Longitudinal studies can present a number of challenges, especially retention of participants. Initial recruiting oversampled the population of interest in anticipation of some participants leaving the study. Data analyses requiring a large sample size were completed early in the study. In this study we have been particularly interested in retaining participants who are not on hormones. This subgroup was oversampled during the initial recruiting and now comprises around 32% of the total sample of 205 women. Another challenge has been identifying when menopause occurs for women who are on hormones. Participants are asked if they went 12 months with no period before beginning hormones.

We have also addressed the challenge of retention by assigning each research associate a caseload, facilitating regular personal contact with participants. Longevity and flexibility of the staff have been key characteristics as well. Participants are paid yearly, receive birthday cards and received a certificate of participation. Two social gatherings/health information fairs were held for participants in November 2002 and February 2003. The 3rd issue of our newsletter for participants was mailed in December 2002. This web site was created in response to participants' interest in learning more about the study and about menopause in general.

Participants have received a written summary of group data. This web site and the newsletter provide information about how the study is carried out, some preliminary findings, and give general resources about menopause. At the end of the study, participants will receive the final written summary.

We are attempting to avoid missing data over time. We have given participants a variety of helpful aids to assist them in keeping track of menstrual data and the correct days to complete the diary and collect the urine sample, as well as the date by when the samples must be given to the research team. These include phone calls and emails, reminder stickers on the 12 month menstrual calendars, and stickers with the community site drop-off dates to put in their personal calendars. They are reminded to write any medication changes, particularly changes in hormones, on the back of the menstrual calendar. All participants are asked to follow the same schedule for data collection and are encouraged to call their research associate for clarification. Questions and scales used in interviews and questionnaires are identical for all participants.

Determining the reliability of questionnaire data is difficult. Participants use varying definitions of key terms such as cycle, period, irregular, irregularity, and skipped period. Some participants may have difficulty with recall of information and some may have a low level of tolerance for filling out detailed questionnaires.