Table 1.

The association between retinal venular caliber and IQ at age 38 years after excluding cases for potential confounders (poor health factors, lifestyle factors or environmental factors). The standardized regression coefficients (β) reflect change in neuropsychological test performance as a function of a 1-SD unit change in retinal venular caliber.

	IQ at Age 38 Years Model 1			IQ at Age 38 Years Model 2
	β	95% CI	p	β	95% CI	p
Total Cohort (n=916)*	−.155	−.219, −.091	< .001	−.194	−.282, −.108	< .001
Excluding cases of (n= total number of study members after exclusion):
High hsCRP (n=708)	−.161	−.234, −.088	< .001	−.233	−.334, −.131	< .001
Pre-diabtetes or Diabetes (n=718)	−.174	−.238, −.099	< .001	−.221	−.309, −.118	< .001
High blood pressure (n=800)	−.155	−.222, −.086	< .001	−.178	−.272, −.083	< .001
Obesity (n=699)	−.186	−.267, −.117	< .001	−.204	−.313, −.107	< .001
Smokers (n=709)	−.124	−.198, −.051	.001	−.192	−.292, −.094	< .001
Low Adult SES (n=638)	−.175	−.249, −.097	< .001	−.210	−.309, −.107	< .001

Note: Model 1: control for sex. Model 2: Model 1 + control for arteriolar caliber.

^*Out of the n=922 with retinal imaging information, 6 Study members did not have neuropsychological tests scores leaving an effective group size of n=916.

hsCRP = high sensitive C reactive protein; SES = Socio-economic status.