1. Preconceptional Health Promotion in Low- Income Rural Communities: Randomized Trial Results From The Central Pennsylvania Women’s Health Study (CePAWHS) Marianne M. Hillemeier, Carol S. Weisman, John J. Botti, Danielle Symons Downs, Mark Feinberg, Gary A. Chase, Sara A. Baker, Anne-Marie Dyer, and Diana Velott The Pennsylvania State University

2. To conduct a randomized trial of a new behavior change intervention – Strong Healthy Women – designed to improve the health and health behaviors of pre- and interconceptional women in low-income rural communities Research Objectives

3. Prevalent Risk Factors in CePAWHS Population-based Survey, Compared with PA and U.S.* (Women ages 18-45, weighted data) CePAWHS PA U.S. Obesity (BMI = 30+) 23% 18% 19% Depression/anxiety dx 29% -- 16% Nutritional deficits: Fruit < daily 68% 57% 60% Vegetables < daily 56% 31% 34% Alcohol use (ever) 48% -- 32% Binge drinking (among drinkers) 34% 29% 23% Smoking 28% 32% 23% Folic acid supplementation 38% 53% 50% * Comparison data sets include BRFSS 2003, Commonwealth Fund Survey of Women’s Health 1998, National Health Interview Survey 2003

4. Curriculum developed by team led by Drs. Mark Feinberg, Prevention Research Center and Danielle Symons Downs, Department of Kinesiology, Pennsylvania State University

5. Intervention Health information & behavior change skills Intention to change behavior and self-efficacy Health behavior change Health status Improvements Long-term Outcomes Unintended pregnancies Complications Preterm birth LBW

6. Intervention Framework and Outcomes Risk Dimension Behavior Change Goals Learning Objective (Example ) Behavioral Objective (Example) Stress Manage- ment Decrease psychosocial stress Understand causes of stress and behavioral responses Practice relaxation techniques Nutrition Increase healthy food choices Understand nutrition and identify barriers to healthy eating Eat healthier foods Physical Activity Achieve exercise recommendations Understand guidelines and practice exercises Exercise regularly per guidelines Tobacco/ Alcohol Decrease tobacco & alcohol use and exposure Understand impact on pregnancy, triggers, and alternatives Decrease behaviors and exposures Gynecolo- gic Infections Decrease gynecologic infx Understand causes and infection Decrease risk behaviors and seek care Pregnancy planning Strategize for pregnancy planning Understand maternal health and contraception Discuss plan with provider; use folic acid

7. Intervention Process Six 2-hour sessions over 12-weeks Groups facilitated by 2 trained personnel Mix of topics at each session Active learning (discussions, physical activity sessions, cooking, handouts) Goal-setting (“baby steps”) Social support (buddy system; facilitator phone calls) Incentives (gift cards, supplies)

8. A group using core balls

9. Study Design Recruitment (n = 692) Baseline Risk Assessment Random Assignment Intervention (12 weeks)Control Follow-up Risk Assessment Follow-up telephone surveys at 6 and 12 months; birth records appended

10. Risk Assessment Content (Behavioral and Biological Markers) Questionnaire (health status, health behaviors, psychosocial stress, access to health care, behavioral intent, self-efficacy, etc.) Anthropometric measurements (height, weight, BMI, waist circumference) Blood pressure Non-fasting blood glucose and lipid panel using fingerstick blood and CardioChek analyzer

11. Intervention Communities (n = 15)

12. Eligibility Age 18-35 years at enrollment* Resides in Central PA target area Not pregnant at enrollment Capable of becoming pregnant (no hysterectomy or tubal ligation) Exclusions: non-English speaking ______________________ * This age group accounts for > 85% of pregnancies in Central PA

14. Recruitment Methods Tailored to communities with help of local Steering Committee members Active One-on-one recruitment at social service agencies (e.g., WIC programs, childcare), schools, and retail settings Passive Presentations in social service agencies, schools Posters and tear-off flyers in businesses, churches, community centers Kiosks at local health fairs, farm shows Inserts in utility bills Postcards to parents of subsidized child care

16. Randomized Trial Enrollees, Compared with Pre- and Interconceptional Women Ages 18-35 in Target Counties Target Counties * Enrolleesp-value (n = 257)(n = 692) Poor or near poor a 34%63% <.0001 Rural b 33%51% <.0001 Education < college35%41% ns Non-white 3% 9%.003 Unmarried 28%49% <.0001 No usual source of care 7%24% <.0001 No health insurance 20%29%.004 Preconceptional c 37%43% ns * CePAWHS survey a Based on federal poverty level b Based on zipcode-based RUCA codes C Never pregnant

17. Total Phase II Enrollees n = 692 Intervention 473 (68%) Control 219 (32%) Did not complete FURA 221 (47%) Completed FURA 252 (53%) Completed FURA 110 (50%) Did not Complete FURA 109 (50%) Base n for pre-post change in biomarkers = 362 Telephone interviews (ongoing) Telephone interviews (ongoing) Mean number of sessions attended = 3.9 0 (14%) 1 – 2 (11%) 3 – 4 (21%) 5 – 6 (54%)

18. Key Retention Issues Those who dropped out of the study tended to be younger and to have lower educational levels than those who were retained. ProblemSolution Inadequate inter-sessionImprove training reminder contacts by and supervision of group facilitatorsgroup facilitators; use email and telephone contacts Lack of reliable childcareProvide on-site childcare

19. Fidelity Videotaped group sessions were rated for: –Percent of content delivered in each domain (77% overall; worst for smoking reduction) –Group facilitator performance (e.g., engaging and involved; confident; provides positive feedback to participants) –Variation across sites (found only for facilitator performance)

20. Statistical Analyses Intent-to-treat pre-post analyses conducted with analysis of variance Baseline measure (pre-test) included as covariate to adjust for differences in baseline measures; follow-up (post-test) measure is the dependent variable Used GLM, ordinal logistic regression, or ordinary logistic regression (depending on dependent variable) Age and educational attainment controlled to account for potential confounding

21. Significant Pre-Post Intervention Effects NOTE: GLM and logistic regression models also included pre-intervention measurement, age, and educational attainment

22. Pre-Post Biomarker Assessments NOTE: GLM models also included pre-intervention measurement, age, and educational attainment

23. Significant Intervention Dose Effects Among Intervention Participants NOTE: GLM and logistic regression models also included pre-intervention measurement, age, and educational attainment

24. Next Steps in Analyses Moderation effects (e.g., Did the intervention work better for certain subgroups of women? Did degree of fidelity affect impact?) Maintenance of attitudinal and behavior change over the 12-month follow-up period Any impact on pregnancy complications or outcomes during follow-up period (~ 115 live births occurred)

25. Future Research Plans Modify the Strong Healthy Women intervention based on Phase II data Test the modified intervention in a randomized trial of alternative modes of delivery in a more diverse target population

26. Acknowledgements Funded, in part, by a grant from the Pennsylvania Department of Health Collaborating organizations: Pennsylvania State University Family Health Council of Central Pennsylvania Franklin & Marshall College Lock Haven University of Pennsylvania Steering Committee of community representatives