AEP Vol. 10, No. 7 October 2000: 449–483
ABSTRACTS (ACE)
asked participants to do, African-Americans indicated they were less likely to participate than Whites on only 2 specific prompts (p ⬍ 0.001 for each), more likely to participate on 2 prompts (p ⬍ 0.001), and were equally or near equally likely to participate on 5 prompts. CONCLUSIONS: The findings from this study show that African-Americans self-reported a lower willingness to participate in biomedical studies than did non-Hispanic Whites, and that African-American subjects were more influenced by the factor of ‘who’ was running the study than by the factor of ‘what’ they might have to do as study subjects. Study supported by grant P 50 DE10592 from the NIDCR at NIH. PII S1047-2797(00)00153-8
ELECTRONIC VERSUS PAPER INSTRUMENTS FOR DAILY DATA COLLECTION CB Johannes, J Woods, S Crawford, H Cochran, D Tran, B Schuth, New England Research Institutes, Watertown, MA PURPOSE: Collecting daily information on a series of similar, recurring events such as menstrual bleeding, headaches, or insulin levels using paper instruments is subject to problems such as missing or incorrectly recorded data, and retrospective data entry. The authors are developing and evaluating electronic data collection using hand-held personal computers (H/PC) for a variety of healthrelated applications, including tracking premenstrual syndrome (PMS), fertility awareness, and headaches, to improve accuracy and timeliness of data collection. METHODS: ProCycle is a prototype electronic diary for collection of daily data on menstrual bleeding, medications, and health symptoms. In a 3-month pilot test in 25 regularly cycling women, we compared its performance with a paper calendar regarding missing and incorrect data, data entry lag, data cleaning time, and users’ preferences with respect to factors such as remembering to enter data, convenience, and overall preference. Additional programs for PMS, fertility, and headaches are being field tested on subjects from the Boston and New York areas, comparing performance with paper versions. RESULTS: In the pilot test, missing data occurred less frequently with ProCycle than with paper, particularly for any days with missing symptoms (4% vs. 35%, p ⬍ 0.05). ProCycle did not permit any data recording mistakes such as circling contradictory responses (e.g., bleeding and no bleeding), compared with incorrect data on 13% of paper calendars. Data entry/cleaning time was 81% lower for ProCycle. 70% of users preferred ProCycle overall, compared with 9% preferring paper (p ⬍ 0.01). CONCLUSIONS: Although the initial cost of the H/PC is significantly higher than paper, there are no recurring charges for printing or data entry, and data cleaning is minimal. Electronic instruments on an H/PC provide an efficient, accurate method of data collection, applicable to a number of areas of health-related research involving daily data collection. PII S1047-2797(00)00141-1
457
DIGIT PREFERENCE IN YEAR AT MENOPAUSE: DATA FROM THE STUDY OF WOMEN’S HEALTH ACROSS THE NATION SL Crawford, CB Johannes, J Bradsher, RK Stellato, S Sherman, S Samuels, New England Research Institutes, Watertown, MA PURPOSE: To assess possible digit preference in self-reported year at natural menopause, and to determine whether equal proportions is an appropriate reference distribution. METHODS: Data are from the cross-sectional telephone interview from the Study of Women’s Health Across the Nation (SWAN), a multi-site, multi-ethnic study of women aged 40–55. Analyses included 2151 naturally menopausal respondents, who were asked the year their periods stopped. Using a chi-square test, we compared the distribution of the terminal digit for year of final menstrual period (FMP) to an equal proportions (EP) distribution assigning 10% probability of each of the 10 digits. Departures from EP, however, may be accurate and may reflect the observed age distribution of subjects rather than digit preference. Thus, we stratified by year of interview to determine if the distribution was the same across years—suggesting digit preference—or if it varied systematically. We then estimated an expected distribution of terminal digit for year of FMP, using prospectively collected data (not subject to digit preference or recall bias) from the Massachusetts Women’s Health Study, applied to the SWAN age distribution. The SWAN terminal digit distribution was compared with this expected distribution. RESULTS: Terminal digit for year at FMP exhibited a strong departure from EP (2 ⫽ 841.60, p ⬍ 0.001), with highest frequencies for digits 4 and 5. Stratifying by year of interview, the distribution was shifted one digit in 1997 compared with 1996, consistent with unbiased reporting. Using the expected distribution as the reference reduced the chi-square statistic by a factor of 7 (2 ⫽ 119.51, p ⬍ 0.001). CONCLUSIONS: The distribution of terminal digits for reported year at FMP was far from uniform, but departures from EP were consistent with results expected from an independent prospective study. These results suggest that EP may not be an appropriate reference for studies of digit preference, particularly those with a restricted age range. PII S1047-2797(00)00140-X
CHRONIC DISEASE TRENDS IN REPRODUCTIVE, SMOKING, AND OTHER CHRONIC DISEASE RISK FACTORS BY BIRTH COHORT IN A LARGE OCCUPATIONAL STUDY DM Freedman, MS Linet, MM Doody, AK Mohan, BH Alexander, JD Boice Jr, JS Mandel, RE Tarone, Radiation Epidemiology Branch, National Cancer Institute, Rockville, MD PURPOSE: To illustrate the value of using large cohort studies to identify birth cohort trends in several chronic disease risk factors. METHODS: In collaboration with the American Registry of Radiologic Technologists (ARRT) and the University of Minnesota, the National Cancer Institute (NCI) initiated a cohort
458
ABSTRACTS (ACE)
AEP Vol. 10, No. 7 October 2000: 449–483
study of radiologic technologists who were certified by ARRT for at least two years between 1926 and 1982. Over 90,000 technologists (nearly four-fifths female) from all 50 states responded to a mailed questionnaire on reproductive, medical, work, and lifestyle factors. Ten, mostly five-year, birth cohorts, from before 1920 through 1960 and later, were evaluated. RESULTS: In this population, the mean height of both men and women generally rose in each subsequent birth cohort. The proportion of men who smoked before age 18 fell among those born since the late 1920s. In contrast, the proportion of women smoking before age 18 rose among those born since the early 1950s, reaching 14.2% among those born in 1960 and later. The mean age at menarche fell, until leveling off at 12.5, among those born after 1940. Recent birth cohorts (since 1950) show among the highest mean ages at birth of first child (⬎26 yeras), highest rates of nulliparity at age 25 (⭓63%), and lowest mean parity levels (⭐1.7), compared with earlier birth cohorts. The proportion of women ever using oral contraceptives rose steadily by birth cohort, reaching 86 percent in the most recent birth cohort. CONCLUSIONS: Some of the trends observed may help explain historic chronic disease incidence patterns and predict future disease burdens. PII S1047-2797(00)00159-9
PSYCHIATRIC CORRELATES IN CHRONIC FATIGUE SYNDROME EK Axe, P Satz, Departments of Epidemiology and Neuropsychology, University of California–Los Angeles, Los Angeles, CA PURPOSE: This study presents psychiatric correlates in Chronic Fatigue Syndrome (CFS) that emerged from the CDC’s Surveillance Study. It seeks to determine the time of onset and rates of syndromal psychiatric disorders and identify the predominant disorder. Other goals are to ascertain whether depression is associated with CFS symptomatology, compare syndromal to selfreported depression, and test for the specificity of the 1988 CDC case definition for CFS. METHODS: All 565 enrolled subjects had fatiguing illnesses and were evaluated for CFS. They completed the Diagnostic Interview Schedule for the DSM-III-R and the Beck Depression Inventory. Prevalence estimates for current syndromal psychiatric disorders were calculated. CFS symptoms were compared by depression status. Syndromal and self-reported depression were contrasted. Groups that did and did not meet the case definition were compared by three outcome variables. RESULTS: Rates of current psychiatric disorders were high in CDC subjects compared to the community. The predominant disorder was depression. Although prior disorders tended to persist (75%), many disorders were incident to the fatiguing illness (57%). Depression was not associated with increased CFS symptomatology. There was only weak agreement between measures of syndromal and self-reported depression (kappa ⫽ 0.3219). Subjects designated as CFS had similar rates of syndromal psychiatric
disorders, syndromal depression, and self-reported depression as did non-CFS subjects. CONCLUSIONS: Current syndrome; psychiatric disorders appear associated with fatiguing illnesses. While prior psychiatric disorders are risk factors for current, the onset was largely concurrent with the fatiguing illnesses. The BDI should probably not be used as a measure for psychiatric morbidity in CFS subjects. Regardless of outcome, there was no evidence of specificity of psychiatric features to the CDC case definition. PII S1047-2797(00)00083-1
LONGITUDINAL ANALYSIS OF SYMPTOMS REPORTED BY PATIENTS WITH CHRONIC FATIGUE SYNDROME R Nisenbaum, AB Jones, JF Jones, WC Reeves, Centers for Disease Control and Prevention, Atlanta, GA PURPOSE: To determine the effect of chronic fatigue syndrome (CFS) illness duration and onset type on the likelihood of reporting a symptom during successive follow-up periods. METHODS: In 1997, a two-phase RDD survey in Wichita, Kansas, was conducted to estimate the prevalence of CFS. Phase I identified 56,154 respondents 18–69 years of age and screened for severe fatigue, extreme tiredness or exhaustion lasting for 1 month or longer. In phase II an equal number of fatigued (n ⫽ 7,176) and randomly selected non-fatigued subjects were asked about 8 CFS and 13 non-CFS symptoms, as well as the presence of specific medical and psychiatric conditions. Eligible respondents were clinically evaluated to establish CFS diagnosis. Phase II respondents were re-contacted at 12- (n ⫽ 4,331) and 24-months (n ⫽ 4,266) for additional follow-up and diagnosis. In this study we considered symptoms reported as being present most of the time during each successive observation period. Generalized estimating equations were used to model symptoms over time and to address study questions. Such a model accounts for correlations among repeated symptoms for each subject. We used an auto-regressive structure for the correlation matrix, assuming the correlations between each pair of repeated symptoms should decrease as the time between symptoms increased. RESULTS: There were 74 CFS patients who had been ill for 1 to 20 years (median ⫽ 6.3 years). Among these, 46 reported gradual and 28 reported sudden onset. Symptoms fluctuated over the course of illness. However, only stomach pain (non-CFS symptom) was more likely to be reported as duration of illness increased (p ⬍ 0.05). There was no association between onset type and the likelihood of reporting a symptom during an interview, except that chills and severe headaches were more likely to be reported by sudden cases. CONCLUSIONS: The likelihood of expressing CFS and nonCFS symptom “most of the time” is the same across years of illness. More analyses are warranted to consider expression of symptoms for ⭓6 months and severe symptoms. PII S1047-2797(00)00119-8