Associations between multimorbidity and adverse health outcomes in UK Biobank and the SAIL Databank: A comparison of longitudinal cohort studies

PheMIME: an interactive web app and knowledge base for phenome-wide, multi-institutional multimorbidity analysis

OBJECTIVES

To address the need for interactive visualization tools and databases in characterizing multimorbidity patterns across different populations, we developed the Phenome-wide Multi-Institutional Multimorbidity Explorer (PheMIME). This tool leverages three large-scale EHR systems to facilitate efficient analysis and visualization of disease multimorbidity, aiming to reveal both robust and novel disease associations that are consistent across different systems and to provide insight for enhancing personalized healthcare strategies.

MATERIALS AND METHODS

PheMIME integrates summary statistics from phenome-wide analyses of disease multimorbidities, utilizing data from Vanderbilt University Medical Center, Mass General Brigham, and the UK Biobank. It offers interactive and multifaceted visualizations for exploring multimorbidity. Incorporating an enhanced version of associationSubgraphs, PheMIME also enables dynamic analysis and inference of disease clusters, promoting the discovery of complex multimorbidity patterns. A case study on schizophrenia demonstrates its capability for generating interactive visualizations of multimorbidity networks within and across multiple systems. Additionally, PheMIME supports diverse multimorbidity-based discoveries, detailed further in online case studies.

RESULTS

The PheMIME is accessible at https://prod.tbilab.org/PheMIME/. A comprehensive tutorial and multiple case studies for demonstration are available at https://prod.tbilab.org/PheMIME_supplementary_materials/. The source code can be downloaded from https://github.com/tbilab/PheMIME.

DISCUSSION

PheMIME represents a significant advancement in medical informatics, offering an efficient solution for accessing, analyzing, and interpreting the complex and noisy real-world patient data in electronic health records.

CONCLUSION

PheMIME provides an extensive multimorbidity knowledge base that consolidates data from three EHR systems, and it is a novel interactive tool designed to analyze and visualize multimorbidities across multiple EHR datasets. It stands out as the first of its kind to offer extensive multimorbidity knowledge integration with substantial support for efficient online analysis and interactive visualization.

Association of latent class analysis-derived multimorbidity clusters with adverse health outcomes in patients with multiple long-term conditions: comparative results across three UK cohorts

Background

It remains unclear how to meaningfully classify people living with multimorbidity (multiple long-term conditions (MLTCs)), beyond counting the number of conditions. This paper aims to identify clusters of MLTCs in different age groups and associated risks of adverse health outcomes and service use.

Methods

Latent class analysis was used to identify MLTCs clusters in different age groups in three cohorts: Secure Anonymised Information Linkage Databank (SAIL) (n = 1,825,289), UK Biobank (n = 502,363), and the UK Household Longitudinal Study (UKHLS) (n = 49,186). Incidence rate ratios (IRR) for MLTC clusters were computed for: all-cause mortality, hospitalisations, and general practice (GP) use over 10 years, using <2 MLTCs as reference. Information on health outcomes and service use were extracted for a ten year follow up period (between 01st Jan 2010 and 31st Dec 2019 for UK Biobank and UKHLS, and between 01st Jan 2011 and 31st Dec 2020 for SAIL).

Findings

Clustering MLTCs produced largely similar results across different age groups and cohorts. MLTC clusters had distinct associations with health outcomes and service use after accounting for LTC counts, in fully adjusted models. The largest associations with mortality, hospitalisations and GP use in SAIL were observed for the "Pain+" cluster in the age-group 18-36 years (mortality IRR = 4.47, hospitalisation IRR = 1.84; GP use IRR = 2.87) and the "Hypertension, Diabetes & Heart disease" cluster in the age-group 37-54 years (mortality IRR = 4.52, hospitalisation IRR = 1.53, GP use IRR = 2.36). In UK Biobank, the "Cancer, Thyroid disease & Rheumatoid arthritis" cluster in the age group 37-54 years had the largest association with mortality (IRR = 2.47). Cardiometabolic clusters across all age groups, pain/mental health clusters in younger groups, and cancer and pulmonary related clusters in older age groups had higher risk for all outcomes. In UKHLS, MLTC clusters were not significantly associated with higher risk of adverse outcomes, except for the hospitalisation in the age-group 18-36 years.

Interpretation

Personalising care around MLTC clusters that have higher risk of adverse outcomes may have important implications for practice (in relation to secondary prevention), policy (with allocation of health care resources), and research (intervention development and targeting), for people living with MLTCs.

Funding

This study was funded by the National Institute for Health and Care Research (NIHR; Personalised Exercise-Rehabilitation FOR people with Multiple long-term conditions (multimorbidity)-NIHR202020).

Associations of Circulating Biomarkers with Disease Risks: A Two-Sample Mendelian Randomization Study

Circulating biomarkers play a pivotal role in personalized medicine, offering potential for disease screening, prevention, and treatment. Despite established associations between numerous biomarkers and diseases, elucidating their causal relationships is challenging. Mendelian Randomization (MR) can address this issue by employing genetic instruments to discern causal links. Additionally, using multiple MR methods with overlapping results enhances the reliability of discovered relationships. Here, we report an MR study using multiple methods, including inverse variance weighted, simple mode, weighted mode, weighted median, and MR-Egger. We use the MR-base resource (v0.5.6) from Hemani et al. 2018 to evaluate causal relationships between 212 circulating biomarkers (curated from UK Biobank analyses by Neale lab and from Shin et al. 2014, Roederer et al. 2015, and Kettunen et al. 2016 and 99 complex diseases (curated from several consortia by MRC IEU and Biobank Japan). We report novel causal relationships found by four or more MR methods between glucose and bipolar disorder (Mean Effect Size estimate across methods: 0.39) and between cystatin C and bipolar disorder (Mean Effect Size: -0.31). Based on agreement in four or more methods, we also identify previously known links between urate with gout and creatine with chronic kidney disease, as well as biomarkers that may be causal of cardiovascular conditions: apolipoprotein B, cholesterol, LDL, lipoprotein A, and triglycerides in coronary heart disease, as well as lipoprotein A, LDL, cholesterol, and apolipoprotein B in myocardial infarction. This Mendelian Randomization study not only corroborates known causal relationships between circulating biomarkers and diseases but also uncovers two novel biomarkers associated with bipolar disorder that warrant further investigation. Our findings provide insight into understanding how biological processes reflecting circulating biomarkers and their associated effects may contribute to disease etiology, which can eventually help improve precision diagnostics and intervention.

Associations of Circulating Biomarkers with Disease Risks: a Two-Sample Mendelian Randomization Study

Background

Circulating biomarkers play a pivotal role in personalized medicine, offering potential for disease screening, prevention, and treatment. Despite established associations between numerous biomarkers and diseases, elucidating their causal relationships is challenging. Mendelian Randomization (MR) can address this issue by employing genetic instruments to discern causal links. Additionally, using multiple MR methods with overlapping results enhances the reliability of discovered relationships.

Methods

Here we report an MR study using multiple methods, including inverse variance weighted, simple mode, weighted mode, weighted median, and MR Egger. We use the MR-base resource (v0.5.6)1 to evaluate causal relationships between 212 circulating biomarkers (curated from UK Biobank analyses by Neale lab and from Shin et al. 2014, Roederer et al. 2015, and Kettunen et al. 2016)2-4 and 99 complex diseases (curated from several consortia by MRC IEU and Biobank Japan).

Results

We report novel causal relationships found by 4 or more MR methods between glucose and bipolar disorder (Mean Effect Size estimate across methods: 0.39) and between cystatin C and bipolar disorder (Mean Effect Size: -0.31). Based on agreement in 4 or more methods, we also identify previously known links between urate with gout and creatine with chronic kidney disease, as well as biomarkers that may be causal of cardiovascular conditions: apolipoprotein B, cholesterol, LDL, lipoprotein A, and triglycerides in coronary heart disease, as well as lipoprotein A, LDL, cholesterol, and apolipoprotein B in myocardial infarction.

Conclusions

This Mendelian Randomization study not only corroborates known causal relationships between circulating biomarkers and diseases but also uncovers two novel biomarkers associated with bipolar disorder that warrant further investigation. Our findings provide insight into understanding how biological processes reflecting circulating biomarkers and their associated effects may contribute to disease etiology, which can eventually help improve precision diagnostics and intervention.

Participant characteristics and exclusion from phase 3/4 industry funded trials of chronic medical conditions: meta-analysis of individual participant level data

Objectives

To assess whether age, sex, comorbidity count, and race and ethnic group are associated with the likelihood of trial participants not being enrolled in a trial for any reason (ie, screen failure).

Design

Bayesian meta-analysis of individual participant level data.

Setting

Industry funded phase 3/4 trials of chronic medical conditions.

Participants

Participants were identified using individual participant level data to be in either the enrolled group or screen failure group. Data were available for 52 trials involving 72 178 screened individuals of whom 24 733 (34%) were excluded from the trial at the screening stage.

Main outcome measures

For each trial, logistic regression models were constructed to assess likelihood of screen failure in people who had been invited to screening, and were regressed on age (per 10 year increment), sex (male v female), comorbidity count (per one additional comorbidity), and race or ethnic group. Trial level analyses were combined in Bayesian hierarchical models with pooling across condition.

Results

In age and sex adjusted models across all trials, neither age nor sex was associated with increased odds of screen failure, although weak associations were detected after additionally adjusting for comorbidity (odds ratio of age, per 10 year increment was 1.02 (95% credibility interval 1.01 to 1.04) and male sex (0.95 (0.91 to 1.00)). Comorbidity count was weakly associated with screen failure, but in an unexpected direction (0.97 per additional comorbidity (0.94 to 1.00), adjusted for age and sex). People who self-reported as black seemed to be slightly more likely to fail screening than people reporting as white (1.04 (0.99 to 1.09)); a weak effect that seemed to persist after adjustment for age, sex, and comorbidity count (1.05 (0.98 to 1.12)). The between-trial heterogeneity was generally low, evidence of heterogeneity by sex was noted across conditions (variation in odds ratios on log scale of 0.01-0.13).

Conclusions

Although the conclusions are limited by uncertainty about the completeness or accuracy of data collection among participants who were not randomised, we identified mostly weak associations with an increased likelihood of screen failure for age, sex, comorbidity count, and black race or ethnic group. Proportionate increases in screening these underserved populations may improve representation in trials.

Trial registration number

PROSPERO CRD42018048202.

Ethnic differences in multimorbidity after accounting for social-economic factors, findings from The Health Survey for England

BACKGROUND

Social-economic factors and health behaviours may be driving variation in ethnic health inequalities in multimorbidity including among distinct ethnic groups.

METHODS

Using the cross-sectional nationally representative Health Surveys for England 2011-18 (N = 54 438, aged 16+), we performed multivariable logistic regression on the odds of having general multimorbidity (≥2 longstanding conditions) by ethnicity [British White (reference group), White Irish, Other White, Indian, Pakistani, Bangladeshi, Chinese, African, Caribbean, White mixed, Other Mixed], adjusting for age, sex, education, area deprivation, obesity, smoking status and survey year. This was repeated for cardiovascular multimorbidity (N = 37 148, aged 40+: having ≥2 of the following: self-reported diabetes, hypertension, heart attack or stroke) and multiple cardiometabolic risk biomarkers (HbA1c ≥6.5%, raised blood pressure, total cholesterol ≥5mmol/L).

RESULTS

Twenty percent of adults had general multimorbidity. In fully adjusted models, compared with the White British majority, Other White [odds ratio (OR) = 0.63; 95% confidence interval (CI) 0.53-0.74], Chinese (OR = 0.58, 95% CI 0.36-0.93) and African adults (OR = 0.54, 95% CI 0.42-0.69), had lower odds of general multimorbidity. Among adults aged 40+, Pakistani (OR = 1.27, 95% CI 0.97-1.66; P = 0.080) and Bangladeshi (OR = 1.75, 95% CI 1.16-2.65) had increased odds, and African adults had decreased odds (OR = 0.63, 95% CI 0.47-0.83) of general multimorbidity. Risk of cardiovascular multimorbidity was higher among Indian (OR = 3.31, 95% CI 2.56-4.28), Pakistani (OR = 3.48, 95% CI 2.52-4.80), Bangladeshi (OR = 3.67, 95% CI 1.98-6.78), African (OR = 1.61, 95% CI 1.05-2.47), Caribbean (OR = 2.18, 95% CI 1.59-2.99) and White mixed (OR = 1.98, 95% CI 1.14-3.44) adults. Indian adults were also at risk of having multiple cardiometabolic risk biomarkers.

CONCLUSION

Ethnic inequalities in multimorbidity are independent of social-economic factors. Ethnic minority groups are particularly at risk of cardiovascular multimorbidity, which may be exacerbated by poorer management of cardiometabolic risk requiring further investigation.

Phenotypes and rates of cancer-relevant symptoms and tests in the year before cancer diagnosis in UK Biobank and CPRD Gold

Early diagnosis of cancer relies on accurate assessment of cancer risk in patients presenting with symptoms, when screening is not appropriate. But recorded symptoms in cancer patients pre-diagnosis may vary between different sources of electronic health records (EHRs), either genuinely or due to differential completeness of symptom recording. To assess possible differences, we analysed primary care EHRs in the year pre-diagnosis of cancer in UK Biobank and Clinical Practice Research Datalink (CPRD) populations linked to cancer registry data. We developed harmonised phenotypes in Read v2 and CTV3 coding systems for 21 symptoms and eight blood tests relevant to cancer diagnosis. Among 22,601 CPRD and 11,594 UK Biobank cancer patients, 54% and 36%, respectively, had at least one consultation for possible cancer symptoms recorded in the year before their diagnosis. Adjusted comparisons between datasets were made using multivariable Poisson models, comparing rates of symptoms/tests in CPRD against expected rates if cancer site-age-sex-deprivation associations were the same as in UK Biobank. UK Biobank cancer patients compared with those in CPRD had lower rates of consultation for possible cancer symptoms [RR: 0.61 (0.59-0.63)], and lower rates for any primary care consultation [RR: 0.86 (95%CI 0.85-0.87)]. Differences were larger for 'non-alarm' symptoms [RR: 0.54 (0.52-0.56)], and smaller for 'alarm' symptoms [RR: 0.80 (0.76-0.84)] and blood tests [RR: 0.93 (0.90-0.95)]. In the CPRD cohort, approximately representative of the UK population, half of cancer patients had recorded symptoms in the year before diagnosis. The frequency of non-specific presenting symptoms recorded in the year pre-diagnosis of cancer was substantially lower among UK Biobank participants. The degree to which results based on highly selected biobank cohorts are generalisable needs to be examined in disease-specific contexts.

Personalised lung cancer risk stratification and lung cancer screening: do general practice electronic medical records have a role?

BACKGROUND

In the United Kingdom (UK), cancer screening invitations are based on general practice (GP) registrations. We hypothesize that GP electronic medical records (EMR) can be utilised to calculate a lung cancer risk score with good accuracy/clinical utility.

METHODS

The development cohort was Secure Anonymised Information Linkage-SAIL (2.3 million GP EMR) and the validation cohort was UK Biobank-UKB (N = 211,597 with GP-EMR availability). Fast backward method was applied for variable selection and area under the curve (AUC) evaluated discrimination.

RESULTS

Age 55-75 were included (SAIL: N = 574,196; UKB: N = 137,918). Six-year lung cancer incidence was 1.1% (6430) in SAIL and 0.48% (656) in UKB. The final model included 17/56 variables in SAIL for the EMR-derived score: age, sex, socioeconomic status, smoking status, family history, body mass index (BMI), BMI:smoking interaction, alcohol misuse, chronic obstructive pulmonary disease, coronary heart disease, dementia, hypertension, painful condition, stroke, peripheral vascular disease and history of previous cancer and previous pneumonia. The GP-EMR-derived score had AUC of 80.4% in SAIL and 74.4% in UKB and outperformed ever-smoked criteria (currently the first step in UK lung cancer screening pilots).

DISCUSSION

A GP-EMR-derived score may have a role in UK lung cancer screening by accurately targeting high-risk individuals without requiring patient contact.

Breast cancer risk stratification using genetic and non-genetic risk assessment tools for 246,142 women in the UK Biobank

PURPOSE

The benefit of using individual risk prediction tools to identify high-risk individuals for breast cancer (BC) screening is uncertain, despite the personalized approach of risk-based screening.

METHODS

We studied the overlap of predicted high-risk individuals among 246,142 women enrolled in the UK Biobank. Risk predictors assessed include the Gail model (Gail), BC family history (FH, binary), BC polygenic risk score (PRS), and presence of loss-of-function (LoF) variants in BC predisposition genes. Youden J-index was used to select optimal thresholds for defining high-risk.

RESULTS

In total, 147,399 were considered at high risk for developing BC within the next 2 years by at least 1 of the 4 risk prediction tools examined (Gail2-year > 0.5%: 47%, PRS2-yea r > 0.7%: 30%, FH: 6%, and LoF: 1%); 92,851 (38%) were flagged by only 1 risk predictor. The overlap between individuals flagged as high-risk because of genetic (PRS) and Gail model risk factors was 30%. The best-performing combinatorial model comprises a union of high-risk women identified by PRS, FH, and, LoF (AUC2-year [95% CI]: 62.2 [60.8 to 63.6]). Assigning individual weights to each risk prediction tool increased discriminatory ability.

CONCLUSION

Risk-based BC screening may require a multipronged approach that includes PRS, predisposition genes, FH, and other recognized risk factors.

Sex differences in associations between creatinine and cystatin C-based kidney function measures with stroke and major bleeding

PURPOSE

We sought to explore whether adding kidney function biomarkers based on creatinine (eGFRCr), cystatin C (eGFRCys) or a combination of the two (eGFRCr-Cys) could improve risk stratification for stroke and major bleeding, and whether there were sex differences in any additive value of kidney function biomarkers.

METHOD

We included participants from the UK Biobank who had not had a previous ischaemic or haemorrhagic stroke or major bleeding episode, and who had kidney function measures available at baseline. Cause-specific Cox proportional hazards models tested associations between eGFRCr, eGFRCys and eGFRCr-Cys (mL/min/1.73 m2) with ischaemic and haemorrhagic stroke, major bleeding (gastrointestinal or intracranial, including haemorrhagic stroke) and all-cause mortality.

FINDINGS

Among 452,879 eligible participants, 246,244 (54.4%) were women. Over 11.5 (IQR 10.8-12.2) years, there were 3706 ischaemic strokes, 795 haemorrhagic strokes, 26,025 major bleeding events and 28,851 deaths. eGFRCys was more strongly associated with ischaemic stroke than eGFRCr: an effect that was more pronounced in women (men - HR: 1.16, 95% CI: 1.12-1.19; female to male comparison - HR: 1.11, 95% CI: 1.05-1.16, per 10 mL/min/1.73 m2 decline in eGFRCys). This interaction effect was also demonstrated for eGFRCr-Cys, but not eGFRCr. eGFRCys and eGFRCr-Cys were more strongly associated with major bleeding and all-cause mortality than eGFRCr in both men and women. Event numbers were small for haemorrhagic stroke.

DISCUSSION

To a greater degree than is seen in men, eGFRCr underestimates risk of ischaemic stroke and major bleeding in women compared to eGFRCys. The difference between measures is likely explained by non-GFR biology of creatinine and cystatin C.

CONCLUSION

Enhanced measurement of cystatin C may improve risk stratification for ischaemic stroke and major bleeding and clinical treatment decisions in a general population setting, particularly for women.

Impact of data source choice on multimorbidity measurement: a comparison study of 2.3 million individuals in the Welsh National Health Service

BACKGROUND

Measurement of multimorbidity in research is variable, including the choice of the data source used to ascertain conditions. We compared the estimated prevalence of multimorbidity and associations with mortality using different data sources.

METHODS

A cross-sectional study of SAIL Databank data including 2,340,027 individuals of all ages living in Wales on 01 January 2019. Comparison of prevalence of multimorbidity and constituent 47 conditions using data from primary care (PC), hospital inpatient (HI), and linked PC-HI data sources and examination of associations between condition count and 12-month mortality.

RESULTS

Using linked PC-HI compared with only HI data, multimorbidity was more prevalent (32.2% versus 16.5%), and the population of people identified as having multimorbidity was younger (mean age 62.5 versus 66.8 years) and included more women (54.2% versus 52.6%). Individuals with multimorbidity in both PC and HI data had stronger associations with mortality than those with multimorbidity only in HI data (adjusted odds ratio 8.34 [95% CI 8.02-8.68] versus 6.95 (95%CI 6.79-7.12] in people with ≥ 4 conditions). The prevalence of conditions identified using only PC versus only HI data was significantly higher for 37/47 and significantly lower for 10/47: the highest PC/HI ratio was for depression (14.2 [95% CI 14.1-14.4]) and the lowest for aneurysm (0.51 [95% CI 0.5-0.5]). Agreement in ascertainment of conditions between the two data sources varied considerably, being slight for five (kappa < 0.20), fair for 12 (kappa 0.21-0.40), moderate for 16 (kappa 0.41-0.60), and substantial for 12 (kappa 0.61-0.80) conditions, and by body system was lowest for mental and behavioural disorders. The percentage agreement, individuals with a condition identified in both PC and HI data, was lowest in anxiety (4.6%) and highest in coronary artery disease (62.9%).

CONCLUSIONS

The use of single data sources may underestimate prevalence when measuring multimorbidity and many important conditions (especially mental and behavioural disorders). Caution should be used when interpreting findings of research examining individual and multiple long-term conditions using single data sources. Where available, researchers using electronic health data should link primary care and hospital inpatient data to generate more robust evidence to support evidence-based healthcare planning decisions for people with multimorbidity.

PheMIME: An Interactive Web App and Knowledge Base for Phenome-Wide, Multi-Institutional Multimorbidity Analysis

Motivation

Multimorbidity, characterized by the simultaneous occurrence of multiple diseases in an individual, is an increasing global health concern, posing substantial challenges to healthcare systems. Comprehensive understanding of disease-disease interactions and intrinsic mechanisms behind multimorbidity can offer opportunities for innovative prevention strategies, targeted interventions, and personalized treatments. Yet, there exist limited tools and datasets that characterize multimorbidity patterns across different populations. To bridge this gap, we used large-scale electronic health record (EHR) systems to develop the Phenome-wide Multi-Institutional Multimorbidity Explorer (PheMIME), which facilitates research in exploring and comparing multimorbidity patterns among multiple institutions, potentially leading to the discovery of novel and robust disease associations and patterns that are interoperable across different systems and organizations.

Results

PheMIME integrates summary statistics from phenome-wide analyses of disease multimorbidities. These are currently derived from three major institutions: Vanderbilt University Medical Center, Mass General Brigham, and the UK Biobank. PheMIME offers interactive exploration of multimorbidity through multi-faceted visualization. Incorporating an enhanced version of associationSubgraphs, PheMIME enables dynamic analysis and inference of disease clusters, promoting the discovery of multimorbidity patterns. Once a disease of interest is selected, the tool generates interactive visualizations and tables that users can delve into multimorbidities or multimorbidity networks within a single system or compare across multiple systems. The utility of PheMIME is demonstrated through a case study on schizophrenia.

Availability and implementation

The PheMIME knowledge base and web application are accessible at https://prod.tbilab.org/PheMIME/. A comprehensive tutorial, including a use-case example, is available at https://prod.tbilab.org/PheMIME_supplementary_materials/. Furthermore, the source code for PheMIME can be freely downloaded from https://github.com/tbilab/PheMIME.

Data availability statement

The data underlying this article are available in the article and in its online web application or supplementary material.

A prognostic risk score for development and spread of chronic pain

Chronic pain is a complex condition influenced by a combination of biological, psychological and social factors. Using data from the UK Biobank (n = 493,211), we showed that pain spreads from proximal to distal sites and developed a biopsychosocial model that predicted the number of coexisting pain sites. This data-driven model was used to identify a risk score that classified various chronic pain conditions (area under the curve (AUC) 0.70-0.88) and pain-related medical conditions (AUC 0.67-0.86). In longitudinal analyses, the risk score predicted the development of widespread chronic pain, the spreading of chronic pain across body sites and high-impact pain about 9 years later (AUC 0.68-0.78). Key risk factors included sleeplessness, feeling 'fed-up', tiredness, stressful life events and a body mass index >30. A simplified version of this score, named the risk of pain spreading, obtained similar predictive performance based on six simple questions with binarized answers. The risk of pain spreading was then validated in the Northern Finland Birth Cohort (n = 5,525) and the PREVENT-AD cohort (n = 178), obtaining comparable predictive performance. Our findings show that chronic pain conditions can be predicted from a common set of biopsychosocial factors, which can aid in tailoring research protocols, optimizing patient randomization in clinical trials and improving pain management.

Treatment effect modification due to comorbidity: Individual participant data meta-analyses of 120 randomised controlled trials

BACKGROUND

People with comorbidities are underrepresented in clinical trials. Empirical estimates of treatment effect modification by comorbidity are lacking, leading to uncertainty in treatment recommendations. We aimed to produce estimates of treatment effect modification by comorbidity using individual participant data (IPD).

METHODS AND FINDINGS

We obtained IPD for 120 industry-sponsored phase 3/4 trials across 22 index conditions (n = 128,331). Trials had to be registered between 1990 and 2017 and have recruited ≥300 people. Included trials were multicentre and international. For each index condition, we analysed the outcome most frequently reported in the included trials. We performed a two-stage IPD meta-analysis to estimate modification of treatment effect by comorbidity. First, for each trial, we modelled the interaction between comorbidity and treatment arm adjusted for age and sex. Second, for each treatment within each index condition, we meta-analysed the comorbidity-treatment interaction terms from each trial. We estimated the effect of comorbidity measured in 3 ways: (i) the number of comorbidities (in addition to the index condition); (ii) presence or absence of the 6 commonest comorbid diseases for each index condition; and (iii) using continuous markers of underlying conditions (e.g., estimated glomerular filtration rate (eGFR)). Treatment effects were modelled on the usual scale for the type of outcome (absolute scale for numerical outcomes, relative scale for binary outcomes). Mean age in the trials ranged from 37.1 (allergic rhinitis trials) to 73.0 (dementia trials) and percentage of male participants range from 4.4% (osteoporosis trials) to 100% (benign prostatic hypertrophy trials). The percentage of participants with 3 or more comorbidities ranged from 2.3% (allergic rhinitis trials) to 57% (systemic lupus erythematosus trials). We found no evidence of modification of treatment efficacy by comorbidity, for any of the 3 measures of comorbidity. This was the case for 20 conditions for which the outcome variable was continuous (e.g., change in glycosylated haemoglobin in diabetes) and for 3 conditions in which the outcomes were discrete events (e.g., number of headaches in migraine). Although all were null, estimates of treatment effect modification were more precise in some cases (e.g., sodium-glucose co-transporter-2 (SGLT2) inhibitors for type 2 diabetes-interaction term for comorbidity count 0.004, 95% CI -0.01 to 0.02) while for others credible intervals were wide (e.g., corticosteroids for asthma-interaction term -0.22, 95% CI -1.07 to 0.54). The main limitation is that these trials were not designed or powered to assess variation in treatment effect by comorbidity, and relatively few trial participants had >3 comorbidities.

CONCLUSIONS

Assessments of treatment effect modification rarely consider comorbidity. Our findings demonstrate that for trials included in this analysis, there was no empirical evidence of treatment effect modification by comorbidity. The standard assumption used in evidence syntheses is that efficacy is constant across subgroups, although this is often criticised. Our findings suggest that for modest levels of comorbidities, this assumption is reasonable. Thus, trial efficacy findings can be combined with data on natural history and competing risks to assess the likely overall benefit of treatments in the context of comorbidity.

Socioeconomic gradient in mortality of working age and older adults with multiple long-term conditions in England and Ontario, Canada

BACKGROUND

There is currently mixed evidence on the influence of long-term conditions and deprivation on mortality. We aimed to explore whether number of long-term conditions contribute to socioeconomic inequalities in mortality, whether the influence of number of conditions on mortality is consistent across socioeconomic groups and whether these associations vary by working age (18-64 years) and older adults (65 + years). We provide a cross-jurisdiction comparison between England and Ontario, by replicating the analysis using comparable representative datasets.

METHODS

Participants were randomly selected from Clinical Practice Research Datalink in England and health administrative data in Ontario. They were followed from 1 January 2015 to 31 December 2019 or death or deregistration. Number of conditions was counted at baseline. Deprivation was measured according to the participant's area of residence. Cox regression models were used to estimate hazards of mortality by number of conditions, deprivation and their interaction, with adjustment for age and sex and stratified between working age and older adults in England (N = 599,487) and Ontario (N = 594,546).

FINDINGS

There is a deprivation gradient in mortality between those living in the most deprived areas compared to the least deprived areas in England and Ontario. Number of conditions at baseline was associated with increasing mortality. The association was stronger in working age compared with older adults respectively in England (HR = 1.60, 95% CI 1.56,1.64 and HR = 1.26, 95% CI 1.25,1.27) and Ontario (HR = 1.69, 95% CI 1.66,1.72 and HR = 1.39, 95% CI 1.38,1.40). Number of conditions moderated the socioeconomic gradient in mortality: a shallower gradient was seen for persons with more long-term conditions.

CONCLUSIONS

Number of conditions contributes to higher mortality rate and socioeconomic inequalities in mortality in England and Ontario. Current health care systems are fragmented and do not compensate for socioeconomic disadvantages, contributing to poor outcomes particularly for those managing multiple long-term conditions. Further work should identify how health systems can better support patients and clinicians who are working to prevent the development and improve the management of multiple long-term conditions, especially for individuals living in socioeconomically deprived areas.

Association of frailty with the incidence risk of cardiovascular disease and type 2 diabetes mellitus in long-term cancer survivors: a prospective cohort study

BACKGROUND

Comorbidities among cancer survivors remain a serious healthcare burden and require appropriate management. Using two widely used frailty indicators, this study aimed to evaluate whether frailty was associated with the incidence risk of cardiovascular disease (CVD) and type 2 diabetes mellitus (T2DM) among long-term cancer survivors.

METHODS

We included 13,388 long-term cancer survivors (diagnosed with cancer over 5 years before enrolment) free of CVD and 6101 long-term cancer survivors free of T2DM, at the time of recruitment (aged 40-69 years), from the UK Biobank. Frailty was assessed by the frailty phenotype (FP_Frailty, range: 0-5) and the frailty index (FI_Frailty, range: 0-1) at baseline. The incident CVD and T2DM were ascertained through linked hospital data and primary care data, respectively. The associations were examined using Cox proportional hazards regression models.

RESULTS

Compared with non-frail participants, those with pre-frailty (FP_Frailty [met 1-2 of the components]: hazard ratio [HR]=1.18, 95% confidence interval [CI]: 1.05, 1.32; FI_Frailty [0.10< FI ≤0.21]: HR=1.51, 95% CI: 1.32, 1.74) and frailty (FP_Frailty [met ≥3 of the components]: HR=2.12, 95% CI: 1.73, 2.60; FI_Frailty [FI >0.21]: HR=2.19, 95% CI: 1.85, 2.59) had a significantly higher risk of CVD in the multivariable-adjusted model. A similar association of FI_Frailty with the risk of incident T2DM was observed. We failed to find such an association for FP_Frailty. Notably, the very early stage of frailty (1 for FP_Frailty and 0.1-0.2 for FI_Frailty) was also positively associated with the risk of CVD and T2DM (FI_Frailty only). A series of sensitivity analyses confirmed the robustness of the findings.

CONCLUSIONS

Frailty, even in the very early stage, was positively associated with the incidence risk of CVD and T2DM among long-term cancer survivors, although discrepancies existed between frailty indicators. While the validation of these findings is required, they suggest that routine monitoring, prevention, and interventive programs of frailty among cancer survivors may help to prevent late comorbidities and, eventually, improve their quality of life. Especially, interventions are recommended to target those at an early stage of frailty when healthcare resources are limited.

Multidisciplinary ecosystem to study lifecourse determinants and prevention of early-onset burdensome multimorbidity (MELD-B) - protocol for a research collaboration

Background

Most people living with multiple long-term condition multimorbidity (MLTC-M) are under 65 (defined as 'early onset'). Earlier and greater accrual of long-term conditions (LTCs) may be influenced by the timing and nature of exposure to key risk factors, wider determinants or other LTCs at different life stages. We have established a research collaboration titled 'MELD-B' to understand how wider determinants, sentinel conditions (the first LTC in the lifecourse) and LTC accrual sequence affect risk of early-onset, burdensome MLTC-M, and to inform prevention interventions.

Aim

Our aim is to identify critical periods in the lifecourse for prevention of early-onset, burdensome MLTC-M, identified through the analysis of birth cohorts and electronic health records, including artificial intelligence (AI)-enhanced analyses.

Design

We will develop deeper understanding of 'burdensomeness' and 'complexity' through a qualitative evidence synthesis and a consensus study. Using safe data environments for analyses across large, representative routine healthcare datasets and birth cohorts, we will apply AI methods to identify early-onset, burdensome MLTC-M clusters and sentinel conditions, develop semi-supervised learning to match individuals across datasets, identify determinants of burdensome clusters, and model trajectories of LTC and burden accrual. We will characterise early-life (under 18 years) risk factors for early-onset, burdensome MLTC-M and sentinel conditions. Finally, using AI and causal inference modelling, we will model potential 'preventable moments', defined as time periods in the life course where there is an opportunity for intervention on risk factors and early determinants to prevent the development of MLTC-M. Patient and public involvement is integrated throughout.

Assessing trial representativeness using serious adverse events: an observational analysis using aggregate and individual-level data from clinical trials and routine healthcare data

BACKGROUND

The applicability of randomised controlled trials of pharmacological agents to older people with frailty/multimorbidity is often uncertain, due to concerns that trials are not representative. However, assessing trial representativeness is challenging and complex. We explore an approach assessing trial representativeness by comparing rates of trial serious adverse events (SAE) to rates of hospitalisation/death in routine care.

METHODS

This was an observational analysis of individual (125 trials, n=122,069) and aggregate-level drug trial data (483 trials, n=636,267) for 21 index conditions compared to population-based routine healthcare data (routine care). Trials were identified from ClinicalTrials.gov . Routine care comparison from linked primary care and hospital data from Wales, UK (n=2.3M). Our outcome of interest was SAEs (routinely reported in trials). In routine care, SAEs were based on hospitalisations and deaths (which are SAEs by definition). We compared trial SAEs in trials to expected SAEs based on age/sex standardised routine care populations with the same index condition. Using IPD, we assessed the relationship between multimorbidity count and SAEs in both trials and routine care and assessed the impact on the observed/expected SAE ratio additionally accounting for multimorbidity.

RESULTS

For 12/21 index conditions, the pooled observed/expected SAE ratio was <1, indicating fewer SAEs in trial participants than in routine care. A further 6/21 had point estimates <1 but the 95% CI included the null. The median pooled estimate of observed/expected SAE ratio was 0.60 (95% CI 0.55-0.64; COPD) and the interquartile range was 0.44 (0.34-0.55; Parkinson's disease) to 0.87 (0.58-1.29; inflammatory bowel disease). Higher multimorbidity count was associated with SAEs across all index conditions in both routine care and trials. For most trials, the observed/expected SAE ratio moved closer to 1 after additionally accounting for multimorbidity count, but it nonetheless remained below 1 for most.

CONCLUSIONS

Trial participants experience fewer SAEs than expected based on age/sex/condition hospitalisation and death rates in routine care, confirming the predicted lack of representativeness. This difference is only partially explained by differences in multimorbidity. Assessing observed/expected SAE may help assess the applicability of trial findings to older populations in whom multimorbidity and frailty are common.

Frailty in COPD: an analysis of prevalence and clinical impact using UK Biobank

BACKGROUND

Frailty, a state of reduced physiological reserve, is common in people with chronic obstructive pulmonary disease (COPD). Frailty can occur at any age; however, the implications in younger people (eg, aged <65 years) with COPD are unclear. We assessed the prevalence of frailty in UK Biobank participants with COPD; explored relationships between frailty and forced expiratory volume in 1 second (FEV1) and quantified the association between frailty and adverse outcomes.

METHODS

UK Biobank participants (n=3132, recruited 2006-2010) with COPD aged 40-70 years were analysed comparing two frailty measures (frailty phenotype and frailty index) at baseline. Relationship with FEV1 was assessed for each measure. Outcomes were mortality, major adverse cardiovascular event (MACE), all-cause hospitalisation, hospitalisation with COPD exacerbation and community COPD exacerbation over 8 years of follow-up.

RESULTS

Frailty was common by both definitions (17% frail using frailty phenotype, 28% moderate and 4% severely frail using frailty index). The frailty phenotype, but not the frailty index, was associated with lower FEV1. Frailty phenotype (frail vs robust) was associated with mortality (HR 2.33; 95% CI 1.84 to 2.96), MACE (2.73; 1.66 to 4.49), hospitalisation (incidence rate ratio 3.39; 2.77 to 4.14) hospitalised exacerbation (5.19; 3.80 to 7.09) and community exacerbation (2.15; 1.81 to 2.54), as was frailty index (severe vs robust) (mortality (2.65; 95% CI 1.75 to 4.02), MACE (6.76; 2.68 to 17.04), hospitalisation (3.69; 2.52 to 5.42), hospitalised exacerbation (4.26; 2.37 to 7.68) and community exacerbation (2.39; 1.74 to 3.28)). These relationships were similar before and after adjustment for FEV1.

CONCLUSION

Frailty, regardless of age or measure, identifies people with COPD at risk of adverse clinical outcomes. Frailty assessment may aid risk stratification and guide-targeted intervention in COPD and should not be limited to people aged >65 years.

Psychometric Properties of the Short Form-36 (SF-36) in Parents of Children with Mental Illness

Given the stressful experiences of parenting children with mental illness, researchers and health professionals must ensure that the health-related quality of life of these vulnerable parents is measured with sufficient validity and reliability. This study examined the psychometric properties of the SF-36 in parents of children with mental illness. The data come from 99 parents whose children were currently receiving mental health services. The correlated two-factor structure of the SF-36 was replicated. Internal consistencies were robust (α > 0.80) for all but three subscales (General Health, Vitality, Mental Health). Inter-subscale and component correlations were strong. Correlations with parental psychopathology ranged from r = −0.32 to −0.60 for the physical component and r = −0.39 to −0.75 for the mental component. Parents with clinically relevant psychopathology had significantly worse SF-36 scores. SF-36 scores were inversely associated with the number of child diagnoses. The SF-36 showed evidence of validity and reliability as a measure of health-related quality of life in parents of children with mental illness and may be used as a potential outcome in the evaluation of family-centered approaches to care within child psychiatry. Given the relatively small sample size of this study, research should continue to examine its psychometric properties in more diverse samples of caregivers.

Frailty in rheumatoidrmdopen-2021-002111 arthritis and its relationship with disease activity, hospitalisation and mortality: a longitudinal analysis of the Scottish Early Rheumatoid Arthritis cohort and UK Biobank

OBJECTIVE

To assess the prevalence of frailty in rheumatoid arthritis (RA) and its association with baseline and longitudinal disease activity, all-cause mortality and hospitalisation.

PARTICIPANTS

People with RA identified from the Scottish Early Rheumatoid Arthritis (SERA) inception cohort (newly diagnosed, mean age 58.2 years) and UK Biobank (established disease identified using diagnostic codes, mean age 59 years). Frailty was quantified using the frailty index (both datasets) and frailty phenotype (UK Biobank only). Disease activity was assessed using Disease Activity Score in 28 joints (DAS28) in SERA. Associations between baseline frailty and all-cause mortality and hospitalisation was estimated after adjusting for age, sex, socioeconomic status, smoking and alcohol, plus DAS28 in SERA.

RESULTS

Based on the frailty index, frailty was common in SERA (12% moderate, 0.2% severe) and UK Biobank (20% moderate, 3% severe). In UK Biobank, 23% were frail using frailty phenotype. Frailty index was associated with DAS28 in SERA, as well as age and female sex in both cohorts. In SERA, as DAS28 lessened over time with treatment, mean frailty index also decreased. The frailty index was associated with all-cause mortality (HR moderate/severe frailty vs robust 4.14 (95% CI 1.49 to 11.51) SERA, 1.68 (95% CI 1.26 to 2.13) UK Biobank) and unscheduled hospitalisation (incidence rate ratio 2.27 (95% CI 1.45 to 3.57) SERA 2.74 (95% CI 2.29 to 3.29) UK Biobank). In UK Biobank, frailty phenotype also associated with mortality and hospitalisation.

CONCLUSION

Frailty is common in early and established RA and associated with hospitalisation and mortality. Frailty in RA is dynamic and, for some, may be ameliorated through controlling disease activity in early disease.