Burden of disease review assesses the burden attributable to risk factors across Europe

By Pooja Toshniwal PahariaJun 27 2023Reviewed by Danielle Ellis, B.Sc.

In a recent study published in the Archives of Public Health, researchers performed a systematic review to map burden of disease (BoD) assessments, including risk factors performed in Europe, and identify the methodological choices in comparative risk assessment (CRA) and other risk assessment methods.

Background

Within the BoD framework, injury and disease burden estimates attributable to risk factors can guide policymaking and help identify priority individuals for disease prevention programs. Previous BoD assessment mapping studies performed in Europe showed non-harmony in methodological designs and modeling parameters for BoD and injury assessments.

The World Health Organization (WHO) and International Labour Organization (ILO) collaborators performed systematic reviews to assess the work-associated burdens of disease and injuries. They highlighted the importance of bias risks, evidence quality, and evidence strength in disease burden studies. However, the reviews did not map the specific methodological designs of disease burden studies evaluating disease burdens related to risk factors.

About the study

In the present systematic review, researchers presented studies that assessed the BoD related to risk factors performed in Europe and assessed their methodologies using the Comparative Risk Assessment (CRA) framework.

Data were searched without language restrictions on the Cochrane Library, Embase, Medline, Web of Science, and Google Scholar databases, including gray literature from the CABDirect, OAIster, WHO, and OpenGrey websites and targeted population health agencies’ data. In addition, the European BoD Network (burden-eu) platform members contributed to the final publication list with additional data accessible in their nations.

Only formal reports and burden of disease studies published from January 1990 to April 2020 assessing the risk factor-related disease burdens based on the Years of Life Lost (YLL), Years Lived with Disability (YLD), or Disability-Adjusted Life Years (DALY) values for the Global Burden of Disease (GBD) in 45 European nations across Eastern, Central, and Western Europe were included in the review. Multinational studies were included if they provided information for the concerned European nations.

A risk factor was defined as an individual’s behavioral preference or occupational, metabolic, or environmental factor affecting the risk related to the disease outcome, including diseases considered risk factors, e.g., type 2 diabetes mellitus (T2D), that can increase the risk of cancer. The burden of disease studies not assessing risk factors’ influences and only focussing on injuries or diseases, and those including indicators or metrics apart from DALY, YLD, or YLL, were excluded.

In addition, theses, books, editorials, conference presentations, and editorial letters were excluded. Two researchers screened the data independently, and a third researcher resolved the disagreements. The included studies were categorized based on the type of risk factor analyzed and the study type (independent vs. GBD-associated study).

Results

The data search yielded 8,167 records, of which only 559 (including five systematic reviews) underwent full-text screening, and of the 74 records that satisfied the eligibility criteria, 68 were considered, and six records from the reference lists of the systematic reviews reviewed were analyzed. In addition, 39 studies were included following burden-eu member consultation. In total, 113 studies were analyzed, of which 54 were independent BoD studies and 59 were linked to the GBD.

Among the records, 107 utilized CRA for estimating the attributable burden; however, CRA approaches varied considerably across the studies, with no observable patterns based on risk factors, country, or year, and diverse terminologies were used. Differences in methodological choices could impact the comparability of estimates due to significant influences on burden quantification.

Bottom-up approaches were preferred to CRA approaches for occupational and environmental risk factors and health outcomes in the case of unavailable disease envelopes. The findings showed a propensity to evaluate behavioral risk factors more than the others in Europe. Alcohol intake, tobacco consumption, and air pollution were the most investigated risk factors, and surveys (38% of independent studies) were the most common data sources, followed by literature and registries.

However, other risk factors were not studied widely, including poor dietary quality, cardiovascular comorbidities, and a lack of physical exercise. Among the independent burden of disease studies, the top-down method, including Population attributable Fraction (PAF) use (87%), with fixed counterfactual values (81%), based on relative risks (RR), was most commonly used to estimate BoD attributable to risk factors. For RR, odds ratio (OR), and hazards ratio (HR) determinations, data were mainly sourced from single studies (43%), literature reviews, and meta-analyses (27%), deriving the estimates from within the study context.

Among independent studies, only 13 assessed causalities, and half performed uncertainty analysis, reporting parametric uncertainty using Monte Carlo simulations and bootstrapping. There were many differences related to the exposure-response functions concerning sources and definitions. Non-CRA studies used analytical methods such as logistic regression and stratification, Markov modeling, or risk assessments with burden estimates derived by integrating dose-response relationships with exposure to the risk factors.

Conclusion

Overall, the findings, examining independent studies that assessed the BoD attributable to risk factors in Europe since the 1990s, showed misreporting, limited causality, and uncertainty considerations, and non-transparency in reporting techniques in the BoD studies.

Considerable variations across nations and risk factors were observed, with profound differences between CRA and other risk assessment methods hampering the comparability of results. The development and use of recommendations to perform and report BoD studies could aid in understanding the differences and avoiding misinterpretations, thereby improving the comparability of findings among studies.