Fundscopically identified risk factors for cerebrovascular disease (CVD) and heart disease (HD) were prospectively studied in 2, 112 men and women aged 30-59 who initially had no history of either disease and who lived in the Nishi-izu district of Shizuoka Prefecture.
Baseline medical examinations of the retinal arteries were made in a sample of 2, 112 residents in Nishi-izu Machi and Kamo Mura in 1964-1966 who were followed up until 1985. During the follow-up period, 93 died from CVD and 64 from HD.
Using a case-control study in a cohort study design, fundoscopically identified risk variables (hypertensive or arteriosclerotic changes in ocular fundus: fundoscopic classification by Scheie) were compared between the 157 cases (93 CVD cases and 64 HD cases) at the last health examination before death and 314 control survivors matched for gender, age (±2 years), and residential district. Using the same design, the progression of risk variables for 5±1 years prior to the last examination was followed to identify factors associated with circulatory disease.
From conditional logistic regression analyses using findings of the retinal arteries at the last health examination, significant risk variables for CVD were found to be the narrowing of retinal arteries and increases in reflex, whereas risk variables for HD were arteriovenous crossing (concealment) and increase in reflex.
From the same multivariate analyses using the progression of findings, significant risk variables for CVD were the ingravescence of the narrowing of retinal arteries and progressive increases in reflex, whereas for HD the only risk variables was progressive increases in reflex.
The present study suggested that, in addition to the cross-sectional findings of retinal arteries on a given occasion, the progression of findings through serial health examinations yields useful information for controlling the health of residents.