Katherine G. Young
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Katherine G. Young has a total of 181 co-authors. Here are the top co-authors by number of shared publications:
- John M. Dennis
Institute of Clinical & Biomedical Sciences, University of Exeter Medical School, Exeter, UK
22 shared publications - Andrew Tym Hattersley
Department of Clinical and Biomedical Sciences, Faculty of Health and Life Sciences, University of Exeter, Exeter, UK
17 shared publications - Beverley M. Shields
Department of Clinical and Biomedical Sciences University of Exeter Medical School Exeter UK
17 shared publications - Angus G. Jones
Department of Clinical and Biomedical Sciences, University of Exeter, Exeter, UK
15 shared publications - Andrew Peter McGovern
Clinical and Biomedical Sciences, University of Exeter Medical School, Exeter, UK, EX2 5GL
11 shared publications - Ewan Robert Pearson
Population Health & Genomics, School of Medicine, University of Dundee, Dundee, UK
11 shared publications
Latest Publications
- Open Access
- 14 October 2026
- Published by Wiley in Diabetic Medicine
- p. e70493
The DePICtion study: Implementing a diabetes classification tool in UK primary care—A qualitative study with primary care staff and patients
Background: The correct classification of diabetes (type 1, type 2 or rarer genetic forms) is essential for optimal treatment, but misclassification is common. The Diabetes in Primary Care Improving Classification (DePICtion) tool uses routinely collected clinical data to flag patients at risk of misclassification and prioritise them for further diagnostic testing. Aim: To explore the acceptability and implementation of the DePICtion tool in primary care. Design and Setting: A qualitative study in 11 general practices across the Southwest and East Midlands, UK. Method: Twelve practices were recruited to the study; 11 practices ran the tool and reviewed flagged patients. Semi‐structured interviews were conducted with 10 staff and 11 patients to explore their experiences of the tool, reactions to the results and barriers and facilitators to use. Data were analysed thematically in NVivo. Results: Staff found the tool straightforward, compatible with existing workflows and clinically useful. Implementation required time and ownership of the process. Patients' responses were shaped by trust in staff and the personal meaning of their diabetes: Some viewed reclassification as a threat to their diagnostic identity, while others welcomed further diagnostic testing. Data privacy concerns were minimal but competing demands on staff time and resources were ubiquitously reported. Conclusion: The DePICtion tool is acceptable and clinically useful for improving diabetes classification in primary care. Routine implementation is technically feasible but will require protected time, clear pathways for follow‐up and support for patient communication and engagement.
- 14 October 2026
- Published by Wiley in Diabetic Medicine
- p. e70493
Open AccessThe DePICtion study: Implementing a diabetes classification tool in UK primary care—A qualitative study with primary care staff and patients
Background: The correct classification of diabetes (type 1, type 2 or rarer genetic forms) is essential for optimal treatment, but misclassification is common. The Diabetes in Primary Care Improving Classification (DePICtion) tool uses routinely collected clinical data to flag patients at risk of misclassification and prioritise them for further diagnostic testing. Aim: To explore the acceptability and implementation of the DePICtion tool in primary care. Design and Setting: A qualitative study in 11 general practices across the Southwest and East Midlands, UK. Method: Twelve practices were recruited to the study; 11 practices ran the tool and reviewed flagged patients. Semi‐structured interviews were conducted with 10 staff and 11 patients to explore their experiences of the tool, reactions to the results and barriers and facilitators to use. Data were analysed thematically in NVivo. Results: Staff found the tool straightforward, compatible with existing workflows and clinically useful. Implementation required time and ownership of the process. Patients' responses were shaped by trust in staff and the personal meaning of their diabetes: Some viewed reclassification as a threat to their diagnostic identity, while others welcomed further diagnostic testing. Data privacy concerns were minimal but competing demands on staff time and resources were ubiquitously reported. Conclusion: The DePICtion tool is acceptable and clinically useful for improving diabetes classification in primary care. Routine implementation is technically feasible but will require protected time, clear pathways for follow‐up and support for patient communication and engagement.
- Open Access
A consensus diabetes core dataset for research using NHS data: outputs from a Diabetes Data Science Catalyst workshop
Aims: To develop a ‘core’ dataset of diabetes related variables to support reproducible research using UK routinely collected health data. Methods: A workshop was conducted bringing together diabetes healthcare professionals, researchers, and patient and public representatives to discuss and prioritise variables for inclusion in the Diabetes Core Dataset. Core variables were those considered to be highest priority for diabetes research and available at high quality in NHS data routinely used for research (primary care [GP] and Hospital Episode Statistics [HES] data). Candidate variables for inclusion in the Diabetes Core Dataset were from a review of existing core datasets and expert opinion. Participants scored variables anonymously based on priority for diabetes research. Results: 25 variables from existing diabetes core datasets and 87 other candidate variables were considered for inclusion in the Diabetes Core Dataset. All 25 of those from existing diabetes core datasets and 5 of the 87 candidate variables met the core requirements for inclusion. In addition, 7 variables were identified as high priority but not included in the core dataset as they are not currently available in GP/HES data; these were labelled as ‘future high priority’ variables for diabetes research. Conclusions: A new diabetes core dataset for UK EHR research has been developed using a consensus-based process. The core dataset is openly available and can be flexibly applied in UK EHR ( https://healthdatagateway.org/en/tool/426 ), including in new NHS Research Secure Data Environment platforms, to enhance reproducible research to improve the clinical care of people with diabetes and associated conditions.
Open AccessA consensus diabetes core dataset for research using NHS data: outputs from a Diabetes Data Science Catalyst workshop
Aims: To develop a ‘core’ dataset of diabetes related variables to support reproducible research using UK routinely collected health data. Methods: A workshop was conducted bringing together diabetes healthcare professionals, researchers, and patient and public representatives to discuss and prioritise variables for inclusion in the Diabetes Core Dataset. Core variables were those considered to be highest priority for diabetes research and available at high quality in NHS data routinely used for research (primary care [GP] and Hospital Episode Statistics [HES] data). Candidate variables for inclusion in the Diabetes Core Dataset were from a review of existing core datasets and expert opinion. Participants scored variables anonymously based on priority for diabetes research. Results: 25 variables from existing diabetes core datasets and 87 other candidate variables were considered for inclusion in the Diabetes Core Dataset. All 25 of those from existing diabetes core datasets and 5 of the 87 candidate variables met the core requirements for inclusion. In addition, 7 variables were identified as high priority but not included in the core dataset as they are not currently available in GP/HES data; these were labelled as ‘future high priority’ variables for diabetes research. Conclusions: A new diabetes core dataset for UK EHR research has been developed using a consensus-based process. The core dataset is openly available and can be flexibly applied in UK EHR ( https://healthdatagateway.org/en/tool/426 ), including in new NHS Research Secure Data Environment platforms, to enhance reproducible research to improve the clinical care of people with diabetes and associated conditions.
- Open Access
- 29 July 2026
- Published by Elsevier in SSRN Electronic Journal
Implementing a Clinical Prediction Model in UK Primary Care Data to Identify Individuals with Misclassified Diabetes: A Population Based Retrospective Cohort Study
Background: Correct classification of type 1 (T1) and type 2 (T2) diabetes is important as treatment differs. Misclassification is common, especially in younger adults, yet biomarker testing to confirm subtype (autoantibodies/C-peptide) is rare in primary care. We evaluated whether a validated clinical prediction model estimating T1 probability (T1/T2 model) could be applied in UK primary care data to identify individuals at high risk for misclassification. Methods: The T1/T2 model was applied to insulin-treated adults with diabetes diagnosed aged 18-50 years, registered on 01/03/2023 in UK primary care data (Clinical Practice Research Datalink). Potential misclassification was identified where the patient’s most recent type-specific diabetes code was discordant with model probability based on predefined thresholds (≥80% for T1; <5% for T2), with potential miscoding considered if the record included both T1 and T2 codes. Outcomes associated with T1 (insulin deficiency) were compared between potentially misclassified and correctly classified (model probability concordant with diabetes type): hospitalisation for diabetic ketoacidosis (DKA) or hypoglycaemia, early insulin initiation (≤1 year of diagnosis), and basal-bolus insulin use. Findings: Model inputs were available for 93.6% of eligible individuals (n=62,016/66,273). 158 (0.6%) of 43,180 with a current T2 diagnosis were identified as potentially misclassified. Compared to correctly classified T2, this group had higher hospitalisation for DKA (0.45 vs 0.17 per 100 person-years, p<0.001), and more initiated insulin early and used a basal-bolus insulin regime. 1,110 (7.8%) of 18,836 with a current T1 diagnosis were identified as potentially misclassified. Compared to correctly classified T1, fewer initiated insulin early or used a basal-bolus regime. Findings were consistent across demographic subgroups. Interpretation: A simple model using routine primary care data can identify individuals with features suggestive of diabetes misclassification, supporting targeted biomarker testing to confirm diabetes subtype and data quality review in primary care.
- 29 July 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessImplementing a Clinical Prediction Model in UK Primary Care Data to Identify Individuals with Misclassified Diabetes: A Population Based Retrospective Cohort Study
Background: Correct classification of type 1 (T1) and type 2 (T2) diabetes is important as treatment differs. Misclassification is common, especially in younger adults, yet biomarker testing to confirm subtype (autoantibodies/C-peptide) is rare in primary care. We evaluated whether a validated clinical prediction model estimating T1 probability (T1/T2 model) could be applied in UK primary care data to identify individuals at high risk for misclassification. Methods: The T1/T2 model was applied to insulin-treated adults with diabetes diagnosed aged 18-50 years, registered on 01/03/2023 in UK primary care data (Clinical Practice Research Datalink). Potential misclassification was identified where the patient’s most recent type-specific diabetes code was discordant with model probability based on predefined thresholds (≥80% for T1; <5% for T2), with potential miscoding considered if the record included both T1 and T2 codes. Outcomes associated with T1 (insulin deficiency) were compared between potentially misclassified and correctly classified (model probability concordant with diabetes type): hospitalisation for diabetic ketoacidosis (DKA) or hypoglycaemia, early insulin initiation (≤1 year of diagnosis), and basal-bolus insulin use. Findings: Model inputs were available for 93.6% of eligible individuals (n=62,016/66,273). 158 (0.6%) of 43,180 with a current T2 diagnosis were identified as potentially misclassified. Compared to correctly classified T2, this group had higher hospitalisation for DKA (0.45 vs 0.17 per 100 person-years, p<0.001), and more initiated insulin early and used a basal-bolus insulin regime. 1,110 (7.8%) of 18,836 with a current T1 diagnosis were identified as potentially misclassified. Compared to correctly classified T1, fewer initiated insulin early or used a basal-bolus regime. Findings were consistent across demographic subgroups. Interpretation: A simple model using routine primary care data can identify individuals with features suggestive of diabetes misclassification, supporting targeted biomarker testing to confirm diabetes subtype and data quality review in primary care.
- Open Access
- 20 May 2026
- Published by American Diabetes Association in Diabetes Care
- Vol. 49 (6) , 1115-1123
Precision Prescribing of SGLT2 Inhibitors in Individuals With Type 2 Diabetes for Primary Prevention of Heart Failure: Model Development and Validation Study
OBJECTIVE: Sodium–glucose cotransporter 2 inhibitors (SGLT2i) reduce heart failure (HF) risk in type 2 diabetes (T2D) and are recommended for patients with T2D who have atherosclerotic cardiovascular disease (ASCVD), HF, or chronic kidney disease (CKD). However, most individuals with T2D do not have these conditions, and current guidelines for this group do not indicate which individuals may benefit most from SGLT2i. We aimed to develop and validate a model to predict the individual-level HF benefit of SGLT2i in individuals with T2D without ASCVD, HF, or CKD. RESEARCH DESIGN AND METHODS: We developed the SGLT2i Absolute Benefit Response (SABRE) model, combining absolute HF risk from the validated QDiabetes-HF model with the SGLT2i-associated hazard ratio (HR) for HF hospitalization from a trial meta-analysis (HR 0.63) to estimate individual 5-year HF benefit. Model components and predictions were validated using U.K. primary care data with linked hospital and death records from 2013 to 2020. RESULTS: Among 57,368 SGLT2i initiators and 111,673 comparator (dipeptidyl peptidase 4 inhibitor or sulfonylurea) initiators, SGLT2i use was associated with a 30% lower risk of new-onset HF (HR 0.70 [95% CI 0.63–0.78]), consistent with trial evidence. Relative HF benefit did not vary by baseline absolute HF risk (P = 0.82). The SABRE model–predicted 5-year absolute HF benefit with SGLT2i ranged from <0.1% to 14.1% (median 1.0% [interquartile range 0.6–1.8%]) and calibrated well against observed HF outcomes. SABRE provided more targeted HF prevention than current guidelines in those with T2D without ASCVD, HF, or CKD. CONCLUSIONS: The SABRE model is an easily deployed clinical prediction model integrating trial evidence and allowing more precise targeting of SGLT2i for primary HF prevention in T2D.
- 20 May 2026
- Published by American Diabetes Association in Diabetes Care
- Vol. 49 (6) , 1115-1123
Open AccessPrecision Prescribing of SGLT2 Inhibitors in Individuals With Type 2 Diabetes for Primary Prevention of Heart Failure: Model Development and Validation Study
OBJECTIVE: Sodium–glucose cotransporter 2 inhibitors (SGLT2i) reduce heart failure (HF) risk in type 2 diabetes (T2D) and are recommended for patients with T2D who have atherosclerotic cardiovascular disease (ASCVD), HF, or chronic kidney disease (CKD). However, most individuals with T2D do not have these conditions, and current guidelines for this group do not indicate which individuals may benefit most from SGLT2i. We aimed to develop and validate a model to predict the individual-level HF benefit of SGLT2i in individuals with T2D without ASCVD, HF, or CKD. RESEARCH DESIGN AND METHODS: We developed the SGLT2i Absolute Benefit Response (SABRE) model, combining absolute HF risk from the validated QDiabetes-HF model with the SGLT2i-associated hazard ratio (HR) for HF hospitalization from a trial meta-analysis (HR 0.63) to estimate individual 5-year HF benefit. Model components and predictions were validated using U.K. primary care data with linked hospital and death records from 2013 to 2020. RESULTS: Among 57,368 SGLT2i initiators and 111,673 comparator (dipeptidyl peptidase 4 inhibitor or sulfonylurea) initiators, SGLT2i use was associated with a 30% lower risk of new-onset HF (HR 0.70 [95% CI 0.63–0.78]), consistent with trial evidence. Relative HF benefit did not vary by baseline absolute HF risk (P = 0.82). The SABRE model–predicted 5-year absolute HF benefit with SGLT2i ranged from <0.1% to 14.1% (median 1.0% [interquartile range 0.6–1.8%]) and calibrated well against observed HF outcomes. SABRE provided more targeted HF prevention than current guidelines in those with T2D without ASCVD, HF, or CKD. CONCLUSIONS: The SABRE model is an easily deployed clinical prediction model integrating trial evidence and allowing more precise targeting of SGLT2i for primary HF prevention in T2D.
- Open Access
Using routine clinical features to classify adult-onset diabetes at diagnosis: the StartRight prospective observational study
Background It is not known which clinical features optimally differentiate type 1 and 2 diabetes at diagnosis. We aimed to determine which clinical features differentiate these conditions at diagnosis and develop classification models combining clinical features with and without islet-autoantibodies. Methods In this prospective cohort study, we recruited 1,800 adults (aged ≥18 years) diagnosed with diabetes within the preceding 12 months, excluding secondary or monogenic diabetes. The primary outcome was diabetes subtype defined by a combination of insulin treatment and endogenous insulin production (measured by C-peptide) at ≥3 years post-diagnosis. Models were developed in participants aged 18 to 50 years and validated internally, alongside validation in an older cohort (aged >50 years) and using UK primary care data (n=188,232). Results Eleven clinical features and routinely measured biomarkers discriminated type 1 from type 2 diabetes independently of diagnosis age and BMI. Lower age of diagnosis, BMI and waist-hip ratio, unintentional weight-loss, and higher presentation HbA1c or glucose were the most discriminative, with other features only weakly discriminative. Models integrating routine features with and without islet-autoantibodies, developed in those age 18 to 50 years at diabetes diagnosis, had high performance in internal validation (clinical features only: Area Under the Receiver Operating Characteristic curve (AUCROC) (95% CI) 0.94 (0.93, 0.96), clinical features and islet-autoantibodies: AUCROC 0.97 (0.96, 0.98)), and maintained high discrimination in older adults (age >50 years AUCROC 0.93 (0.90, 0.96), and 0.97 (0.94, 0.99)). Simplifying the models to a points-based score resulted in similar performance. In primary care data models and score were strongly predictive of outcomes associated with type 1 diabetes, including in those initially treated as type 2. Interpretation Lower age-at-diagnosis, BMI, and wait-hip ratio, unintentional weight loss and high presentation glycaemia are the most discriminative features for diagnosis of type 1 diabetes in adults. Models combining routine clinical features, with or without islet-autoantibodies, have high accuracy and could assist clinical classification and prioritisation of classification biomarker testing. Funding UK National Institute of Health and Care Research (NIHR) and Diabetes UK.
Open AccessUsing routine clinical features to classify adult-onset diabetes at diagnosis: the StartRight prospective observational study
Background It is not known which clinical features optimally differentiate type 1 and 2 diabetes at diagnosis. We aimed to determine which clinical features differentiate these conditions at diagnosis and develop classification models combining clinical features with and without islet-autoantibodies. Methods In this prospective cohort study, we recruited 1,800 adults (aged ≥18 years) diagnosed with diabetes within the preceding 12 months, excluding secondary or monogenic diabetes. The primary outcome was diabetes subtype defined by a combination of insulin treatment and endogenous insulin production (measured by C-peptide) at ≥3 years post-diagnosis. Models were developed in participants aged 18 to 50 years and validated internally, alongside validation in an older cohort (aged >50 years) and using UK primary care data (n=188,232). Results Eleven clinical features and routinely measured biomarkers discriminated type 1 from type 2 diabetes independently of diagnosis age and BMI. Lower age of diagnosis, BMI and waist-hip ratio, unintentional weight-loss, and higher presentation HbA1c or glucose were the most discriminative, with other features only weakly discriminative. Models integrating routine features with and without islet-autoantibodies, developed in those age 18 to 50 years at diabetes diagnosis, had high performance in internal validation (clinical features only: Area Under the Receiver Operating Characteristic curve (AUCROC) (95% CI) 0.94 (0.93, 0.96), clinical features and islet-autoantibodies: AUCROC 0.97 (0.96, 0.98)), and maintained high discrimination in older adults (age >50 years AUCROC 0.93 (0.90, 0.96), and 0.97 (0.94, 0.99)). Simplifying the models to a points-based score resulted in similar performance. In primary care data models and score were strongly predictive of outcomes associated with type 1 diabetes, including in those initially treated as type 2. Interpretation Lower age-at-diagnosis, BMI, and wait-hip ratio, unintentional weight loss and high presentation glycaemia are the most discriminative features for diagnosis of type 1 diabetes in adults. Models combining routine clinical features, with or without islet-autoantibodies, have high accuracy and could assist clinical classification and prioritisation of classification biomarker testing. Funding UK National Institute of Health and Care Research (NIHR) and Diabetes UK.
- Open Access
- 16 April 2026
- Published by Elsevier in The Lancet Primary Care
- Vol. 2 (4)
Kidney outcomes with GLP-1 receptor agonists in people with type 2 diabetes already receiving SGLT2 inhibitors: a target trial emulation study using UK primary care data
- 16 April 2026
- Published by Elsevier in The Lancet Primary Care
- Vol. 2 (4)
Open AccessKidney outcomes with GLP-1 receptor agonists in people with type 2 diabetes already receiving SGLT2 inhibitors: a target trial emulation study using UK primary care data
- Open Access
- 8 April 2026
- Published by Elsevier in Primary Care Diabetes
- Vol. 20 (3) , 335-343
Impact of depression on treatment progression in type 2 diabetes: A UK retrospective cohort study using the Clinical Practice Research Datalink Aurum database
- 8 April 2026
- Published by Elsevier in Primary Care Diabetes
- Vol. 20 (3) , 335-343
Open AccessImpact of depression on treatment progression in type 2 diabetes: A UK retrospective cohort study using the Clinical Practice Research Datalink Aurum database
- Open Access
Trends in second-line glucose-lowering therapy initiations and treatment outcomes across age and frailty groups in people with type 2 diabetes: UK population-based study, 2019-2024
Objective To assess contemporary UK population-level trends in second-line glucose-lowering therapy initiation and explore treatment outcome trends across age and frailty groups in people with type 2 diabetes (T2D). Methods and Analysis We studied 117,064 adults with T2D who initiated major second-line glucose-lowering drug classes after metformin between 2019 to 2024 in population-representative UK data (Clinical Practice Research Datalink). Outcomes were assessed in three groups: age ≤70 years (n=84,589 [72%]), non-frail >70 years (n=18,933 [16%]), and frail >70 years (n=13,524 [12%]), with frailty specified as an electronic frailty index ≥ 0.24. Annual prescribing trends were calculated as the proportion of each drug class among all second-line initiations. We evaluated 12-month changes in HbA1c, weight, treatment discontinuation, and complications, including composite severe diabetes-related complications, heart failure and kidney failure. Trends were estimated using multivariable regression models adjusted for demographic and clinical characteristics. Results Use of SGLT2-inhibitors increased from 26% of overall second-line initiations in 2019 to 63% in 2024, with the greatest increase among frail adults >70 (6% in 2019 vs 60% in 2024). DPP4-inhibitor use declined substantially over the same period (44% 2019 vs 16% in 2024). In adults aged ≤70 years, mean 12-month HbA1c response improved from -11.3 mmol/mol (95%CI -11.6 to -11.0) in 2019 to -12.9 mmol/mol (95%CI -13.2 to -12.6) in 2023, and mean weight loss improved from -1.7 kg (95%CI -1.8 to -1.5) to -2.8 kg (95%CI -2.9 to -2.6). Treatment discontinuation within 12 months remained stable (30.2% [95%CI 29.4 to 31.0] in 2019 vs 30.7% [95%CI 30.0 to 31.4] in 2022). There was limited evidence of change in short-term complications over 2019-2022: rates (per 1,000 person-years) of severe diabetes-related complications were 9.9 (95%CI 8.1 to 11.7) vs 8.6 (95%CI 6.9 to 10.3), heart failure 2.0 (95%CI 1.1 to 2.9) vs 2.7 (95%CI 1.8 to 3.7), and kidney failure 2.1 (95%CI 1.1 to 3.0) vs 1.2 (95%CI 0.5 to 1.8). All trends were consistent in adults >70, including those with frailty. Conclusion Use SGLT2-inhibitors in people with T2D has rapidly increased in the UK over 2019 to 2024 to represent the majority of second-line initiations, with a 10-fold increase in initiations amongst older adults with frailty. Over the same period, population-level benefits for HbA1c and weight were modest, with no evidence of changes in risk of short-term discontinuation or diabetes complications. Findings support the careful use of SGLT2i across the wider population with type 2 diabetes including frail older adults.
Open AccessTrends in second-line glucose-lowering therapy initiations and treatment outcomes across age and frailty groups in people with type 2 diabetes: UK population-based study, 2019-2024
Objective To assess contemporary UK population-level trends in second-line glucose-lowering therapy initiation and explore treatment outcome trends across age and frailty groups in people with type 2 diabetes (T2D). Methods and Analysis We studied 117,064 adults with T2D who initiated major second-line glucose-lowering drug classes after metformin between 2019 to 2024 in population-representative UK data (Clinical Practice Research Datalink). Outcomes were assessed in three groups: age ≤70 years (n=84,589 [72%]), non-frail >70 years (n=18,933 [16%]), and frail >70 years (n=13,524 [12%]), with frailty specified as an electronic frailty index ≥ 0.24. Annual prescribing trends were calculated as the proportion of each drug class among all second-line initiations. We evaluated 12-month changes in HbA1c, weight, treatment discontinuation, and complications, including composite severe diabetes-related complications, heart failure and kidney failure. Trends were estimated using multivariable regression models adjusted for demographic and clinical characteristics. Results Use of SGLT2-inhibitors increased from 26% of overall second-line initiations in 2019 to 63% in 2024, with the greatest increase among frail adults >70 (6% in 2019 vs 60% in 2024). DPP4-inhibitor use declined substantially over the same period (44% 2019 vs 16% in 2024). In adults aged ≤70 years, mean 12-month HbA1c response improved from -11.3 mmol/mol (95%CI -11.6 to -11.0) in 2019 to -12.9 mmol/mol (95%CI -13.2 to -12.6) in 2023, and mean weight loss improved from -1.7 kg (95%CI -1.8 to -1.5) to -2.8 kg (95%CI -2.9 to -2.6). Treatment discontinuation within 12 months remained stable (30.2% [95%CI 29.4 to 31.0] in 2019 vs 30.7% [95%CI 30.0 to 31.4] in 2022). There was limited evidence of change in short-term complications over 2019-2022: rates (per 1,000 person-years) of severe diabetes-related complications were 9.9 (95%CI 8.1 to 11.7) vs 8.6 (95%CI 6.9 to 10.3), heart failure 2.0 (95%CI 1.1 to 2.9) vs 2.7 (95%CI 1.8 to 3.7), and kidney failure 2.1 (95%CI 1.1 to 3.0) vs 1.2 (95%CI 0.5 to 1.8). All trends were consistent in adults >70, including those with frailty. Conclusion Use SGLT2-inhibitors in people with T2D has rapidly increased in the UK over 2019 to 2024 to represent the majority of second-line initiations, with a 10-fold increase in initiations amongst older adults with frailty. Over the same period, population-level benefits for HbA1c and weight were modest, with no evidence of changes in risk of short-term discontinuation or diabetes complications. Findings support the careful use of SGLT2i across the wider population with type 2 diabetes including frail older adults.
- Open Access
Impact of depression on treatment progression in type 2 diabetes: A UK retrospective cohort study using the Clinical Practice Research Datalink Aurum database
Aims: To investigate the association between depression history, timing, and progression of diabetes treatment in type 2 diabetes (T2D). Methods: We conducted a cohort study using primary care records from the Clinical Practice Research Datalink Aurum database (2011–2023). Adults with T2D initiating oral glucose-lowering monotherapy (index) were included. Depression history, identified using clinical codes, was categorised by the most recent code before index: recent (≤1.7 years), intermediate (1.7–12.8 years), distant (>12.8 years), or none. Outcomes were time to treatment intensification (adding or switching drug class) and insulin initiation. We used Royston-Parmar survival models to estimate odds ratios (ORs) and 95% confidence intervals (CIs), adjusting for demographic and clinical variables. Results: Among 378,935 included individuals, 25.1% had a history of depression. Compared with no prior depression, recent depression was associated with higher odds of treatment intensification (OR 1.20, 95% CI: 1.17-1.23) and insulin initiation (OR 1.29, 95% CI: 1.23-1.36). Intermediate and distant depression were also associated with higher progression-odds, though more modestly. Conclusions: Depression, particularly recent episodes, is associated with earlier treatment progression in T2D, highlighting the importance of integrating mental and physical healthcare in diabetes management.
Open AccessImpact of depression on treatment progression in type 2 diabetes: A UK retrospective cohort study using the Clinical Practice Research Datalink Aurum database
Aims: To investigate the association between depression history, timing, and progression of diabetes treatment in type 2 diabetes (T2D). Methods: We conducted a cohort study using primary care records from the Clinical Practice Research Datalink Aurum database (2011–2023). Adults with T2D initiating oral glucose-lowering monotherapy (index) were included. Depression history, identified using clinical codes, was categorised by the most recent code before index: recent (≤1.7 years), intermediate (1.7–12.8 years), distant (>12.8 years), or none. Outcomes were time to treatment intensification (adding or switching drug class) and insulin initiation. We used Royston-Parmar survival models to estimate odds ratios (ORs) and 95% confidence intervals (CIs), adjusting for demographic and clinical variables. Results: Among 378,935 included individuals, 25.1% had a history of depression. Compared with no prior depression, recent depression was associated with higher odds of treatment intensification (OR 1.20, 95% CI: 1.17-1.23) and insulin initiation (OR 1.29, 95% CI: 1.23-1.36). Intermediate and distant depression were also associated with higher progression-odds, though more modestly. Conclusions: Depression, particularly recent episodes, is associated with earlier treatment progression in T2D, highlighting the importance of integrating mental and physical healthcare in diabetes management.
- Open Access
- 9 November 2025
- Published by Springer Nature in Diabetologia
- Vol. 69 (2) , 374-385
Precision medicine in type 2 diabetes: targeting SGLT2 inhibitor treatment for kidney protection
Aims/hypothesis: Current guidelines recommend use of sodium–glucose cotransporter-2 inhibitors (SGLT2 inhibitors) for kidney protection in people with type 2 diabetes and early-stage chronic kidney disease (CKD) based on a urinary albumin/creatinine ratio (uACR) of ≥3 mg/mmol. However, individuals with a normal uACR or low-level albuminuria were not represented in kidney outcome trials, leaving uncertainty about absolute treatment benefit in this group. To address this gap and support treatment decisions in clinical practice, we developed and validated a model to predict individual-level kidney protection benefit through the use of SGLT2 inhibitors. Methods: This observational cohort study used electronic health record data from UK primary care (Clinical Practice Research Datalink, 2013–2020) of adults with type 2 diabetes, eGFR ≥60 ml/min per 1.73 m 2 and uACR n =25,303, 17.9%) as the albuminuria threshold would prevent over 10% more events over 3 years (253 vs 228) by identifying a subgroup of 6.7% of individuals with uACR p =0.05). Conclusions/interpretation: A model adapting the international CKD-PC risk score can accurately predict the individual-level kidney protection benefit from treatment with SGLT2 inhibitors in people with type 2 diabetes and no or early-stage CKD. This could guide treatment decisions in clinical practice worldwide and could target treatment more effectively than the ≥3 mg/mmol albuminuria threshold recommended by current international guidelines. Graphical
- 9 November 2025
- Published by Springer Nature in Diabetologia
- Vol. 69 (2) , 374-385
Open AccessPrecision medicine in type 2 diabetes: targeting SGLT2 inhibitor treatment for kidney protection
Aims/hypothesis: Current guidelines recommend use of sodium–glucose cotransporter-2 inhibitors (SGLT2 inhibitors) for kidney protection in people with type 2 diabetes and early-stage chronic kidney disease (CKD) based on a urinary albumin/creatinine ratio (uACR) of ≥3 mg/mmol. However, individuals with a normal uACR or low-level albuminuria were not represented in kidney outcome trials, leaving uncertainty about absolute treatment benefit in this group. To address this gap and support treatment decisions in clinical practice, we developed and validated a model to predict individual-level kidney protection benefit through the use of SGLT2 inhibitors. Methods: This observational cohort study used electronic health record data from UK primary care (Clinical Practice Research Datalink, 2013–2020) of adults with type 2 diabetes, eGFR ≥60 ml/min per 1.73 m 2 and uACR n =25,303, 17.9%) as the albuminuria threshold would prevent over 10% more events over 3 years (253 vs 228) by identifying a subgroup of 6.7% of individuals with uACR p =0.05). Conclusions/interpretation: A model adapting the international CKD-PC risk score can accurately predict the individual-level kidney protection benefit from treatment with SGLT2 inhibitors in people with type 2 diabetes and no or early-stage CKD. This could guide treatment decisions in clinical practice worldwide and could target treatment more effectively than the ≥3 mg/mmol albuminuria threshold recommended by current international guidelines. Graphical