Get Newsletter
AlzRisk Paper Detail

Reference: Brenner, 1994
Cohort: Alzheimer Disease Patient Registry - Group Health Cooperative
Risk Factor: Hormone Therapy

Average Follow-up Time Detail
Eligible cases were enrolled in the Alzheimer's Disease Patient Registry between April 1987 and February 1992. The control subjects were selected from a stratified random sample of the same Group Health Cooperative of Puget Sound population, and matched to cases by age (within 2 years).

These are the results for those who took all estrogens, which includes both oral and vaginal estrogens. Therefore, the number of subjects is different from the rows for oral or vaginal estrogens.

Exposure Detail
HRT exposure was determined using computerized pharmacy records from the Group Health Cooperative of Puget Sound. Pharmacies in the health plan have records of filled prescriptions dating back to January 1977.

"A filled estrogen prescription was considered as exposure up to the reference year for both cases and controls. For cases, it was defined as the year before reported onset of disease. For the controls, it was specified as the year of study intake minus 3."

"The distribution of the different estrogen formulations is the same for both the cases and controls. Oral conjugated estrogens accounted for two thirds and dienestrol vaginal cream accounted for 19 percent of all estrogen prescriptions."

Results were reported for three separate categories in this paper:

1) Any v. no HRT use
2) Duration of HRT use (defined by number of prescriptions filled) v. no HRT use
3) Timing of HRT use (current, former) v. no HRT use

For each of these categories, results were provided for route of administration (all estrogens, oral, and vaginal).

Results reported here compared users of estrogen administered by any route (both oral and vaginal) with never users, further distinguishing HRT use according to duration of use (in tertiles of total number of prescriptions).

Age Detail
The age provided is a calculated weighted average from the 107 cases and 120 age- and sex- matched controls. Age corresponds to age at the index year.

Screening and Diagnosis Detail
Screening Method:

AD Diagnosis:
DSM IIIR Diagnostic and Statistical Manual III-Revised
NINCDS ADRDA National Institute of Neurological and Communicative Diseases and Stroke/Alzheimer's Disease and Related Disorders Association Criteria (McKhann 1984)

Screening: "The University of Washington has established an Alzheimer's Disease Patient Registry in a defined part of this health plan (n = 23,000 persons aged more than 60 years) (9, 10). Briefly, this Registry has implemented a surveillance system that identifies individuals with symptoms of suspected dementia by different sources (logs of neuroradiologic examinations, discharge records, clinic registration lists, and referrals of primary care physicians)."

AD Diagnosis: "A standardized examination including medical history, physical and neurologic examinations, 2-3 hours of neuropsychologic testing, laboratory tests, and imaging techniques is the basis for a standardized diagnosis by an interdisciplinary study team."

Covariates & Analysis Detail
Analysis Type:
Logistic regression

AD Covariates:
HYShistory of hysterectomy

Adjustment for the following variables reportedly did not change the results, and they were NOT included in the models: education (more than high school v. high school or less), marital status, ethnicity (Caucasian v. other), and history of either smoking or progestogen use.

The final models included the matching variable age (in years) and two indicator variables for history of hysterectomy (before and after age 55).