The umbrella review pooled existing systematic reviews and meta-analyses to examine how composite dietary patterns relate to risk of endometrial cancer. Investigators reported that adherence to a Mediterranean diet was associated with lower odds of EC in pooled case‑control analyses (pooled OR = 0.71, 95% CI 0.51–0.98). That association was graded as class IV evidence under the prespecified Ioannidis/Fusar-Poli criteria. Short declarative statement: pooled case-control data suggested reduced odds. The source noted that a prospective cohort study addressing the same exposure reported hazard ratios rather than odds ratios and was therefore not pooled with the case-control data; the authors treated cohort evidence separately and narratively.
The review team also pooled data on empirically derived dietary patterns and found that a Western dietary pattern corresponded to higher odds of EC (pooled OR = 1.20, 95% CI 1.03–1.39), likewise graded class IV. The umbrella review cautioned that heterogeneity between primary studies existed for several pooled estimates and that some primary studies were excluded from pooled analyses because of insufficient reporting of effect measures or data needed for quantitative synthesis. These analytic choices were described in the methods and applied, for example, to the Mediterranean diet score (MDS) synthesis, where design‑ and effect‑measure stratification was performed.
Pooled evidence from systematic reviews assessing composite lifestyle indices — measures that integrate diet, body mass index, physical activity, and smoking status — indicated a stronger association with incident EC than for single dietary indices. The authors reported a combined healthy diet and lifestyle approach associated with a 46% lower hazard of EC incidence (pooled HR = 0.54, 95% CI 0.42–0.71). This finding was graded class III (suggestive) under the review’s evidence scheme. Heterogeneity for this pooled estimate was high, and the source explicitly reported that between-study variability limited the certainty of the association.
The umbrella review documented that prespecified indices evaluated included the Mediterranean Diet Score (MDS), Healthy Eating Index (HEI), Alternative Healthy Eating Index (AHEI), DASH, WCRF/AICR score, American Cancer Society (ACS) guidelines, and generic Healthy Lifestyle Index. The authors noted that some indices incorporate BMI and physical activity in addition to diet; therefore, pooled associations for lifestyle indices may reflect combined contributions of adiposity and behaviour as captured by those composite scores.
Randomised and non-randomised lifestyle interventions that combined dietary modification and physical activity were assessed for effects on quality of life (QoL) among EC survivors. Data indicated no significant differences in global or domain‑specific QoL across follow-ups ranging from nine weeks to 12 months, regardless of the validated instruments used, which included FACT‑G, FACT‑F, and SF‑12. The authors pooled two RCTs reporting SF‑12 physical health at six months and converted reported medians to means and SDs where necessary to allow synthesis; even with those adjustments, pooled QoL effects were non‑significant. The umbrella review therefore concluded that current intervention evidence does not demonstrate measurable QoL benefits in the short-to-medium term after diagnosis.
The review also reported very limited evidence on post-diagnosis outcomes beyond QoL. Data addressing recurrence, overall or cancer‑specific survival, mortality, and new‑onset cardiometabolic disease were sparse across systematic reviews and, where reported, were insufficient to support strong conclusions. The authors emphasized the gap in reliable post-diagnosis outcome data.
The authors systematically appraised methodological quality of included reviews using AMSTAR 2 and graded pooled evidence using a well‑defined Ioannidis/Fusar-Poli framework. They recomputed meta-analyses using random-effects models implemented in the metaumbrella R package, and they reported heterogeneity (I2), 95% prediction intervals, small-study effects (Egger’s test), and excess significance bias where appropriate. The umbrella review excluded individual primary studies from pooled analyses when mandatory data fields were missing, and they stratified analyses by study design and effect measure when comparability was limited.
Key limitations reported in the source include high heterogeneity in several pooled estimates, limited numbers of primary studies for some exposures and outcomes (preventing robust tests for small-study effects or excess significance bias), and incomplete reporting in some primary studies that precluded their inclusion in quantitative synthesis. The review also noted that most included systematic reviews focused on incidence rather than survivorship outcomes, leaving important clinical questions about recurrence, long-term survival, and cardiometabolic sequelae after EC incompletely addressed.
The umbrella review followed PRISMA guidance and registered a protocol in PROSPERO. Investigators searched PubMed, Embase, Scopus, Web of Science, CINAHL, and the Cochrane Library from inception until 3 December 2024. Search terms covered diet scores, dietary patterns, lifestyle indices, and cancer prevention guideline nomenclature; filters for systematic reviews and meta-analyses were applied. Two reviewers independently screened records, extracted data, and assessed methodological quality using AMSTAR 2. For meta-analytic recomputation, the team required specific primary-study data fields (effect sizes, CIs, counts) and applied REML to estimate heterogeneity. The evidence grading applied the Ioannidis classification: class I (convincing) through class IV (weak) and non-significant, with thresholds based on case numbers, p-values, heterogeneity, prediction intervals, and bias assessments.
Final reporting in the source emphasised that suggestive (class III–IV) evidence linked healthier dietary patterns — particularly the Mediterranean diet — and combined healthy lifestyle approaches with lower EC incidence. The review explicitly concluded that evidence for post-diagnosis outcomes is currently insufficient or non‑significant and called for further high‑quality research to address these gaps and reduce heterogeneity and reporting limitations across primary studies.