A case-crossover analysis of pollen as a urologic chronic pelvic pain syndrome flare trigger in the Multidisciplinary Approach to the Study of Chronic Pelvic Pain (MAPP) Research Network

Javed I1, Li J1, Pakpahan R1, Milbrandt M1, Andriole G1, Lai H1, Colditz G1, Sutcliffe S1

Research Type


Abstract Category

Pelvic Pain Syndromes / Sexual Dysfunction

Abstract 537
Urogynaecology 5
Scientific Podium Short Oral Session 29
Friday 6th September 2019
12:37 - 12:45
Hall G3
Painful Bladder Syndrome/Interstitial Cystitis (IC) Prevention Prospective Study
1.Washington University in Saint Louis

Siobhan Sutcliffe



Hypothesis / aims of study
Pollen has anecdotally been reported to trigger flares in patients with urologic chronic pelvic pain syndrome (UCPPS), a syndrome that frequently co-exists with allergies. To test the hypothesis that pollen triggers UCPPS flares, we linked local pollen count data with our case-crossover study of flare triggers embedded in the Multidisciplinary Approach to the Study of Chronic Pelvic Pain (MAPP) Epidemiology and Phenotyping longitudinal Study.
Study design, materials and methods
The MAPP Epidemiology and Phenotyping Study was a one-year longitudinal study of UCPPS patients designed to investigate the “usual-care” natural history of UCPPS and to identify patient sub-groups with possible differing etiology and clinical course. It included women with interstitial cystitis/bladder pain syndrome and men with interstitial cystitis/bladder pain syndrome or chronic prostatitis/chronic pelvic pain syndrome. Participants completed a lengthy battery of questionnaires at biannual study visits and a shorter set of questionnaires at biweekly online assessments.

To investigate flare triggers, participants were asked whether they were currently “experiencing a flare of their urologic or pelvic pain symptoms […] meaning symptoms that are much worse than usual” at each biweekly assessment. Participants who responded affirmatively were asked to complete an additional set of questions about their flare symptoms, flare start date, and exposures in the 3 days leading up to their flare. These questions were asked for the first 3 flares and at 3 randomly selected times when participants did not report a flare. 

We limited the case-crossover analysis to participants who completed at least one flare and one non-flare assessment (n=290 of 422 participants). Flare and non-flare data were linked to daily pollen count values (obtained from IMS Health) by 3-digit residential zip code. Pollen count levels in the 3 days before (day -1, day -2, and day -3) and the day of a flare (day 0), as well as daily changes in these values, were compared to corresponding non-flare values by conditional logistic regression. Poisson regression was used to estimate flare rates in the 3 weeks following increases in pollen count in the full longitudinal study. Analyses were performed in all participants and separately in those who reported allergies. With a sample size of 290 participants, we had at least 80% power to detect odds ratios as small as 1.6-1.7, assuming 1:1 matching, a prevalence of exposure in controls of 20-30%, a correlation between observations from the same participant of 0.1-0.2, two-sided tests, and an α-level of 0.05.
Overall, the 290 eligible participants contributed information on 574 flares and 791 non-flare time periods. Pollen count values ranged from 0.1 to 11.8 (on a scale of 12) during follow-up. Pollen count on any of the 3 days before or the day of a flare was not associated with flare onset in the full case-crossover sample or among participants with allergies (Table 1). In general, pollen count increases of at least one level were also not associated with flare onset, although a significant positive association was observed for an increase in pollen in the two days before a flare in the case-crossover sample. No association was observed between the start of a period of time with increased pollen counts and flare rates in the full longitudinal study.
Interpretation of results
In general, no consistent pattern of associations was observed between pollen count and the onset or frequency of flares.
Concluding message
Overall, we did not find strong evidence to support a role for pollen as a flare trigger, although we cannot rule out this possibility entirely, at least in a subset of participants. If pollen is ultimately found to trigger flares, these could potentially be prevented and/or treated by pollen avoidance and anti-histamines in susceptible patients.
Figure 1 Table 1: Associations between pollen count in the past three days and UCPPS flare onset, Multidisciplinary Approach to the Study of Chronic Pelvic Pain Epidemiology and Phenotyping Study, 2009-2013.
Funding Funded by NIDDK grants: U01 DK082315, U01 DK82316, U01 DK82325, U01 DK82333, U01 DK82342, U01 DK82344, U01 DK82345, and U01 DK82370 Clinical Trial No Subjects Human Ethics Committee The MAPP Study was approved by the Institutional Review Board at each participating site and the Data Coordinating Center. All participants provided written informed consent. Helsinki Yes Informed Consent Yes