- Click here for explanation of classification scheme.

To examine whether declining selenium levels over time are associated with cognitive decline in a cohort of community-dwelling French elderly.

Participants in waves two and six of the Etude de Vieillissement Artériel (EVA) who completed a general questionnaire, cognitive evaluation and a blood sample.

Participants in waves two and six of the EVA who did not provide a general questionnaire, cognitive evaluation and a blood sample.

Recruitment

Men and women born between 1922 and 1932 and residing in the town of Nantes, Western France were recruited from electoral rolls and to a lesser extent through informational campaigns.  EVA is a nine-year longitudinal study with six waves of follow-up.

Design

Prospective cohort study. 

Statistical Analysis

• The characteristics of subjects at inclusion were described in two groups: 
  • Those who completed the follow-up
  • Those who did not complete the follow-up.
• To test the differences between these two groups, x² test and the Student T-test were used. Percentile distribution and means with standard deviations (SDs) were described for both cognitive and selenium change variables. Classic linear regressions were used to assess the association between nine-year selenium change and cognitive change, the association between two-year selenium  change and cognitive  changes during the two, four, six and nine years of follow-up.
• To simultaneously take into account cognitive changes at each wave of the study and the within-subject correlation of measurements, mixed models were used to analyze associations between cognitive and selenium changes. Cognitive decline was dichotomized by using two cutoffs. The cognitive score difference between that wave and baseline was calculated. Cutoff points corresponded to the 25th and 10th percentile of the distributions of the mean of the difference. To analyze these dichotomous cognitive variables, mixed logistic model with Gaussian random effect was used.
• Selenium changes were analyzed as continuous variables. Analyses were first adjusted for time and selenium level at baseline. Second, analyses were adjusted for the other potential confounding factors associated with cognition or selenium level including sociodemographic factors, consumption habits and health factors. Results of mixed linear models were expressed by linear regression coefficient (β) with their 95% CI. Results of mixed logistic models were expressed by odds ratio (OR) with their 95% CI.

Timing of Measurements

1991-1993, June 2000 and December 2001.

Dependent Variables

Cognitive evaluation based on:

• Mini-Mental Status Examination (MMSE)
• Trail Making Test part B (TMTB)
• Digit Symbol Substitution (DSS)
• Finger Tapping Test (FTT).

Independent Variables

Selenium levels at EVA waves zero, two and six.

Control Variables

• Sex
• High school
• Smoking status
• Alcohol consumption
• Diabetes
• Hypertension
• Cardiovascular disease
• Dyslipidemia
• Age
• Body mass index (BMI)
• Baseline selenium level
• Two-year selenium decrease. 

• Initial N: 1,389
• Attrition (final N): 702
• Age: 65±3 years
• Ethnicity: French.

Other Relevant Demographics

Baseline Characteristics of Subjects According to Nine-Year Follow-up Status
	Yes	No
Sex (women)	62.1	55.2
High school	50.8	46.7
Smoking status:
Current	8.3	9.0
Former	30.5	35.5
Alcohol consumption	28.8	30.1
Diabetes	4.9	6.0
Hypertension	46.5	52.8
Cardiovascular diseases	9.7	12.7
Dyslipidemia	47.1	45.3
MMSE score 25th percentile or less	17.7	25.1
DSS score 25th percentile or less	20.3	27.5
FTT score 25th percentile or less	22.7	26.7
TMTB score 25th percentile or less	20.8	30.2
BMI	24.8	25.6
Two-year selenium decrease (μmol/L), mean	0.049	0.063

Anthropometrics

Baseline mean plasma selenium level (±SD) was 1.09μmol per L (±0.2μmol per L).

Location

Nantes, Western France. 

Key Findings

Classic Linear Regression Analysis

• A decrease of plasma selenium was found over the entire follow-up period. Means of selenium declined -0.055 (SD±.2) μmol per L at two years and -0.096 (SD±21) μmol per L at the nine-year follow-up
• Two-year selenium change was not associated with cognitive change during two-, four-, six- or nine-year follow-up for any of the cognitive tests
• Nine-year selenium change was associated with nine-year cognitive change for MMSE (β=0.38; 95% CI=0.14 to 062), but not for the other cognitive tests.

Mixed Models: Factors Associated with Cognitive Change and Selenium Change

• Cognitive change was associated with time of follow-up (β=0.09; 95% CI=0.07 to 0.11) but not with age at inclusion
• Diabetes and hypertension were modestly associated with a higher decrease of cognitive performance [β=0.31 (0.004 to 0.62) and β=0.14 (-0.02 to 0.3)], respectively.
• During the follow-up, occurrence of cardiovascular events, as well as obesity, were associated with greater declines in plasma selenium.

Mixed Models: Association Between Two-year Selenium Change and Nine-year Cognitive Change

• Using crude mixed linear models cognitive change (two-, four-, six- and nine-year) by the first two- year selenium change, there was no association with any of the four cognitive tests
• After controlling for time and plasma selenium level at baseline, the short-term plasma selenium decrease was only weekly associated with cognitive decline. These results were not changed after controlling for potential confounding factors.

Association Between Selenium Change and Cognitive Change During the Nine Years of Follow-up

• Cognitive and selenium variables were considered simultaneously measured at inclusion, two years and nine years
• Cognitive performances (as continuous variables) were related to selenium levels and time of follow-up. Interaction terms between selenium and time of follow-up expressed the change of selenium according to cognitive change during follow-up.
• Selenium changes were associated with DSS and FTT, but not with TMTB or MMSE. Analyses adjusted for time, sex, education, diabetes, hypertension, dyslipidemia and history of cardiovascular diseases gave similar results.
• In sensitivity analysis, the same models were applied to subjects who had all three selenium measurements and cognitive evaluation. Results were quite similar although with lower power because analysis were carried out on 570 subjects instead of the initial 1,371:

	Crude Association		Adjusted Association		Sensitivity Analysis
	β	95% CI	β	95% CI	β	95% CI
Decrease in MMSE score	-0.008	(-0.12 to 0.12)	0.0037	(-0.12 to 0.13)	.041	(-0.09 to 0.17)
Decrease in DSS score	0.476	(0.12 to 0.83)	0.38	(0.03 to 0.73)	.34	(0.03 to 0.71)
Delay in TMTB time	-0.4	(-2.87 to 2.07)	-0.197	(-2.65 to 2.25)	-.24	(-2.98 to 2.50)
Decrease in FTT score	1.25	(0.35 to 2.15)	1.18	(0.28 to 2.08)	.83	(0.17 to 1.83)

• After studying associations between selenium change and cognitive decline using a mixed logistic model with Gaussian random effect the following was found after adjusting for confounding factors:
  • Among subjects who had an decrease of their plasma selenium levels, the higher the decrease of plasma selenium, the higher the probability of cognitive decline
  • Among subjects who had an increase of their plasma selenium levels, the probability of cognitive decline was higher in subjects with the smallest selenium increase 
  • Plasma selenium change was associated with MMSE decline at the two-point cutoff (OR=2.31; CI: 1.12 to 4.77) but not at the three-point cutoff (1.41; 0.52 to 3.83)
  • An association was observed for  the DSS at the 25th cutoff (2.6; 1.22 to 5.57) and for TMTB at the 10th percentile (3.18; 1.14 to 8.88)
  • For FTT, decline increased when selenium decreased for the two cutoffs considered [the 10th percentile, 2.4 (1.21 to 4.77) and for the 25th percentile, 4.33 (1.6 to 11.72)].   

The results of the study suggest that plasma selenium decease is associated with cognitive decline. The real importance of selenium in the brain and the capacity of the brain to manage selenium depletion is just beginning to be explored. These results, together with information on involvement of selenoproteins in brain functions, support possible relationships between selenium status and neuropsychologic functions in aging people. The preventative effects of selenium supplementation at a nutritional level needs to be evaluated with large-scale studies. This approach could shed new light on the potential benefits of supplementation.

Government:	Inserm
Industry:	Merck, Sharp and Dohme-Chibret Laboratories, EISAI Laboratory Pharmaceutical/Dietary Supplement Company:
Not-for-profit	French Alzheimer's Disease Association Foundation associated with industry:

• The authors note the following limitations:
  • The main limit was sample selection during follow-up. Subjects who did not complete the whole follow-up had the lowest cognitive performances at inclusion; however, they did not differ with regard to selenium level at baseline or plasma selenium decrease for the first two years.
  • A relationship between plasma selenium and cognition change was found in subjects with weak cognitive decline. Sample selection may limit the interpretation of the results, but should not affect the relationship itself
• Specific inclusion and exclusion criteria for the EVA was not discussed in this study. It was unclear if this was a representative sample of the population. Specific reasons for withdrawals from EVA were not fully discussed. Blinding was not used in the study.