Get Newsletter
AlzRisk Paper Detail

Reference: Yoshitake, 1995
Cohort: Hisayama Study
Risk Factor: Physical Activity

Average Follow-up Time Detail
A cohort of individuals free from dementia at study start was followed for 7 years (1985 to 1992).

"In addition, less than 2% of the original cohort of the study have been lost to follow-up form 1961 to 1993."

Exposure Detail
"We defined the physically active group as those including daily exercise during the leisure period or moderate to severe physical activity at work."

Ethnicity Detail
"Since 1961, we have carried out a prospective cohort study in a Japanese subrural community, Hisayama Town, which is adjacent to the metropolitan area of Fukuoka on Kyushu Island, Japan. This study has investigated the epidemiology of cerebrovascular disease in the general Japanese population.16,17 The population of the town has distributions of age, occupational status, and nutrient intake that are almost identical with those for the whole of Japan.17,18"

Age Detail
All residents were 65 years or older as of 1985.

Screening and Diagnosis Detail
Screening Method:
HDSHasegawa's dementia scale

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

"We carried out a two-stage survey of dementia among Hisayama residents 65 years or older in 1985.19 In the first stage we screened 887 subjects (94.6% of the elderly population) by Hasegawa's dementia scale (HDS),20 a neuropyschological test widely utilized in Japan, and an interview based on comprehensive questionnaires concerning psychological and medical symptoms, chronic conditions, treatment, and activities of daily living (ADL). When we suspected dementia, the subject underwent the second stage of the survey, which consisted of physical and neurologic examinations, an interview with the family or attending physician, and review of the clinical records. By the criteria of the Diagnostic and Statistical Manual of Mental Disorders, third edition (DSM-III)21, we diagnosed 59 subjects as demented. The remaining 828 subjects (334 men and 494 women; 88.3% of the elderly population) constituted a nondemented cohort and were followed prospectively from 1985 to 1992.

"During the follow-up period, 214 subjects died, 176 of whom (82%) underwent brain examinations at autopsy. Autopsies were performed at the Department of Pathology of Kyushu University. We reviewed all the available clinical information and interviewed the attending physicians and the families of the deceased subjects.

"From April to November 1992, we did a follow-up screening survey of all 614 survivors. The examinations included administration of a battery of neuropsychological tests and an interview in which the standard questionnaires of the first screening were used. This survey was part of the regular health check included in the Hisayama study. We surveyed those subjects who did not show up for examination at their homes or in hospitals or other institutions. In this survey, 577 subjects (94.0% of the survivors) underwent neuropsychological tests that consisted of the Mini_mental State Examination (MMSE),22 the HDS, and the HDS revised version (HDS-R).23 The HDS consists of 11 questions that test orientation, memory function, common knowledge, and calculation, the possible scores ranging from 0 to 32.5 points. In the HDS-R, five of the original HDS questions have been altered and condensed into three new questions, the maximum score thereby being changed to 30 points. When one of these test scores were below the cutoff point (21 for MMSE, 22.0 for the HDS, and 21 for the HDS-R), dementia was suspected. We also contacted by telephone and completed and interview with 35 subjects (or their families) who had refused the neuropsychological tests or who had moved out of town. We lost only two subjects to the follow-up. Subjects with suspected dementia underwent the additional examinations described above. Zung's self-rating scale24 was used to exclude depression.

"The diagnosis of dementia was made clinically based on the guidelines of the DSM-III-R.25 The criteria used for the diagnosis of definite, probable, or possible AD were those of the NINCDS-ADRDA.

"Because it is difficult to determine the onset of AD and VD when there is no antecedent stroke episode, the tentative time of onset, when the family or attending physician first noticed abnormal behavior by the subject, was used."

Covariates & Analysis Detail
Analysis Type:
Cox proportional hazards regression

"To determine the risk factors for dementia, we collected the following data from the initial survey: age, sex, educational level, marital status (single or not), physical activity (four categories each for leisure and for work), left-handedness, alcohol consumption (yes or no), smoking (yes or no), ADL (six categories), HDS score, and history of diabetes and stroke (yes or no). We also obtained the following measurements from the regular examinations included in the Hisayama study for the years 1983-1985: average of three systolic and diastolic blood pressures (mmHg); body mass index (kg/m2); subscapular/triceps skinfold thickness ratio; proteinuria (presence or not); abnormal ECG findings, consisting of left ventricular hypertrophy (Minnesota code 3-1), ST depression (4-1,2,3), or atrial fibrillation (8-3); white blood cell count (cells/ul); hemoglobin (g/dl); hematocrit (%); serum total protein and albumin (g/dl); gamma-glutamyl transpeptidase (mU/ml); glycosylated hemoglobin A1 (%); serum total cholesterol (mg/dl); triglycerides (mg/dl); high-density lipoprotein cholesterol (mg/dl); low-density lipoprotein cholesterol (mg/dl). We defined the physically active group as those including daily exercise during the leisure period or moderate to severe physical activity at work. A low educational level constituted completion of only 6 years of elementary school. We divided the ADL into six categories and defined poor ADL by the last four categories (ie, from needing assistance when going out of the home to being bedridden).
"We estimated the age-adjusted and multivariate risks of each potential risk factor by using the beta-coefficients from stepwise Cox's proportional hazards analysis27...
"Age and gender were included in all multivariate analyses.
"In order to exclude risk factors that were significant by chance after age adjustment and to determine the independent risk factors for dementia, we made a multivariate analysis by the significant risk factors available in table 4 and sex (table 5).
"In contrast, age and a low HDS score were independent risk factors for Ad, and regular moderate physical activity was a significant preventive factor."

AD Covariates: