Fruit, Vegetables and NMIBC Recurrence Risk

Key Takeaways

• In a Dutch prospective NMIBC cohort of more than 1,000 patients, higher fruit and vegetable intake 3 months after diagnosis was associated with lower first recurrence risk.
• Increases in total fruit and vegetable intake from before diagnosis to 3 months after diagnosis were also associated with lower first recurrence risk.
• Prediagnosis intake and both dietary and plasma carotenoids were not associated with recurrence or progression outcomes.
• No clear overall progression association was seen, although a higher fruit intake signal at 15 months tracked with higher progression risk and the authors cautioned that this could reflect chance.

Non-muscle-invasive bladder cancer (NMIBC) often leads to prolonged cystoscopic surveillance because recurrence after local treatment is common. In Dutch patients after transurethral resection of a bladder tumor (TURBT), whether postdiagnosis fruit and vegetable intake or circulating carotenoids relate to later recurrence or progression remains uncertain. For U.S. clinicians, these Dutch data add to limited prior observational survivorship evidence rather than establish U.S.-specific dietary guidance.

The UroLife cohort study was a prospective multicenter study across 22 hospitals in the Netherlands that enrolled Dutch-speaking adults aged 18 to 80 years with histologically confirmed Ta, T1, or Tis NMIBC after TURBT, with only stage T1 included from May 2017 onward. Prediagnosis diet analyses included 1396 patients, postdiagnosis diet analyses 1270, and plasma carotenoid analyses 910, with median follow-up of 4.7 years.

Prediagnosis diet was reported 6 weeks after diagnosis for the prior year, postdiagnosis diet at 3 and 15 months for the prior month, and blood was collected at 3 and 15 months. Fruit and vegetable intake came from a validated 163-item food frequency questionnaire, and plasma α-carotene, β-carotene, β-cryptoxanthin, lycopene, lutein, and zeaxanthin were measured by high-performance liquid chromatography (HPLC). First recurrence was the primary outcome, with multiple recurrences and progression as secondary outcomes, and models adjusted for age, sex, smoking, body mass index, physical activity, energy intake, and European Association of Urology (EAU) risk score. Multiple imputation addressed missing data, and complete-case analyses were similar.

In the analysis of postdiagnosis diet and NMIBC outcomes, 466 patients had at least one recurrence, 139 had multiple recurrences, and 153 showed progression. Each 100 g/day higher total fruit and vegetable intake at 3 months postdiagnosis was associated with lower first recurrence risk (HR 0.91, 95% CI 0.85–0.99), and each 100 g/day increase from prediagnosis to 3 months postdiagnosis was also associated with lower first recurrence risk (HR 0.90, 95% CI 0.81–0.99). Vegetable intake moved in the same direction, and spline analyses suggested the linear association was evident only above the median.

Prediagnosis fruit and vegetable intake was not associated with recurrence or progression, multiple-recurrence associations were attenuated, and total and individual dietary and plasma carotenoids were not associated with first recurrence, multiple recurrences, or progression. No clear overall progression association was seen, although two related 15-month fruit findings were associated with higher progression risk—higher fruit intake at 15 months postdiagnosis (HR 1.15, 95% CI 1.03–1.29) and greater increase in fruit intake from prediagnosis to 15 months (HR 1.14, 95% CI 1.01–1.28)—which the authors cautioned could reflect chance. Significant subgroup signals for lower first recurrence were limited to women, former smokers, Ta disease, low or intermediate EAU risk, and patients not treated with bacillus Calmette-Guérin (BCG), with attenuation at 15 months.

Because diet was self-reported, recall error and non-differential misclassification may have affected estimates, especially for prediagnosis intake. The authors noted that plasma carotenoids may not fully represent fruit and vegetable exposure because preparation methods, digestion, absorption, metabolism, and unmeasured urinary metabolites can weaken the link between intake and blood levels. Plasma lipid data were unavailable for adjustment, biomarker analyses came from a subset enrolled between 2014 and 2017, and progression analyses involved fewer events, which is why the unexpected 15-month fruit signal was framed as a possible chance finding.

The authors concluded that the recurrence signal in NMIBC was concentrated in postdiagnosis fruit and vegetable intake, particularly at 3 months and in early increases after diagnosis, while carotenoid measures did not track with outcomes. They cautioned that confirmation is still needed before definitive conclusions or dietary recommendations are made.

Clinician Questions

Which patients with NMIBC were represented in the UroLife diet analysis?

The cohort represented Dutch-speaking adults aged 18 to 80 years with newly diagnosed, histologically confirmed Ta, T1, or Tis NMIBC after TURBT who were recruited from 22 hospitals in the Netherlands. Because only stage T1 continued to be enrolled from May 2017 onward, the findings apply most directly to that newly diagnosed post-TURBT Dutch surveillance population.

Why might plasma carotenoids not have matched the fruit-and-vegetable recurrence signal in NMIBC?

Plasma carotenoids may not fully capture fruit and vegetable exposure because food preparation, digestion, absorption, and metabolism can all affect blood concentrations, and urinary metabolites were not measured. The authors also reported weak agreement between reported intake and plasma measures, leaving open the possibility that compounds other than carotenoids may be more relevant to recurrence biology.

How did the investigators handle 15-month diet data when analyzing recurrence and progression in NMIBC?

The 15-month questionnaires were treated as markers of habitual postdiagnosis exposure rather than as a new starting point for risk accumulation. Follow-up for those analyses therefore began on the same date used for the 3-month analyses while testing whether later-reported diet reflected longer-term patterns.

What remains unresolved after this UroLife analysis of diet and NMIBC outcomes?

Whether the observed recurrence association is reproducible enough to support clinical dietary guidance remains unresolved. The authors highlighted the smaller number of progression events, the possibility that the 15-month fruit-progression signal was a chance finding, and the need to study other biomarkers or bioactive metabolites of fruit and vegetable intake in NMIBC.