degree: 0.73 [0.67-0.79], compared to no qualifications) were associated with lower levels of switching. PGS for non-remission, based on clinical studies, were associated with increased risk of switching (OR: 1.07 [1.02-1.12], p=0.007). MDD PGS and family history of depression were not significantly associated with switching. Using GCTB, the heritability of SSRI switching was approximately 4% (SE: 0.016) on the observed scale. Conclusion: This study identified SSRI switching as a proxy of non-response, scalable across biobanks with EHR, capturing demographic and genetics of treatment non-response, and independent of MDD genetics.
- Open Access
Sociodemographic, clinical, and genetic factors associated with self-reported antidepressant response outcomes in the UK Biobank
Background: In major depressive disorder (MDD), only ∼35% achieve remission after first-line antidepressant therapy. Using UK Biobank data, we identify sociodemographic, clinical, and genetic predictors of antidepressant response through self-reported outcomes, aiming to inform personalized treatment strategies. Methods: In UK Biobank Mental Health Questionnaire 2, participants with MDD reported whether specific antidepressants helped them. We tested whether retrospective lifetime response to four selective serotonin reuptake inhibitors (SSRIs) (N=19,516)-citalopram (N=8,335), fluoxetine (N=8,476), paroxetine (N=2,297) and sertraline (N=5,883)-was associated with sociodemographic (e.g., age, gender) and clinical factors (e.g., episode duration). Genetic analyses evaluated the association between CYP2C19 variation and self-reported response, while polygenic score (PGS) analysis assessed whether genetic predisposition to psychiatric disorders and antidepressant response predicted self-reported SSRI outcomes. Results: 71%-77% of participants reported positive responses to an SSRI. Non-response was significantly associated with self-medication (OR=1.56, p=2.08×10 −19 ), younger age (OR=0.87, p=2.49×10 −11 ), male gender (OR=1.26, p=3.18×10 −08 ), and lower income (OR=1.34, p=6.17×10 −07 ). A worst episode lasting over two years (OR=1.97, p=1.25×10 −16 ) and no mood improvement from positive events (OR=1.33, p=1.35×10 −06 ) were also associated with non-response. CYP2C19 poor metabolizers had higher non-response rates (OR=1.31, p=1.77×10 −02 ). Additionally, higher PGS for depression (OR=1.08, p=3.37×10 −05 ), ADHD (OR=1.06, p=1.06×10 −03 ), and schizophrenia (OR =1.06, p=1.53×10 −03 ) predicted negative SSRI outcomes. Conclusion: Self-reported antidepressant response in the UK Biobank is influenced by sociodemographic, clinical, and genetic factors, mirroring clinical response measures. While positive outcomes are more frequent than remission reported in clinical trials, these self-reports replicate known treatment associations, suggesting they capture meaningful aspects of antidepressant effectiveness from the patient’s perspective.
Open AccessSociodemographic, clinical, and genetic factors associated with self-reported antidepressant response outcomes in the UK Biobank
Background: In major depressive disorder (MDD), only ∼35% achieve remission after first-line antidepressant therapy. Using UK Biobank data, we identify sociodemographic, clinical, and genetic predictors of antidepressant response through self-reported outcomes, aiming to inform personalized treatment strategies. Methods: In UK Biobank Mental Health Questionnaire 2, participants with MDD reported whether specific antidepressants helped them. We tested whether retrospective lifetime response to four selective serotonin reuptake inhibitors (SSRIs) (N=19,516)-citalopram (N=8,335), fluoxetine (N=8,476), paroxetine (N=2,297) and sertraline (N=5,883)-was associated with sociodemographic (e.g., age, gender) and clinical factors (e.g., episode duration). Genetic analyses evaluated the association between CYP2C19 variation and self-reported response, while polygenic score (PGS) analysis assessed whether genetic predisposition to psychiatric disorders and antidepressant response predicted self-reported SSRI outcomes. Results: 71%-77% of participants reported positive responses to an SSRI. Non-response was significantly associated with self-medication (OR=1.56, p=2.08×10 −19 ), younger age (OR=0.87, p=2.49×10 −11 ), male gender (OR=1.26, p=3.18×10 −08 ), and lower income (OR=1.34, p=6.17×10 −07 ). A worst episode lasting over two years (OR=1.97, p=1.25×10 −16 ) and no mood improvement from positive events (OR=1.33, p=1.35×10 −06 ) were also associated with non-response. CYP2C19 poor metabolizers had higher non-response rates (OR=1.31, p=1.77×10 −02 ). Additionally, higher PGS for depression (OR=1.08, p=3.37×10 −05 ), ADHD (OR=1.06, p=1.06×10 −03 ), and schizophrenia (OR =1.06, p=1.53×10 −03 ) predicted negative SSRI outcomes. Conclusion: Self-reported antidepressant response in the UK Biobank is influenced by sociodemographic, clinical, and genetic factors, mirroring clinical response measures. While positive outcomes are more frequent than remission reported in clinical trials, these self-reports replicate known treatment associations, suggesting they capture meaningful aspects of antidepressant effectiveness from the patient’s perspective.
- Open Access
- 26 August 2024
- Published by Springer Nature in Genome Medicine
- Vol. 16 (1) , 1-17
Ancestry-aligned polygenic scores combined with conventional risk factors improve prediction of cardiometabolic outcomes in African populations
Cardiovascular diseases (CVD) are a major health concern in Africa. Improved identification and treatment of high-risk individuals can reduce adverse health outcomes. Current CVD risk calculators are largely unvalidated in African populations and overlook genetic factors. Polygenic scores (PGS) can enhance risk prediction by measuring genetic susceptibility to CVD, but their effectiveness in genetically diverse populations is limited by a European-ancestry bias. To address this, we developed models integrating genetic data and conventional risk factors to assess the risk of developing cardiometabolic outcomes in African populations. We used summary statistics from a genome-wide association meta-analysis (n = 14,126) in African populations to derive novel genome-wide PGS for 14 cardiometabolic traits in an independent African target sample (Africa Wits-INDEPTH Partnership for Genomic Research (AWI-Gen), n = 10,603). Regression analyses assessed relationships between each PGS and corresponding cardiometabolic trait, and seven CVD outcomes (CVD, heart attack, stroke, diabetes mellitus, dyslipidaemia, hypertension, and obesity). The predictive utility of the genetic data was evaluated using elastic net models containing multiple PGS (MultiPGS) and reference-projected principal components of ancestry (PPCs). An integrated risk prediction model incorporating genetic and conventional risk factors was developed. Nested cross-validation was used when deriving elastic net models to enhance generalisability. Our African-specific PGS displayed significant but variable within- and cross- trait prediction (max.R2 = 6.8%, p = 1.86 × 10−173). Significantly associated PGS with dyslipidaemia included the PGS for total cholesterol (logOR = 0.210, SE = 0.022, p = 2.18 × 10−21) and low-density lipoprotein (logOR = − 0.141, SE = 0.022, p = 1.30 × 10−20); with hypertension, the systolic blood pressure PGS (logOR = 0.150, SE = 0.045, p = 8.34 × 10−4); and multiple PGS associated with obesity: body mass index (max. logOR = 0.131, SE = 0.031, p = 2.22 × 10−5), hip circumference (logOR = 0.122, SE = 0.029, p = 2.28 × 10−5), waist circumference (logOR = 0.013, SE = 0.098, p = 8.13 × 10−4) and weight (logOR = 0.103, SE = 0.029, p = 4.89 × 10−5). Elastic net models incorporating MultiPGS and PPCs significantly improved prediction over MultiPGS alone. Models including genetic data and conventional risk factors were more predictive than conventional risk models alone (dyslipidaemia: R2 increase = 2.6%, p = 4.45 × 10−12; hypertension: R2 increase = 2.6%, p = 2.37 × 10−13; obesity: R2 increase = 5.5%, 1.33 × 10−34). In African populations, CVD and associated cardiometabolic trait prediction models can be improved by incorporating ancestry-aligned PGS and accounting for ancestry. Combining PGS with conventional risk factors further enhances prediction over traditional models based on conventional factors. Incorporating data from target populations can improve the generalisability of international predictive models for CVD and associated traits in African populations.
- 26 August 2024
- Published by Springer Nature in Genome Medicine
- Vol. 16 (1) , 1-17
Open AccessAncestry-aligned polygenic scores combined with conventional risk factors improve prediction of cardiometabolic outcomes in African populations
Cardiovascular diseases (CVD) are a major health concern in Africa. Improved identification and treatment of high-risk individuals can reduce adverse health outcomes. Current CVD risk calculators are largely unvalidated in African populations and overlook genetic factors. Polygenic scores (PGS) can enhance risk prediction by measuring genetic susceptibility to CVD, but their effectiveness in genetically diverse populations is limited by a European-ancestry bias. To address this, we developed models integrating genetic data and conventional risk factors to assess the risk of developing cardiometabolic outcomes in African populations. We used summary statistics from a genome-wide association meta-analysis (n = 14,126) in African populations to derive novel genome-wide PGS for 14 cardiometabolic traits in an independent African target sample (Africa Wits-INDEPTH Partnership for Genomic Research (AWI-Gen), n = 10,603). Regression analyses assessed relationships between each PGS and corresponding cardiometabolic trait, and seven CVD outcomes (CVD, heart attack, stroke, diabetes mellitus, dyslipidaemia, hypertension, and obesity). The predictive utility of the genetic data was evaluated using elastic net models containing multiple PGS (MultiPGS) and reference-projected principal components of ancestry (PPCs). An integrated risk prediction model incorporating genetic and conventional risk factors was developed. Nested cross-validation was used when deriving elastic net models to enhance generalisability. Our African-specific PGS displayed significant but variable within- and cross- trait prediction (max.R2 = 6.8%, p = 1.86 × 10−173). Significantly associated PGS with dyslipidaemia included the PGS for total cholesterol (logOR = 0.210, SE = 0.022, p = 2.18 × 10−21) and low-density lipoprotein (logOR = − 0.141, SE = 0.022, p = 1.30 × 10−20); with hypertension, the systolic blood pressure PGS (logOR = 0.150, SE = 0.045, p = 8.34 × 10−4); and multiple PGS associated with obesity: body mass index (max. logOR = 0.131, SE = 0.031, p = 2.22 × 10−5), hip circumference (logOR = 0.122, SE = 0.029, p = 2.28 × 10−5), waist circumference (logOR = 0.013, SE = 0.098, p = 8.13 × 10−4) and weight (logOR = 0.103, SE = 0.029, p = 4.89 × 10−5). Elastic net models incorporating MultiPGS and PPCs significantly improved prediction over MultiPGS alone. Models including genetic data and conventional risk factors were more predictive than conventional risk models alone (dyslipidaemia: R2 increase = 2.6%, p = 4.45 × 10−12; hypertension: R2 increase = 2.6%, p = 2.37 × 10−13; obesity: R2 increase = 5.5%, 1.33 × 10−34). In African populations, CVD and associated cardiometabolic trait prediction models can be improved by incorporating ancestry-aligned PGS and accounting for ancestry. Combining PGS with conventional risk factors further enhances prediction over traditional models based on conventional factors. Incorporating data from target populations can improve the generalisability of international predictive models for CVD and associated traits in African populations.
- Open Access
- 7 December 2023
- Published by BMJ in BMJ Global Health
- Vol. 8 (12) , e013509
Multimorbidity in African ancestry populations: a scoping review
Objectives Multimorbidity (MM) is a growing concern linked to poor outcomes and higher healthcare costs. While most MM research targets European ancestry populations, the prevalence and patterns in African ancestry groups remain underexplored. This study aimed to identify and summarise the available literature on MM in populations with African ancestry, on the continent, and in the diaspora. Design A scoping review was conducted in five databases (PubMed, Web of Science, Scopus, Science Direct and JSTOR) in July 2022. Studies were selected based on predefined criteria, with data extraction focusing on methodology and findings. Descriptive statistics summarised the data, and a narrative synthesis highlighted key themes. Results Of the 232 publications on MM in African-ancestry groups from 2010 to June 2022—113 examined continental African populations, 100 the diaspora and 19 both. Findings revealed diverse MM patterns within and beyond continental Africa. Cardiovascular and metabolic diseases are predominant in both groups (80% continental and 70% diaspora). Infectious diseases featured more in continental studies (58% continental and 16% diaspora). Although many papers did not specifically address these features, as in previous studies, older age, being women and having a lower socioeconomic status were associated with a higher prevalence of MM, with important exceptions. Research gaps identified included limited data on African-ancestry individuals, inadequate representation, under-represented disease groups, non-standardised methodologies, the need for innovative data strategies, and insufficient translational research. Conclusion The growing global MM prevalence is mirrored in African-ancestry populations. Recognising the unique contexts of African-ancestry populations is essential when addressing the burden of MM. This review emphasises the need for additional research to guide and enhance healthcare approaches for African-ancestry populations, regardless of their geographic location.
- 7 December 2023
- Published by BMJ in BMJ Global Health
- Vol. 8 (12) , e013509
Open AccessMultimorbidity in African ancestry populations: a scoping review
Objectives Multimorbidity (MM) is a growing concern linked to poor outcomes and higher healthcare costs. While most MM research targets European ancestry populations, the prevalence and patterns in African ancestry groups remain underexplored. This study aimed to identify and summarise the available literature on MM in populations with African ancestry, on the continent, and in the diaspora. Design A scoping review was conducted in five databases (PubMed, Web of Science, Scopus, Science Direct and JSTOR) in July 2022. Studies were selected based on predefined criteria, with data extraction focusing on methodology and findings. Descriptive statistics summarised the data, and a narrative synthesis highlighted key themes. Results Of the 232 publications on MM in African-ancestry groups from 2010 to June 2022—113 examined continental African populations, 100 the diaspora and 19 both. Findings revealed diverse MM patterns within and beyond continental Africa. Cardiovascular and metabolic diseases are predominant in both groups (80% continental and 70% diaspora). Infectious diseases featured more in continental studies (58% continental and 16% diaspora). Although many papers did not specifically address these features, as in previous studies, older age, being women and having a lower socioeconomic status were associated with a higher prevalence of MM, with important exceptions. Research gaps identified included limited data on African-ancestry individuals, inadequate representation, under-represented disease groups, non-standardised methodologies, the need for innovative data strategies, and insufficient translational research. Conclusion The growing global MM prevalence is mirrored in African-ancestry populations. Recognising the unique contexts of African-ancestry populations is essential when addressing the burden of MM. This review emphasises the need for additional research to guide and enhance healthcare approaches for African-ancestry populations, regardless of their geographic location.