Dietary habits in relation to outcome and therapy-related toxicity in patients with glioblastoma - a retrospective cohort study

Abstract

Purpose: Glioblastoma is the most prevalent primary malignant brain tumour in adults and bears poor survival. Data on associations of nutrition with outcome parameters are scarce. Our aim was to investigate whether dietary habits influence chemotherapy-related toxicity and overall survival in patients with glioblastoma.

Methods: In this retrospective cohort study, we included glioblastoma patients who received treatment between January 2010 and December 2019. We used a 35-item food frequency questionnaire to calculate a dietary score based on Mediterranean-like diet recommendations. We analysed relationships between nutrition score, side effects and survival using chi square tests, binary logistic regression, Kaplan-Meier and multivariable Cox regression analyses.

Results: From an initial population of 1,448 patients, a homogenous cohort of 128 glioblastoma cases were included in our final analysis. Patients with a higher score than the median, tending to a more Mediterranean-like diet, showed more infections (13.8% vs. 3.2%, p = 0.031) and a trend for more myelodepression (32.3% vs. 17.5%, p = 0.052). A higher score was predictive for infections (OR = 12.33, 95% CI: 1.36-111.98; p = 0.026). Median survival was worse in the higher score group (16.6 vs. 19.4 months, p = 0.004), confirmed by multivariable cox regression analysis (HR 1.60, 95% CI 1.04–2.46; p = 0.034).

Conclusion: We identified associations between dietary patterns and chemotherapy-related toxicity as well as outcome. Our findings underscore the potential impact of nutrition on cancer treatment outcomes. Further research is needed to validate these results and to address possible underlying mechanisms.

Supplementary Information: The online version contains supplementary material available at 10.1007/s11060-025-05137-3

Keywords: Diet; Food frequency questionnaire; Glioblastoma; Survival; Therapy-related toxicity.

Fig. 1

Consort flow chart of patients with prospective data collection

Fig. 2

Overall survival (OS) and Progression free survival (PFS) of patients with higher dietary scores compared to lower dietary scores. A Median OS of the higher score group was worse (16.6 months vs. 19.4 months, p = 0.004). This finding was confirmed by multivariable cox regression analysis (HR 1.60, 95% CI 1.04–2.46; p = 0.034). B PFS was not significantly different between both groups with 7.3 months vs. 8.0 months (HR 1.31, 95% CI 0.86–1.98; p = 0.205)