The Adventist Health Study-1: Gathering Data
The Adventist Health Study, a cohort investigation which began in 1974, had some very basic differences from the earlier mortality studies.
For one, it was designed to find out which components of the Adventist lifestyle give protection against disease. Never has it been a major aim of this study to compare the rates of disease or mortality between Adventists and non-Adventists. Another difference was the addition of data reporting nonfatal, as well as fatal, disease events. The Adventist Health Study-1 also added a more detailed investigation of diet. In the beginning, the Adventist Health Study-1 was primarily a cancer investigation. In 1981, the cardiovascular component was added.
The initial mailing of 63,530 census questionnaires was sent to every Adventist household in California in an attempt to enroll every adult in these households over the age of 25 years.
Two years later, a second more extensive lifestyle questionnaire was sent to those who had responded to the census questionnaire. This second questionnaire included previous medical history, drug therapy, more than 60 dietary questions, a variety of psychosocial questions, and questions about physical activity. The response rate from White non-Hispanic subjects was the highest of any ethnic group at 75 percent, and numbered 34,198 individuals.
For the next six years, researchers followed this group closely, looking for both fatal and nonfatal cases of either cancer or heart disease. A short questionnaire was mailed to each individual once a year, asking about any hospitalization within the previous 12 months and, if so, where the hospitalization occurred.
Subsequently, a field staff of Adventist Health Study researchers was dispatched to visit the medical records departments of the hospitals mentioned by the subjects. The field staff reviewed all hospital records of each admission, looking for any hint of cancer or coronary heart disease that took place in a hospital. For suspected acute coronary heart disease events, copies of all physician records, electrocardiograms, and abstractions of serum enzyme reports were obtained. For suspected cancer cases, copies of histology and other relevant tests were taken.
Fatal cases of cancer or heart disease, where death took place outside the hospital, were found using several means. Annual questionnaires were often returned by friends or relatives for some deceased individuals. Church records were often available as well. In addition, researchers were able to compare these records to state death records for California. After 1978, the National Death Index became available, helping researchers track deaths for those who had moved away from California.
Diagnosing a new cancer event required a histological confirmation of a malignant tumor, this being made sometime after the subject had already entered the study. Confirming a new case of acute myocardial infarction or fatal coronary heart disease was made by applying published international diagnostic criteria for these events.
Great effort was made to establish the validity and accuracy of the questionnaire information initially gathered. When compared to the information gained from face-to-face interviews, the data obtained through the questionnaire and follow-up was remarkably accurate. Missing data occurred in between one and seven percent of the responses for different variables. Interestingly, the rate for missing data did not increase for questions that might be sensitive to Adventists, such as alcohol consumption, eating pork and other meats, or drinking coffee.
When the gathering of follow-up data was officially concluded for the Adventist Health Study, 32,000 hospitalizations were reported, representing 28,000 hospital charts on 18,053 different individuals. In California, 698 different hospitals were involved, and 960 hospitals out-of-state.
Researchers from many studies of this type are pleased when they receive a response rate of 50 percent. The rate of returns from the annual follow-up questionnaires which asked about hospitalization, critical to the entire research process, received response rates in excess of 90 percent, usually above 95 percent. The final and most critical mailing received an incredible 99.5 percent response.
In the decade since the last of the formal data collection, researchers at the Center for Health Research, headquarters for the Adventist Health Study, have been looking at the data from a variety of angles. Following are some of the exciting discoveries which have proven beyond a doubt that certain aspects of the Adventist lifestyle have markedly improved the health of those individuals.