Hormonal Influences on Inflammatory Bowel Diseases

Aim 1. Identify preferences and reasons for contraceptive method selection (or non-use) and willingness to participate in a future randomized controlled trial in 200 participants with IBD (100 naturally cycling and 100 hormonal contraception users). Our a priori cut off requires ≥ 75% of participants accepting randomization.

Hypothesis: The majority of participants will accept randomization if they may switch methods if not satisfied.

Aim 2. Obtain estimates of means and standard deviations for the validated Crohn's Disease and Ulcerative Colitis PRO Instruments across the menstrual cycle (menses, ovulation, pre-menstrual) in the naturally-cycling participants and between bleeding and non-bleeding days in the hormonal contraception users.

Hypothesis: IBD PRO instruments will reliably reflect IBD symptom changes related to menstrual phases.

Aim 3. Assess the correlation between inflammatory marker changes (fecal calprotectin \& high sensitivity C-reactive protein), menstrual timing and/or bleeding days, and IBD PRO responses in a subset of 30% of Aim 1 participants (30 naturally cycling and 30 hormonal contraception users).

Hypothesis: Changes in inflammatory markers will be correlated with menstrual phases and IBD PRO responses.

Study Design: A pilot prospective cohort study of 200 women with IBD (100 naturally cycling and 100 hormonal contraception users).

Recruitment. We will recruit via four different sources: 1. With the support of the co-investigators who are IBD subspecialists at the University of Utah (Drs. Flynn and Johnson), we will send recruitment emails to their patients who meet inclusion criteria via the Epic patient portal (MyChart). 2. We will post fliers regarding the study in the gastroenterology, primary care, and women's health clinics across University of Utah sites. 3. We will send recruitment materials to other non-university gastroenterology sites in Utah. 4. We will advertise via social media nationally with a link to a University of Utah Research Enterprise Data Capture (REDCap) survey to assess eligibility and provide contact information for study coordinators to connect for a phone screen.

AIM 1 Sample Size: This is a pilot study designed to collect baseline data to inform future trials and not powered to assess statistical differences between study groups. We plan to recruit 200 participants in an IBD setting who represent common contraceptive method choices and compare hormonal method users to non-users regarding preferences in method selection. The sample size will also give a breadth of perspectives on willingness to be randomized to different hormonal methods under variable study conditions and designs to influence a future trial. Our sample size takes into consideration feasibility for recruitment given our recruitment timeline, as well as sample size recommendations for pilot clinical trials.

Participant procedures: Total study engagement is 12 weeks. Enrollment processes. Study staff will complete a phone screening, discuss study procedures, and if they meet inclusion criteria, either email the potential participant the consent and enrollment survey for completion via REDCap or schedule an in-person enrollment visit if participating in the Aim 3 sub study to allow for inflammatory marker collection. The screening phone call will include request of documentation confirming IBD diagnosis and screening questions regarding timing of IBD diagnosis, current disease activity, prior endoscopy and surgery dates, current and prior treatments, and any hospitalizations. These responses and documentation of diagnoses will be reviewed for eligibility by the IBD subspecialists before consent and enrollment is completed.

Survey content. The enrollment survey will assess sociodemographic and reproductive characteristics, IBD history as above, menstrual cycle timing and symptoms including bleeding characteristics and pain scores, sexual satisfaction surveys, baseline IBD symptom PRO instruments, menstrual-related IBD symptoms, prior and current contraceptive experiences, and preferences regarding contraceptive methods. Using an adapted 7-item Attitudes to Randomized Trials Questionnaire, participants will be presented with hypothetical RCT scenarios to assess willingness to be randomized based on different methods, different study designs (e.g. participant preference arm), options to switch methods, and then query regarding reasoning if not willing to be randomized.

Study Arms: The "naturally-cycling" cohort will include participants who are not using any form of hormonal contraception and meet inclusion criteria with regular menses. This may include non-hormonal method users, including abstinence, copper IUD, self or partner permanent contraception, barrier methods, or fertility awareness methods. The hormonal contraception user cohort includes the most common hormonal contraceptive options (COC, LNG IUD, ENG implant). The hormonal contraceptive method must have been initiated prior to study enrollment, thus the method choice made outside of this study is not an intervention. If a participant initiates, stops, or switches methods during the study, they will be analyzed in the assigned cohort at enrollment, as changes in hormonal methods may not result in consistent bleeding profiles or resumption of natural cycles for weeks to months.

Follow-up. While the aim 1 outcome assessment occurs with the enrollment survey, all participants will be followed for Aim 2 and a subset for Aim 3 for 12 weeks from the time of enrollment. Participants will receive compensation for time with each completed study activity.

Analyses. The goal of Aim 1 is to define the characteristics of the IBD study population and determine a point estimate of the proportion of participants willing to be randomized to one of three contraceptive methods under each scenario characteristics. A 95% confidence interval around this estimate will describe its variability. The 7-item Attitudes to Randomized Trials Questionnaire30 is not scored in sum, but each item adds characteristics related to the proposed study design that are reported via descriptive statistics as positive or negative attributes influencing randomization willingness. We will report contraceptive method preferences and IBD-specific reasons for method selection (and non-use of hormonal methods) via descriptive statistics, inferential analysis exploring relationships between respondent characteristics and method preferences, adjusting for participant characteristics as well as thematic findings from free text response options. For free text responses, we will use frequency distributions as well as Latent Dirichlet Allocation, a topic modeling technique to identify common themes and patterns in the responses. We will also conduct comparative analysis to compare IBD-specific reasons for method selection or non-use of hormonal methods across different demographics and disease characteristics. While the enrollment survey and all daily and weekly PRO instruments will be administered electronically to all participants, we anticipate that certain significant differences may also exist between participants who agree to enroll in Aim 3 and those who do not. This may be due to their willingness to come for in-person lab assessments and engagement with the study team. We will thus perform sensitivity analyses to determine, for each of these groups, the proportion willing to be randomized and to assess contraceptive method preferences. Similarly, we will perform sensitivity analyses for both those who switch, discontinue or initiate a contraceptive method and for those who are abstinent and may be using hormonal contraception for non-contraceptive purposes only compared to those who need contraception for pregnancy prevention.

AIM 2 Sample size. All study participants recruited in Aim 1 will be included in this aim. We anticipate a 5% attrition in survey completion over 12 weeks.

Procedures. Participants will receive electronic survey links to assess IBD-related and menstrual-related symptoms, QOL measures, sexual activity and health changes across menstrual cycles and bleeding and non-bleeding days in the hormonal contraception users over 12 weeks. Via REDCap, we will send automated reminders and deliver links for data completion through text messages to the participant's mobile phone or device. Participant responses automatically upload from the phone or device to REDCap.

Daily surveys. Daily surveys will begin the day of enrollment and continue for 12 weeks. The survey questions include either the CD or UC PRO instrument depending upon diagnosis. The daily IBD PRO instruments include modules on signs and symptoms, systemic symptoms, and coping strategies with a total of 17 questions (6 minutes) for the CD PRO and 20 questions (7 minutes) for the UC PRO. We will also query chronic pelvic pain severity, vaginal bleeding and amount, and if bleeding, menstrual-related pain and severity (4 follow-up questions if bleeding).

Weekly Surveys. The weekly surveys will begin on day of enrollment and continue every week for 12 weeks (13 total with the enrollment survey). These longer surveys will include the daily questions plus the IBD PRO modules on daily life impact with 9 questions (3 minutes) and emotional impact with 8 questions (3 minutes). Participants will complete a Likert scale on how bothersome IBD symptoms have been in the past week, and report any sexual activity with assessment of dyspareunia, additional treatments for IBD or menstrual symptoms, and any changes to health history.

Analyses. This aim will assess changes in IBD PRO responses over subsequent menstrual cycles and between bleeding and non-bleeding days for hormonal contraceptive users. The UC PRO and CD PRO instrument modules will all be scored based on guidelines for instrument use. Report of bothersome IBD symptoms, chronic pelvic pain, dysmenorrhea, and dyspareunia will be on a Likert scale from 0-10. Any other reported changes in health, treatments or concerns will be reported by frequency and cycle timing. We will compare variations in these means across menstrual cycle timing using linear regression and compare means in symptom scores between cohorts using paired t-tests.

AIM 3 Sample size. We will enroll 30% of each cohort from Aim 1 in this sub study for a total of 60 participants.

Procedures. Potential participants eligible for Aim 1 who are willing and able to come to a study enrollment visit and 12 subsequent lab visits at the University of Utah will be offered participation in this Aim 3 sub study. For those interested, their enrollment will occur in-person, rather than virtually in order to initiate the inflammatory marker collection. At the enrollment visit, the research staff review the same electronic consents used for participants in Aim 1 and 2 and prompt completion of the enrollment survey using the University of Utah's secure, web-based REDCap system. The enrollment visit will also include urine pregnancy testing, lab testing (complete blood count, hsCRP, ferritin, and albumin) and they will receive a kit to collect fecal calprotectin from a home stool sample and return it at the 1-week blood draw. Lab assessments of inflammatory markers will occur at varying intervals. WBC, ferritin and albumin will be assessed at baseline screening visit and again monthly (4 total). The hsCRP will be measured at baseline and then occur weekly (13 total). Fecal calprotectin (from 1st morning stool) will be measured at baseline (sample collected at 1-week blood draw after enrollment) then monthly for 12 weeks (4 total). The blood draws and fecal calprotectin assessments that occur after the single in-person enrollment visit can be completed at local University labs throughout the region.

Analyses. We will compare the proportion of participants with hsCRP and fecal calprotectin levels within a 'normal' range by cohort (naturally cycling vs hormonal contraception users) using chi square. We will compare variations in means across menstrual cycles using mixed effects models. We will look for difference in inflammatory marker levels across IBD PRO responses and specific modules by including these PROs as covariates in our regression models, and will include a random intercept in our models for participants in order to account for variations in individual baseline marker levels.