Table 3.
Year-adjusted relative expected cognitive scores, by baseline characteristics, over time.
| CVLT-II/HVLT | WAIS-II | BDI | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Total Recall | Short Delay | Digit Symbol | ||||||||
| Predictor | Coef. (95% CI) | p | Coef. (95% CI) | p | Coef. (95% CI) | p | Coef. (95% CI) | p | Coef. (95% CI) | p |
| Razor clam consumption | ||||||||||
| None/Low (0–14 clams/mo.) | 0.03 | 0.04 | 0.28 | 0.13 | 0.45 | |||||
| High (15+ clams/mo.) | −1.0 (−2.0, −0.1) | −1.1 (−2.1, −0.0) | −0.6 (−1.6, +0.5) | −0.6 (−1.3, +0.2) | −0.3 (−1.2, +0.5) | |||||
| Gender | ||||||||||
| Male | +0.1 (−1.6, +1.8) | +3.1 (+1.8, +4.5) | ||||||||
| Female | 0.94 | −1.6 (−3.1, −0.1) | 0.04 | −0.4 (−2.1, +1.2) | 0.59 | +6.1 (+4.6, +7.6) | <0.01 | <0.01 | ||
| Marital status | ||||||||||
| Not married | +1.2 (−0.4, +2.8) | +2.0 (+0.2, +3.7) | +3.7 (+2.0, +5.3) | −0.3 (−1.8, +1.2) | ||||||
| Currently married | +1.8 (−0.0, +3.7) | 0.05 | 0.15 | 0.03 | <0.01 | 0.68 | ||||
| Education level | ||||||||||
| High school or less | +5.5 (+3.9, +7.2) | +5.8 (+4.3, +7.3) | ||||||||
| At least some college | <0.01 | +3.2 (+1.7, +4.7) | <0.01 | +4.9 (+3.3, +6.5) | <0.01 | <0.01 | −1.9 (−3.3, −0.5) | 0.01 | ||
| Employment | ||||||||||
| Not working | +1.9 (−1.2, +4.9) | +2.1 (−0.6, +4.8) | +1.6 (−1.3, +4.6) | +0.3 (−2.4, +3.0) | −3.8 (−6.3, −1.3) | |||||
| Working, Manual labor | +6.3 (+4.4, +8.2) | 0.22 | +4.0 (+2.3, +5.7) | 0.13 | +5.2 (+3.3, +7.0) | 0.27 | +6.6 (+4.9, +8.3) | 0.83 | −2.2 (−3.8, −0.7) | <0.01 |
| Working, Office work | +4.0 (+1.7, +6.4) | <0.01 | +2.3 (+0.2, +4.4) | <0.01 | +2.2 (−0.0, +4.4) | <0.01 | +2.5 (+0.4, +4.6) | <0.01 | −1.1 (−3.1, +0.8) | 0.01 |
| Working, unspecified | <0.01 | 0.03 | 0.05 | 0.02 | 0.25 | |||||
CVLT: California Verbal Learning Test. HVLT: Hopkins Verbal Learning Test. WAIS: Wechsler Adult Intelligence Scale. BDI: Beck Depression Inventory. CI: confidence interval. T-scores were used for all outcomes except for BDI, which used raw scores. Confidence intervals and p-values from generalized estimating equations (GEE), adjusted for year of study. A separate GEE model was built, adjusted for year of study, for each combination of independent variable and outcome (e.g., this table includes seven predictor variables and five outcome variables, meaning that 35 total GEE models are represented, each of which has two independent variables: the predictor variable listed, plus year of study). Negative value indicates a group that had a lower expected score than the reference group.