**Background:** The relationship between diet type and endurance running performance remains inconclusive, particularly when comparing omnivorous, vegetarian, and vegan diets. Previous studies have examined limited training variables, and no study had comprehensively assessed training behaviors and racing experience confounders across all three diet types in recreational long-distance runners. The NURMI (Nutrition and Running High Mileage) study aimed to fill this gap.
**Methods:** This cross-sectional, survey-based investigation (Part B of the NURMI Study) included 245 recreational distance runners (104 males, 141 females; median age 39 years, IQR 17) who had completed at least a half-marathon. Participants were recruited primarily from Germany and Austria (n=221; 90%) and classified by self-reported diet type: omnivore (n=109; 45%), vegetarian (n=45; 18%), or vegan (n=91; 37%). The minimum dietary adherence period was 4 weeks. Data were collected on running history (complete years of active running, age at first race), training behaviors (weekly runs, distances, durations, professional support, overall training duration), racing experiences (number of finished races, ratio of half-marathon/marathon to other races, planned races over prior two years), and best race times for half-marathon (HM), marathon (M), and ultra-marathon (UM). Statistical analysis used non-parametric tests (Wilcoxon, Chi-square) and multivariate linear regression with confounders, performed in R (version 4.2.2). Significance was set at p ≤ 0.05.
**Key Results:** Significant differences across diet groups were found for sex (p=0.004), body weight (p=0.001), and BMI (p=0.001). Omnivores were most likely male (53.2%) and had the highest BMI (22.6 kg/m², IQR 3.61). Vegans had the lowest BMI (21.3 kg/m², IQR 3.21) and were predominantly female (46% of females). No significant differences were found for age (p=0.054), height (p=0.088), country of residence (p=0.171), race motive (p=0.223), or preferred race distance (p=0.483). A significant difference was found for academic qualifications (p=0.029), with vegans most likely to hold a university degree or higher (36%). For running history, omnivores had the most complete years of active running (8 years, IQR 10) and vegans the least (5 years, IQR 8; p=0.002). No significant differences were found for best race times: HM (omnivore 106±21 min, vegetarian 111±25 min, vegan 118±46 min; p=0.153), M (omnivore 228±45 min, vegetarian 231±38 min, vegan 233±48 min; p=0.799), or UM (omnivore 705±498 min, vegetarian 829±806 min, vegan 454±270 min; p=0.081). Multivariate linear regression showed no significant diet-type differences in best HM/M race time when adjusting for running history (adjusted R²=0.16; vegetarian: b=−1.66, 95% CI [−11.6–8.27], p>0.05; vegan: b=0.491, 95% CI [−7.76–8.74], p>0.05), training behaviors (adjusted R²=0.22; vegetarian: b=−0.495, 95% CI [−10.5–9.51], p>0.05; vegan: b=1.85, 95% CI [−6.33–10], p>0.05), or racing experiences (adjusted R²=0.14; vegetarian: b=−0.716, 95% CI [−10.7–9.22], p>0.05; vegan: b=1.4, 95% CI [−6.78–9.59], p>0.05). Omnivores were significantly more well-being motivated than vegetarians (b=−12.6; 95% CI [−22.8–−2.47]; p<0.05), but no differences were found for competition or hobby motivations.
**Clinical Implications:** General diet type (omnivorous, vegetarian, or vegan) does not appear to independently influence half-marathon or marathon best race times in recreational distance runners when training behaviors and racing experience are accounted for. The similar performance across diet groups suggests that recreational runners can achieve comparable endurance performance regardless of dietary pattern, provided training is adequate. The higher proportion of females among vegans and the lower BMI in this group warrant further investigation into potential sex-specific performance advantages. Clinicians and sports nutrition professionals should reassure athletes that plant-based diets are compatible with recreational endurance performance, while emphasizing that individual training behaviors and experience are more critical determinants of race time than diet type alone. The study's limitations include its cross-sectional design, self-reported data, short minimum dietary adherence period (4 weeks), and predominantly Western European sample, which limit causal inference and generalizability.