Get Newsletter
AlzRisk Paper Detail
Back

Reference: Hayden, 2006
Cohort: Cache County Study
Risk Factor: Blood Pressure


Average Follow-up Time Detail
"Briefly, all residents of the county aged 65 or older as of January 1, 1995 were invited to participate.... A second evaluation of the cohort took place 3 years later, from 1998 to 1999.... Participants' cognitive status was assessed via a multistage screening procedure at follow-up, approximately 3.2 years (range= 2 to 5y, standard deviation = 0.44) after the baseline assessment."

Exposure Detail
History of hypertension was assessed at baseline, in 1995, when all participants were at least 65 years old.

"Individuals or their informants reported medical histories of CVRFs [cerebrovascular risk factors] at the baseline and follow-up interviews. Participants were asked about a number of factors and history of events including hypertension …. Interviewers recorded a positive history of each condition if the participant, or proxy informant, indicated he/she was ever told about the condition by a doctor or nurse and he/she received treatment for it. If the participant did not specify a doctor's diagnosis or treatment, a negative history was recorded. A negative history was also recorded if the participant reported no history of the condition or event. In the event that the status was unknown, the item was recorded as missing. Information on participants' current use of antihypertensives … gained from a medicine check inventory taken at the time of the interview was also considered. For example, if a participant reported no history of hypertension but the medicine chest inventory revealed a current prescription for antihypertensives, the participant was included in the hypertensive group. Information from proxy informants was used in cases where cognitive difficulties interfered with accurate reporting."

Ethnicity Detail
"[T]his population is primarily white and is genetically similar to other US populations of Northern European decent."

Age Detail
All participants were 65 years or older at study entry.

Screening and Diagnosis Detail
Screening Method:
"Modified" 3MSE"Modified" Modified Mini-Mental State Examination (Tschantz 2002)
Age
DQDementia Questionnaire (Silverman 1986)
IQ-CODEInformant Questionnaire for Cognitive Decline in the Elderly (Jorm 1989)

AD Diagnosis:
NINCDS ADRDA National Institute of Neurological and Communicative Diseases and Stroke/Alzheimer's Disease and Related Disorders Association Criteria (McKhann 1984)

Total dementia definition: Dementia was diagnosed via DSM-III-R criteria using all available information. Severity was additionally assessed using the Clinical Dementia Rating Scale.

"A modified version of the Modified Mini-Mental State Exam (3MS)23 was administered either at the participants home or in the nursing home. The Informant Questionnaire for Cognitive Decline in the Elderly (IQ-CODE)24 was administered to a proxy informant, that is, spouse or next of kin, in the event that participants were unable to complete the 3MS. Individuals who scored beyond predetermined cut-points were referred for further assessment with the dementia questionnaire (DQ),25 which was administered by phone to a proxy informant. At the follow-up evaluation, all those scoring below 87 on the 3MS, our predetermined cut-point,21 and those over age 80 scoring below 84 on the 3MS, and all those scoring above the recommended cut-point of 3.2 on the IQCODE were referred for further assessment with the DQ. A subsample of participants, including all participants aged 90 or older, were also referred for DQ. If participants scored > 4 on the DQ at baseline20 or >+3 at follow-up21 or were part of the designated subsample, they were referred for a full clinical assessment.

A psychometrician and a research nurse conducted assessments at the participants' home or nursing home, in the presence of an informant. The nurse recorded the participant's history of cognitive symptoms if any, medical history, current medications used, and blood pressure, and administered a standardized neurological examination and the Neuropsychiatric Inventory. 26 The Dementia Severity Rating Scale 27 was used to assess cognitive and functional impairment. A family history and a 7-minute videotape of the participants were taken, and a battery of neuropsychologic tests was administered.28 If dementia was suspected, further information was gathered from participants via laboratory tests, that is, blood count, routine chemistries, serum B12, folate, thyroid function test, urinalysis, and standardized magnetic resonance imaging, or in a few cases via computed tomography. Geriatric psychiatrists, blind to the participant's working diagnosis, examined individuals with possible dementia at their place of residence. Clinical history and Neuropsychiatric Inventory results were reviewed with collateral informants. The assignment of initial diagnoses occurred at consensus conferences attended by the staff who conducted the evaluations and a geriatric psychiatrist (D.C.S, J.C.S.B., or M.S.). Final best estimate diagnoses were assigned by consensus at conferences with geriatric psychiatrists, a board-certified neurologist, a senior neuropsychologist, and a cognitive neuroscientist in attendance."

"[This] study focused on the 3123 dementia-free individuals, 104 AD cases, and 37 cases with any VaD (n=3264). Individuals with primary or secondary diagnoses of other dementias such as Lewy Body Dementia, Parkinson disease, or other dementias were set aside for this evaluation (n=44). Therefore, our AD category covers cases of probable or possible AD only with no secondary diagnoses. Individuals diagnosed with dementia believed to be purely cerebrovascular in nature were categorized as VaD cases. Twelve individuals were diagnosed with mixed dementia (AD and VaD). In the interest of identifying risk factors for AD without the influence of VaD, we placed these individuals in the VaD category. We compared analyses using an all AD category versus a "clean" AD category and there were few meaningful differences."

Covariates & Analysis Detail
Analysis Type:
Discrete-time survival modeling

Stratification by gender yielded similar results.

“Discrete-time survival models 35 were used to estimate the relationship between each CVRF and dementia after adjustment for potential confounders. This method considers time in discrete intervals, in this case years, which is appropriate, given the exact time of onset of dementia and CVRFs are not known. Using this method, each year was considered a discrete time interval. Thus, each subject contributed to the analysis one person-year of observation for each year from the age at which the subject entered the study until either his/her onset of dementia or end of the follow-up period.

HRs for dementia subtypes AD and VaD, stratified by sex, were then estimated in models that included multiple covariates. CVRF variables, the main variables of interest, were operationalized with time invariant indicator variables. Thus, individuals with prevalent vascular diseases or risk factors were recorded as such for every person-year of observation. Incident CVRFs were not considered owing to the possibility of conflicts in the temporality of onset of risk factors and dementia outcomes. All models were controlled for current age within each person-year of observation, sex, education as a continuous variable for number of years of schooling, and number of APOE Epsilon4 alleles (0,1,or 2 alleles)"

AD Covariates:
Aage
Eeducation
Ggender
APOE4APOE e4 genotype
CABGcoronary artery bypass graft
DMdiabetes mellitus
HChigh cholesterol
MImycardial infarction history
OBoverweight/obesity
SHstroke history

TD Covariates:
Aage
Eeducation
Ggender
APOE4APOE e4 genotype
CABGcoronary artery bypass graft
DMdiabetes mellitus
HChigh cholesterol
MImycardial infarction history
OBoverweight/obesity
SHstroke history