🟢 Strong Evidence
How researchers define and measure multimorbidity—the presence of multiple chronic conditions—dramatically affects both prevalence estimates and mortality predictions, according to a comprehensive analysis of nearly half a million UK adults. The study found that prevalence rates varied by more than 35-fold depending on the methodology used, raising critical questions about healthcare planning and resource allocation.
Key takeaways
- Multimorbidity prevalence ranged from 1.0% to 35.3% in the same population depending on definition method
- Count-based approaches using broader condition lists identified higher prevalence but clustering methods better predicted blood-based health markers
- Higher diagnostic thresholds identified more severe health profiles but did not improve mortality risk prediction
Study at a Glance
| Source | PLOS Medicine |
| Study type | Cohort study |
| Sample size | N = 474,397 |
| Population | UK Biobank participants aged 40-69 |
| Country | United Kingdom |
Multimorbidity Prevalence Varies Dramatically by Definition Method
Percentage of UK Biobank participants classified as having multimorbidity, 2006-2010
Source: Silva et al., PLOS Medicine 2024 | Georgian Medical Journal News
Count-Based Versus Clustering Approaches Show Different Strengths
The research team, led by Gabriella Silva and colleagues, compared six different count-based definitions of multimorbidity alongside clustering analysis using UK Biobank data collected between 2006 and 2010. Count-based methods simply tally the number of chronic conditions a person has, while clustering approaches identify patterns of diseases that commonly occur together.
Count-based definitions using comprehensive condition lists identified substantially higher prevalence rates. The extended list approach, requiring two or more conditions, classified 35.3% of participants as having multimorbidity, compared to just 1.0% using cluster-based methods.
Researchers found that higher diagnostic thresholds—requiring three rather than two conditions—identified participants with more severe health profiles across all measured outcomes, including polypharmacy, poor self-rated health, frailty, and chronic pain.
Blood Biomarkers Respond Better to Clustering Methods
While count-based approaches captured more participants, clustering methods showed stronger associations with blood-based health markers. The PLOS Medicine study found that clustering approaches better predicted levels of C-reactive protein, Cystatin-C, HDL and LDL cholesterol, and IGF-1.
“Associations with blood-based measures were more pronounced using clustering, with the highest differences from the standard definition distributed across clusters,” the authors reported. This suggests that disease patterns identified through clustering may reflect underlying biological pathways more accurately than simple condition counts.
The study analyzed data separately for men and women, using multivariable regression models adjusted for sociodemographic characteristics and body mass index, strengthening the reliability of the findings across different populations.
Multimorbidity prevalence ranged from 1.0% using cluster-based approaches to 35.3% using count-based extended condition lists in the same population
— Silva et al., PLOS Medicine (2024)
Mortality Predictions Show Unexpected Patterns
Surprisingly, higher diagnostic thresholds did not improve mortality risk prediction despite identifying participants with worse health profiles. Three-year mortality odds ratios ranged from 1.44 to 4.60 for men and 1.35 to similar ranges for women, depending on the definition used.
The findings have important implications for how clinical practice and public health policy approach multimorbidity. Different methodological choices could lead to vastly different estimates of disease burden and healthcare needs in aging populations.
These results also raise questions about optimal approaches for identifying high-risk patients in clinical settings, where both comprehensive screening and targeted interventions are necessary but resource-intensive.
Implications for Healthcare Planning and Research
The dramatic variation in prevalence estimates highlights a critical challenge for healthcare systems planning for aging populations. When the same dataset yields prevalence rates ranging from 1% to 35%, policymakers face significant uncertainty in resource allocation and service planning.
The research suggests that the choice of multimorbidity definition should align with the intended purpose. Count-based approaches may be more suitable for population health surveillance and resource planning, while clustering methods might better inform personalized medicine approaches and treatment strategies.
Future research comparing these methodologies across different populations and healthcare systems will be essential for developing standardized approaches to multimorbidity measurement and management strategies.
What this means
Frequently asked questions
Why do different methods for measuring multimorbidity matter?
Different measurement methods can lead to vastly different prevalence estimates—from 1% to 35% in the same population. This affects healthcare planning, resource allocation, and how patients are identified for specialized care programs.
Which approach is better for identifying high-risk patients?
The study suggests clustering approaches may be better for understanding biological processes and predicting blood-based health markers, while count-based methods may be more practical for routine clinical screening and population health surveillance.
How might this research change clinical practice?
Healthcare providers may need to consider both the number of conditions and patterns of co-occurring diseases when assessing patients. This could lead to more personalized treatment approaches based on disease clustering patterns rather than simple condition counts.
This research underscores the need for healthcare systems to carefully consider their methodological approaches to multimorbidity assessment. As populations continue aging globally, developing consistent and clinically meaningful definitions will be crucial for effective healthcare delivery and policy development. The UK Biobank findings provide a foundation for future work standardizing these critical measurement approaches across different healthcare contexts.
Was this article helpful?
Disclaimer. This article is health journalism intended for general information and education. It is not medical advice and is not a substitute for professional diagnosis or treatment. Always consult a qualified healthcare provider about your individual circumstances. Full disclaimer →
Related Coverage
Medically reviewed by Prof. Giorgi Pkhakadze, MD, MPH, PhD. Spotted an error? Contact the editorial team.
