Abstract
The authors examined long-term antioxidant intake in relation to cognitive decline among older women. Beginning in 1980, Nurses’ Health Study (NHS) participants completed dietary assessments every 4 years; in 1995–2001, 16,010 participants aged ≥70 years completed initial cognitive assessments, which were repeated 3 times at 2-year intervals. Long-term antioxidant intake was averaged from 1980 through the time of initial cognitive interviews. Multivariable-adjusted linear regression was used to estimate mean differences in rates of cognitive decline across categories of vitamin E, vitamin C, and carotenoid intake; statistical tests were 2-sided. No associations were evident for vitamin E or total carotenoid intake and cognitive decline (e.g., after multivariable adjustment, P-trend = 0.44 and P-trend = 0.51, respectively, for a global composite score averaging all 6 cognitive tests), although higher lycopene intake and lower vitamin C intake were related to slower cognitive decline. In alternative analyses of overall cognitive status at older ages (averaging all 4 cognitive assessments), results for vitamins E and C were generally null, but higher carotenoid intake was related to better cognition. Overall, long-term vitamin E and C intakes were not consistently related to cognition, although greater consumption of carotenoids may have cognitive benefits in older adults.
Keywords: antioxidants, cognition, cohort studies, diet
Experimental data indicate that oxidative stress plays an important role in the development of cognitive impairment and Alzheimer's disease, but epidemiologic studies have produced conflicting results on the association between antioxidant vitamins and cognition in older adults (1–3). In particular, several large clinical trials of vitamin E, vitamin C, and beta-carotene supplements have shown little effect on cognitive function, with follow-up periods of 3–7 years (4–8). However, 1 trial suggested that long-term supplementation with beta-carotene, over approximately 18 years, can confer cognitive benefits, and an additional analysis by our research group indicated that long-term exposure to vitamin E supplements might be associated with cognitive health in the Nurses’ Health Study (NHS) (9, 10). These findings are consistent with the prevailing notion that the pathology underlying cognitive impairment begins decades prior to onset of detectable symptoms; thus, long-term antioxidant intake could be most relevant for cognitive outcomes in later life (11). We took advantage of multiple, repeated dietary assessments administered over 2 decades in the NHS to examine relations between long-term antioxidant intake and subsequent cognition in approximately 16,000 older women.
MATERIALS AND METHODS
The NHS began in 1976, when 121,700 female registered nurses aged 30–55 years responded to a mailed questionnaire on health and lifestyle. Biennial questionnaires are used to update this information, and current follow-up exceeds 90%. In 1980, we added a food frequency questionnaire, which is administered approximately every 4 years. In addition, we initiated a telephone-based study of cognitive function in 1995–2001 and invited women to participate if they were ≥70 years of age and had no history of stroke. Ninety-two percent of eligible women participated in the initial cognitive interview, and 3 subsequent follow-up interviews were conducted at approximately 2-year intervals (the median time between interviews 1 and 4 was 6.4 years). Follow-up was ≥90% among women who were still alive at each interview. The institutional review board of Brigham and Women's Hospital (Boston, Massachusetts) approved this study.
Population for analysis
Of the 19,415 nurse-participants who completed the initial cognitive interview, 3,405 were excluded from our analysis because they lacked dietary data; thus, we included 16,010 women in the current analysis. Key characteristics at the time of initial cognitive interview were similar among women who were included in our analysis and those who were excluded (e.g., mean age = 74.2 years vs. 74.4 years; mean body mass index (weight (kg)/height (m)2) = 26.0 vs. 26.3).
Dietary assessment
We used a 61-item, Willett semiquantitative food frequency questionnaire to ascertain dietary habits in 1980 and an expanded 130-item version in 1984, 1986, and every 4 years thereafter (12). Food items were specified in common units or portion sizes (e.g., 1 tomato or a small, 1-ounce (28-g) packet of nuts) and detailed information was collected on methods of food preparation (e.g., types of cooking oils). Participants reported how often, on average, they had consumed each food item over the previous year, using 9 prespecified response categories that ranged from “almost never” to “ ≥6 times per day”. For each food, we multiplied the reported consumption frequency by its nutrient content, which was determined using the Harvard University Food Composition Database (updated regularly). Each participant's nutrient intakes were aggregated across all dietary sources and adjusted for total energy intake using the residual method (13).
In a validation study, we found high correlations for major foods contributing to antioxidant intakes, comparing the food frequency questionnaire with four 1-week dietary records collected over a period of 1 year (14). In particular, correlation coefficients (corrected for within-person variation in dietary records) were 0.67 for chicken, 0.79 for cold cereal, and 0.75 for peanut butter (major sources of vitamin E in this cohort); 0.74 for oranges and 0.84 for orange juice (major sources of vitamin C); and 0.73 for tomatoes and 0.55 for carrots (major sources of carotenoids).
Cognitive function measurement
Initially, we administered the Telephone Interview for Cognitive Status (TICS), a telephone adaptation of the Mini-Mental State Examination; the two tests are highly correlated (ρ = 0.94) (15). After establishing high participation rates, we gradually added 5 other cognitive tests: the East Boston Memory Test—immediate and delayed recalls (16), category fluency (17, 18), delayed recall of the TICS 10-word list, and digit span backward (19). Despite the gradual introduction of cognitive tests, 87% of women completed all 6 tests in our cognitive battery during the initial cognitive interview. We found that our telephone-based cognitive battery was highly valid compared with detailed, in-person interviews (ρ = 0.81 for scores using the two methods in a validation study among 61 women). Inter-interviewer reliability was high across 10 interviewers (ρ > 0.95 for each cognitive test), and participation rates were virtually identical across all cognitive tests and remained stable over time.
We evaluated 3 primary outcomes selected a priori: 2 measures of general cognition (a global composite score and the TICS score) and a verbal composite score, as verbal memory is a strong predictor of the development of Alzheimer's disease (20–22). Because our 6 cognitive tests are scaled differently, we used z scores to create the 2 composite measures; specifically, we calculated the difference between each participant's individual score and the overall mean score for our study population and divided by the standard deviation of the study population. To create the global score, we averaged z scores for all 6 cognitive tests; we created the verbal score by averaging the immediate and delayed recalls of both the East Boston Memory Test and the TICS 10-word list. Composite scores were constructed only for women who completed all contributing tests.
Statistical analysis
Because of the long period of pathogenesis underlying cognitive decline, long-term dietary habits are likely to be most relevant; thus, we averaged antioxidant intakes from 1980 through each woman's last dietary report prior to her initial cognitive interview, to obtain the most stable measures reflecting long-term diet (11). Women had an average of 5 dietary assessments over a mean of 17 years. We evaluated total intakes of vitamin E, vitamin C, and carotenoids (derived from foods and supplements) and dietary intakes of vitamin E, vitamin C, and carotenoids (derived from foods only) in relation to cognitive decline; we also examined associations of antioxidant intake from supplements with cognitive decline. In addition to total carotenoids, we considered the following individual carotenoids: beta-carotene, alpha-carotene, lutein/zeaxanthin, lycopene, and beta-cryptoxanthin. Of the carotenoids, only beta-carotene was regularly included in vitamin supplements throughout the dietary reporting period; thus, we evaluated beta-carotene in foods and supplements but only dietary intakes of alpha-carotene, lutein/zeaxanthin, lycopene, and beta-cryptoxanthin.
In our analyses, we found nonlinearity in cognitive trajectories due to “learning effects,” primarily between interviews 1 and 2 (i.e., there was an average increase in cognitive scores between assessments 1 and 2, followed by an average decline in scores at interviews 3 and 4); thus, we averaged test scores at cognitive assessments 1 and 2 to obtain a “robust baseline” score (23), yielding linear cognitive trajectories over time. Then, we used multivariable-adjusted mixed linear regression to estimate mean differences in rates of cognitive decline over 3 time points across quintiles of various antioxidant intakes. Specifically, we used repeated-measures models with random intercepts and slopes, which permits description of individual cognitive trajectories over time and provides explicit tests regarding the relation of antioxidant intakes and rates of cognitive decline. We calculated 95% confidence intervals and performed linear tests of trend using the median values of quintiles.
In our analyses, we considered multiple potential confounders: age (years; continuous), education (registered nurse degree, bachelor's degree, or graduate degree), antidepressant use (yes, no), alcohol intake (none, 0–14 g/day, ≥15 g/day, or missing data), smoking (never, past, or current), physical activity (quintiles of metabolic equivalent of task-hours/week, or missing data), total energy intake (quintiles of kcal/day), body mass index (<22, 22–24, 25–30, ≥30, or missing data), history of high blood pressure (yes, no), history of high cholesterol (yes, no), history of type 2 diabetes (yes, no), and history of myocardial infarction (yes, no). When relevant, to consider possible confounding by vitamin supplements, we controlled for vitamin E supplement use (yes, no, or missing data), vitamin C supplement use (yes—regularly, yes—seasonally, no, or missing data), and multivitamin supplement use (yes, no, or missing data), except when doing so would have adjusted for the exposure of interest (e.g., we did not adjust for vitamin E supplement use in analyses of total vitamin E). In addition, we adjusted analyses of total and dietary vitamin E for polyunsaturated fat intake (quintiles) because vitamin E and polyunsaturated fat are found in common food sources. Levels of all covariates were determined at the time of initial cognitive interview, except total energy intake, which was averaged to correspond with the method used to calculate antioxidant intakes.
We conducted additional analyses of cognitive status as well. Specifically, we evaluated an alternative outcome by averaging all 4 of our repeated cognitive assessments to obtain “average” cognition in later life. Given the very long exposure period over which we evaluated antioxidant intake (approximately 15–20 years), it seemed plausible that any impact of antioxidants on brain health may have occurred prior to our initial cognitive assessment; thus, relations might be better reflected by a measure of “average” cognition in later life than by analyses of change in cognition over a relatively short period of 6 years subsequent to the initial cognitive assessment. Thus, the average of all 4 repeated cognitive measurements was considered a representation of overall cognitive status at older ages. Finally, we examined the possibility that smoking—a major contributor to oxidative stress—might modify the association between antioxidant intake and cognition.
All statistical analyses were conducted in SAS, version 9 (SAS Institute Inc., Cary, North Carolina), and all statistical tests were 2-sided.
RESULTS
Key health and lifestyle characteristics of participants at the time of initial cognitive interview are presented in Table 1. Overall, characteristics were generally similar across increasing quintiles of long-term total vitamin E, vitamin C, and carotenoid consumption. However, overall, women had slightly higher physical activity levels and baseline cognitive scores across increasing quintiles of antioxidant intake; these trends were most pronounced for total carotenoids. Trends were similar when we considered dietary intakes of vitamin E, vitamin C, and carotenoids from foods only (data not shown in table).
Table 1.
Characteristics of Participants at the Time of the Initial Cognitive Interview, According to Selected Quintiles of Total Vitamin E, Vitamin C, and Carotenoid Intakea, Nurses' Health Study, 1995–2001
| Total Vitamin E |
Total Vitamin C |
Total Carotenoids |
|||||||
|---|---|---|---|---|---|---|---|---|---|
| Q1 | Q3 | Q5 | Q1 | Q3 | Q5 | Q1 | Q3 | Q5 | |
| Median intakeb | 7 | 31 | 159 | 124 | 244 | 761 | 8,631 | 13,841 | 21,310 |
| Mean age, years | 74.2 | 74.2 | 74.3 | 74.1 | 74.2 | 74.3 | 74.3 | 74.2 | 74.2 |
| Education, % with graduate-level degree | 5 | 6 | 6 | 5 | 7 | 7 | 4 | 6 | 9 |
| Median alcohol intake, g/day | 0 | 0.88 | 0.88 | 0 | 0.88 | 0.88 | 0 | 0.88 | 0.88 |
| Current smoking, % | 11 | 7 | 7 | 12 | 7 | 7 | 11 | 7 | 7 |
| Median physical activity, MET-hours/weekc | 7.7 | 10.2 | 10.2 | 7.5 | 10.3 | 10.4 | 7.4 | 9.5 | 14.2 |
| Obesityd, % | 19 | 19 | 16 | 19 | 18 | 16 | 17 | 17 | 17 |
| Use of antidepressant medication, % | 4 | 6 | 6 | 4 | 5 | 7 | 6 | 6 | 5 |
| History of high blood pressure, % | 56 | 55 | 55 | 54 | 56 | 56 | 54 | 55 | 56 |
| History of high cholesterol, % | 59 | 66 | 68 | 61 | 66 | 66 | 63 | 67 | 65 |
| History of type 2 diabetes, % | 10 | 10 | 9 | 11 | 11 | 9 | 9 | 10 | 10 |
| History of myocardial infarction, % | 6 | 6 | 6 | 7 | 5 | 6 | 7 | 6 | 6 |
| Mean total energy intake, kcal/day | 1,612 | 1,736 | 1,613 | 1,649 | 1,713 | 1,617 | 1,685 | 1,680 | 1,653 |
| Mean baseline cognitive scoree | |||||||||
| Global composite | −0.06 | 0.01 | 0.00 | −0.03 | 0.01 | 0.02 | −0.08 | 0.01 | 0.05 |
| Verbal composite | −0.06 | 0.00 | 0.00 | −0.03 | 0.00 | 0.01 | −0.09 | 0.00 | 0.07 |
| TICS | 33.7 | 33.9 | 33.9 | 33.8 | 33.9 | 33.9 | 33.7 | 33.9 | 34.0 |
Abbreviations: MET, metabolic equivalent of task; Q, quintile; TICS, Telephone Interview for Cognitive Status.
a Total intake includes the contributions of foods and supplements.
b Median intakes are expressed in g/day for vitamins E and C and in μg/day for carotenoids.
c One MET-hour is proportional to the amount of energy spent sitting quietly for 1 hour.
d Obesity was defined as body mass index (weight (kg)/height (m)2) ≥30.
e Average of cognitive scores from cognitive interviews 1 and 2.
In models adjusted for multiple potentially confounding factors, total vitamin E intake was not related to cognitive decline (e.g., for the global score, P-trend = 0.44) (Table 2). We initially observed a modest association between greater intake of dietary vitamin E and slower cognitive decline in multivariable-adjusted models (e.g., P-trend = 0.07 for the global score), but further adjustment for polyunsaturated fat intake attenuated this association (e.g., P-trend = 0.20 for the global score). Similarly, no association was found for long-term intake of vitamin E from supplements and cognitive decline (e.g., for the global score in multivariable-adjusted models, P-trend = 0.37; results not shown in table). However, with adjustment for covariates, greater consumption of total vitamin C was related to worse cognitive decline in the global score (P-trend = 0.006), the verbal score (P-trend = 0.04), and TICS score (P-trend = 0.006)—a finding that was inconsistent with our a priori hypothesis. Although dietary consumption of vitamin C was not associated with cognitive decline (e.g., P-trend = 0.99 for the global score), higher intake of supplemental vitamin C was related to worse cognitive decline over the follow-up period (e.g., P-trend = 0.001 for the global score; comparing extreme quintiles of intake, mean difference = − 0.07, 95% confidence interval (CI): −0.13, 0.02) (supplement results not shown in table), which explained the finding for total vitamin C and cognitive decline. In addition, no associations were observed between total carotenoid intake (from supplements and foods) or dietary carotenoid intake (from foods only) and cognitive decline in models adjusted for multiple potentially confounding factors (e.g., P-trend = 0.51 and P-trend = 0.48 for the global score). Supplemental carotenoid intake (from beta-carotene supplements) was unrelated to cognitive decline as well (e.g., P-trend = 0.32 for the global score) (results not shown in table).
Table 2.
Mean Differencea in Rates of Cognitive Decline Over the Follow-Up Period Across Quintiles of Totalb and Dietaryc Intake of Vitamin E, Vitamin C, and Carotenoids, Nurses’ Health Study, 1995–2001
| Antioxidant and Quintile | Median Intaked | Global Composite |
Verbal Composite |
TICS |
|||
|---|---|---|---|---|---|---|---|
| Mean Difference | 95% CI | Mean Difference | 95% CI | Mean Difference | 95% CI | ||
| Total vitamin Ee,f | |||||||
| 1 | 7 | 0.00 | Referent | 0.00 | Referent | 0.00 | Referent |
| 2 | 12 | −0.02 | −0.07, 0.03 | 0.00 | −0.06, 0.06 | 0.08 | −0.13, 0.29 |
| 3 | 31 | −0.01 | −0.06, 0.04 | 0.01 | −0.05, 0.08 | 0.21 | −0.01, 0.42 |
| 4 | 69 | 0.00 | −0.05, 0.05 | 0.02 | −0.04, 0.09 | 0.29 | 0.07, 0.51 |
| 5 | 159 | −0.03 | −0.08, 0.03 | −0.01 | −0.08, 0.05 | 0.06 | −0.17, 0.29 |
| P-trendg | 0.44 | 0.51 | 0.75 | ||||
| Dietary vitamin Ee,h | |||||||
| 1 | 6.2 | 0.00 | Referent | 0.00 | Referent | 0.00 | Referent |
| 2 | 7.1 | 0.02 | −0.02, 0.07 | 0.03 | −0.03, 0.09 | 0.20 | 0.00, 0.40 |
| 3 | 7.8 | 0.04 | −0.01, 0.09 | 0.04 | −0.02, 0.09 | 0.15 | −0.05, 0.35 |
| 4 | 8.5 | 0.03 | −0.02, 0.07 | 0.03 | −0.02, 0.09 | 0.10 | −0.10, 0.30 |
| 5 | 9.9 | 0.04 | −0.01, 0.09 | 0.06 | 0.00, 0.12 | 0.14 | −0.06, 0.34 |
| P-trend | 0.20 | 0.09 | 0.50 | ||||
| Total vitamin Ce,i | |||||||
| 1 | 124 | 0.00 | Referent | 0.00 | Referent | 0.00 | Referent |
| 2 | 178 | 0.00 | −0.05, 0.05 | 0.02 | −0.04, 0.07 | 0.02 | −0.18, 0.21 |
| 3 | 244 | −0.02 | −0.06, 0.03 | −0.01 | −0.07, 0.05 | 0.00 | −0.20, 0.20 |
| 4 | 389 | −0.03 | −0.08, 0.02 | −0.03 | −0.09, 0.03 | −0.03 | −0.23, 0.18 |
| 5 | 761 | −0.06 | −0.11, −0.01 | −0.05 | −0.11, 0.01 | −0.24 | −0.46, −0.03 |
| P-trend | 0.006 | 0.04 | 0.006 | ||||
| Dietary vitamin Ce,j | |||||||
| 1 | 95 | 0.00 | Referent | 0.00 | Referent | 0.00 | Referent |
| 2 | 126 | 0.01 | −0.04, 0.06 | 0.01 | −0.04, 0.07 | 0.08 | −0.11, 0.27 |
| 3 | 149 | −0.03 | −0.07, 0.02 | 0.00 | −0.06, 0.05 | −0.04 | −0.23, 0.16 |
| 4 | 173 | 0.01 | −0.04, 0.06 | 0.01 | −0.04, 0.07 | 0.11 | −0.09, 0.30 |
| 5 | 213 | 0.00 | −0.05, 0.05 | 0.01 | −0.05, 0.07 | 0.08 | −0.12, 0.28 |
| P-trend | 0.99 | 0.76 | 0.40 | ||||
| Total carotenoidse,j | |||||||
| 1 | 8,631 | 0.00 | Referent | 0.00 | Referent | 0.00 | Referent |
| 2 | 11,436 | 0.02 | −0.03, 0.06 | 0.01 | −0.04, 0.07 | 0.12 | −0.07, 0.32 |
| 3 | 13,841 | 0.02 | −0.03, 0.07 | 0.01 | −0.04, 0.07 | 0.10 | −0.09, 0.30 |
| 4 | 16,594 | 0.05 | 0.00, 0.10 | 0.05 | −0.01, 0.11 | 0.18 | −0.01, 0.38 |
| 5 | 21,310 | 0.01 | −0.04, 0.06 | 0.01 | −0.05, 0.07 | 0.09 | −0.11, 0.29 |
| P-trend | 0.51 | 0.44 | 0.36 | ||||
| Dietary carotenoidse,j | |||||||
| 1 | 8,326 | 0.00 | Referent | 0.00 | Referent | 0.00 | Referent |
| 2 | 11,095 | 0.04 | 0.00, 0.09 | 0.04 | −0.02, 0.10 | 0.13 | −0.06, 0.33 |
| 3 | 13,407 | 0.00 | −0.05, 0.05 | 0.00 | −0.06, 0.05 | −0.03 | −0.23, 0.16 |
| 4 | 16,044 | 0.06 | 0.02, 0.11 | 0.07 | 0.01, 0.13 | 0.17 | −0.03, 0.36 |
| 5 | 20,506 | 0.01 | −0.03, 0.06 | 0.02 | −0.03, 0.08 | 0.09 | −0.10, 0.29 |
| P-trend | 0.48 | 0.29 | 0.34 | ||||
Abbreviations: CI, confidence interval; TICS, Telephone Interview for Cognitive Status.
a Mean differences compare the mean rates of cognitive decline over follow-up for women in the quintile of interest with women in quintile 1 (the referent). Therefore, a positive mean difference indicates that women in the quintile of interest have less decline than those in quintile 1; a negative mean difference indicates that women in the quintile of interest have more decline than those in quintile 1.
b Total intake includes the contributions of foods and supplements.
c Dietary intake includes the contribution of foods only.
d Median intakes are expressed in g/day for vitamins E and C and in μg/day for carotenoids.
e Adjusted for age, education, antidepressant use (yes, no), alcohol intake (none, 0–14 g/day, ≥ 15 g/day, or missing data), smoking (never, past, or current), physical activity (quintiles of metabolic equivalent of task-hours/week, or missing data), total energy intake (quintiles of kcal/day), body mass index (<22, 22–24, 25–30, ≥30, or missing data), history of high blood pressure (yes, no), history of high cholesterol (yes, no), history of type 2 diabetes (yes, no), and history of myocardial infarction (yes, no).
f Additionally adjusted for use of vitamin C supplements (yes—regularly, yes—seasonally, no, or missing data), use of multivitamins (yes, no, or missing data), and polyunsaturated fat intake (quintiles).
g Linear tests for trend were performed using the median value of each quintile as a continuous variable in regression models.
h Additionally adjusted for use of vitamin E supplements (yes, no, or missing data), vitamin C supplements, and multivitamins and polyunsaturated fat intake (quintiles).
i Additionally adjusted for use of vitamin E supplements and multivitamins.
j Additionally adjusted for use of vitamin E supplements, vitamin C supplements, and multivitamins.
In analyses of individual carotenoids, we found that higher long-term consumption of lycopene from food sources was related to slower cognitive decline (i.e., a better outcome) in the global score (P-trend = 0.05) and the verbal score (P-trend = 0.009) (results not shown in tables). For example, participants in the highest quintile of lycopene intake had a mean decline in the verbal score that was 0.08 standard units less over follow-up (95% CI: 0.03, 0.14) than participants in the lowest quintile. To help interpret the mean differences observed with greater lycopene intake, we found that 1 year of age was associated with a mean decline of 0.05 standard units on the verbal score; thus, the mean difference that we observed between extreme quintiles of lycopene intake over the follow-up period of 6 years was equivalent to approximately 1–2 years of cognitive aging. Long-term intakes of total beta-carotene, dietary beta-carotene, alpha-carotene, lutein/zeaxanthin, and beta-cryptoxanthin were not related to cognitive decline (e.g., P-trend values for the global score were 0.76, 0.77, 0.61, 0.88, and 0.69, respectively, after adjustment for multiple potential confounders).
In alternate analyses of cognitive function averaged across all 4 time points, greater intake of total vitamin E was not related to cognitive status measured by our 3 primary outcomes (e.g., P-trend = 0.12 for the global score), whereas the relation with dietary vitamin E was significant for these outcomes (e.g., P-trend = 0.0008 for the global score) (Table 3). However, findings for dietary vitamin E were attenuated after polyunsaturated fat intake was included in the models. Because polyunsaturated fat intake is measured with error, there is concern that residual confounding may explain these results. In contrast to our finding above for total vitamin C and cognitive decline, there was no association for total vitamin C intake and average cognitive status after controlling for possible confounding factors (e.g., for the global score, P-trend = 0.51). A positive association between greater dietary vitamin C intake and better cognitive status was observed only for the verbal score (P-trend = 0.009). Furthermore, higher intakes of total and dietary carotenoids were both associated with better cognition at older ages as assessed by the global score (P-trend = 0.0001 and P-trend = 0.0005 for total and dietary intakes of carotenoids, respectively) and the verbal score (P-trend < 0.0001 for both). When we explored associations between individual carotenoids and overall cognition, we found that greater intakes of lutein/zeaxanthin and lycopene were related to better cognitive status (e.g., in multivariable-adjusted models, P-trend = 0.0002 and P-trend < 0.0001, respectively, for the global score) (results not shown in table).
Table 3.
Mean Differencea in “Average” Cognition Score (Average of Scores From All 4 Cognitive Assessments) in Later Life Across Quintiles of Totalb and Dietaryc Intake of Vitamin E, Vitamin C, and Carotenoids, Nurses’ Health Study, 1995–2001
| Antioxidant and Quintile | Median Intaked | Global Composite |
Verbal Composite |
TICS |
|||
|---|---|---|---|---|---|---|---|
| Mean Difference | 95% CI | Mean Difference | 95% CI | Mean Difference | 95% CI | ||
| Total vitamin Ee,f | |||||||
| 1 | 7 | 0.00 | Referent | 0.00 | Referent | 0.00 | Referent |
| 2 | 12 | 0.04 | 0.01, 0.07 | 0.03 | 0.00, 0.07 | 0.17 | 0.04, 0.30 |
| 3 | 31 | 0.05 | 0.02, 0.08 | 0.05 | 0.01, 0.08 | 0.21 | 0.07, 0.34 |
| 4 | 69 | 0.08 | 0.05, 0.11 | 0.09 | 0.05, 0.12 | 0.31 | 0.17, 0.45 |
| 5 | 159 | 0.05 | 0.02, 0.08 | 0.05 | 0.02, 0.09 | 0.16 | 0.02, 0.31 |
| P-trendg | 0.12 | 0.06 | 0.49 | ||||
| Dietary vitamin Ee,h | |||||||
| 1 | 6.2 | 0.00 | Referent | 0.00 | Referent | 0.00 | Referent |
| 2 | 7.1 | 0.03 | 0.00, 0.05 | 0.03 | 0.00, 0.06 | 0.08 | −0.04, 0.21 |
| 3 | 7.8 | 0.04 | 0.01, 0.06 | 0.04 | 0.00, 0.07 | 0.12 | 0.00, 0.25 |
| 4 | 8.5 | 0.04 | 0.01, 0.07 | 0.05 | 0.01, 0.08 | 0.12 | 0.00, 0.25 |
| 5 | 9.9 | 0.05 | 0.02, 0.08 | 0.08 | 0.04, 0.11 | 0.14 | 0.02, 0.27 |
| P-trend | 0.0008 | < 0.0001 | 0.03 | ||||
| Total vitamin Ce,i | |||||||
| 1 | 124 | 0.00 | Referent | 0.00 | Referent | 0.00 | Referent |
| 2 | 178 | −0.02 | −0.05, 0.01 | −0.02 | −0.06, 0.01 | −0.09 | −0.22, 0.03 |
| 3 | 244 | 0.01 | −0.02, 0.04 | 0.01 | −0.02, 0.04 | 0.01 | −0.11, 0.14 |
| 4 | 389 | 0.01 | −0.02, 0.04 | 0.02 | −0.02, 0.05 | 0.00 | −0.13, 0.13 |
| 5 | 761 | 0.00 | −0.03, −0.04 | 0.00 | −0.04, 0.03 | −0.06 | −0.20, 0.07 |
| P-trend | 0.51 | 0.77 | 0.54 | ||||
| Dietary vitamin Ce,j | |||||||
| 1 | 95 | 0.00 | Referent | 0.00 | Referent | 0.00 | Referent |
| 2 | 126 | 0.03 | 0.00, 0.05 | 0.02 | −0.01, 0.05 | 0.16 | 0.04, 0.28 |
| 3 | 149 | 0.01 | −0.02, 0.03 | 0.01 | −0.03, 0.04 | 0.05 | −0.07, 0.18 |
| 4 | 173 | 0.02 | −0.01, 0.05 | 0.03 | 0.00, 0.06 | 0.10 | −0.02, 0.23 |
| 5 | 213 | 0.03 | 0.00, 0.06 | 0.04 | 0.01, 0.07 | 0.16 | 0.03, 0.28 |
| P-trend | 0.12 | 0.009 | 0.05 | ||||
| Total carotenoidse,j | |||||||
| 1 | 8,631 | 0.00 | Referent | 0.00 | Referent | 0.00 | Referent |
| 2 | 11,436 | 0.03 | 0.01, 0.06 | 0.04 | 0.01, 0.07 | 0.09 | −0.03, 0.22 |
| 3 | 13,841 | 0.05 | 0.02, 0.08 | 0.05 | 0.02, 0.09 | 0.13 | 0.01, 0.26 |
| 4 | 16,594 | 0.05 | 0.02, 0.08 | 0.07 | 0.03, 0.10 | 0.11 | −0.02, 0.23 |
| 5 | 21,310 | 0.06 | 0.03, 0.09 | 0.09 | 0.06, 0.13 | 0.13 | 0.00, 0.25 |
| P-trend | 0.0001 | < 0.0001 | 0.08 | ||||
| Dietary carotenoidse,j | |||||||
| 1 | 8,326 | 0.00 | Referent | 0.00 | Referent | 0.00 | Referent |
| 2 | 11,095 | 0.05 | 0.02, 0.08 | 0.07 | 0.04, 0.10 | 0.13 | 0.01, 0.25 |
| 3 | 13,407 | 0.05 | 0.02, 0.08 | 0.06 | 0.03, 0.09 | 0.12 | 0.00, 0.25 |
| 4 | 16,044 | 0.06 | 0.03, 0.09 | 0.08 | 0.05, 0.12 | 0.14 | 0.02, 0.27 |
| 5 | 20,506 | 0.06 | 0.03, 0.09 | 0.10 | 0.07, 0.13 | 0.11 | −0.01, 0.24 |
| P-trend | 0.0005 | < 0.0001 | 0.13 | ||||
Abbreviations: CI, confidence interval; TICS, Telephone Interview for Cognitive Status.
a Mean differences compare the mean “average cognition” in later life for women in the quintile of interest with women in quintile 1 (the referent). Therefore, a positive mean difference indicates that women in the quintile of interest have higher “average cognition” than those in quintile 1; a negative mean difference indicates that women in the quintile of interest have lower “average cognition” than those in quintile 1.
b Total intake includes the contributions of foods and supplements.
c Dietary intake includes the contribution of foods only.
d Median intakes are expressed in g/day for vitamins E and C and in μg/day for carotenoids.
e Adjusted for age (years; continuous), education (registered nurse degree, bachelor's degree, graduate degree), time span between cognitive interviews (years; continuous), antidepressant use (yes, no), alcohol intake (none, 0–14 g/day, ≥15 g/day, or missing data), smoking (never, past, or current), physical activity (quintiles of metabolic equivalent of task-hours/week, or missing data), total energy intake (quintiles of kcal/day), body mass index (<22, 22–24, 25–30, ≥30, or missing data), history of high blood pressure (yes, no), history of high cholesterol (yes, no), history of type 2 diabetes (yes, no), and history of myocardial infarction (yes, no).
f Additionally adjusted for use of vitamin C supplements (yes—regularly, yes—seasonally, no, or missing data), use of multivitamins (yes, no, or missing data), and polyunsaturated fat intake (quintiles).
g Linear tests for trend were performed using the median value of each quintile as a continuous variable in regression models.
h Additionally adjusted for use of vitamin E supplements (yes, no, or missing data), vitamin C supplements, and multivitamins and polyunsaturated fat intake (quintiles).
i Additionally adjusted for use of vitamin E supplements and multivitamins.
j Additionally adjusted for use of vitamin E supplements, vitamin C supplements, and multivitamins.
Finally, there was no evidence of effect modification by smoking status for antioxidant intakes in relation to cognitive decline (e.g., for decline on the global score, P values for interaction were 0.24 for total vitamin E intake, 0.87 for total vitamin C intake, and 0.68 for total carotenoid intake).
DISCUSSION
We found no clear or consistent evidence that greater consumption of vitamin E or vitamin C was associated with cognition in this cohort of older women. However, higher lycopene intake was related to less cognitive decline in this study, and in analyses of overall cognition at older ages (i.e., averaging all 4 of our repeated cognitive measures), we confirmed the relation with lycopene and observed similar associations for total carotenoids and lutein/zeaxanthin as well.
There have been several previous randomized clinical trials of antioxidant supplements and cognition in older adults. In the Age-Related Eye Disease Study (n = 2,166), participants who were given a combination of vitamin E, vitamin C, and beta-carotene over a period of 7 years demonstrated similar levels of cognitive function compared with a placebo group (e.g., mean scores on the Mini-Mental State Examination were 92.7 and 92.1, respectively; P > 0.05) (5). Another trial, the Women's Health Study (n = 6,377), identified a virtually identical risk of substantial cognitive decline in the vitamin E supplement group versus the placebo group over 10 years of supplementation (relative risk = 0.98, 95% CI: 0.83, 1.16) (7). In addition, 2 trials of antioxidants in women with existing cardiovascular risk factors or cardiovascular disease yielded null results. In the Women's Antioxidant and Cardiovascular Study (n = 2,824), Kang et al. (8) found that 9 years of supplementation with vitamin E, vitamin C, and beta-carotene did little to slow cognitive decline (e.g., comparing vitamin E and placebo groups, mean difference in TICS score = −0.01, 95% CI: −0.24, 0.23), and investigators in the Heart Protection Study (n = 20,536) detected no significant differences in mean TICS scores (24.1 vs. 24.0) for persons receiving a combined vitamin E, vitamin C, and beta-carotene supplement versus placebo over a period of 5 years (4). Our data on intakes of vitamins E and C over 15–20 years are generally similar to the results of these trials.
For carotenoids, our data suggest a possible relation between higher levels of very long-term intake and better cognition, especially in our analyses of overall cognitive status at older ages. These results are generally consistent with those of Physicians’ Health Study II (n = 4,052), the longest-term antioxidant trial to date, in which men received beta-carotene supplementation for 18 years and had higher cognitive function scores than those receiving placebo (e.g., mean difference in TICS score = 0.18, 95% CI: 0.01, 0.35) (9). These findings certainly merit further consideration in future research, as increased carotenoid intake would be a simple intervention for maintaining cognitive function, although it appears that consistent consumption over at least 15 years is needed.
Previous observational studies of dietary antioxidants and cognitive decline have yielded inconsistent results, with 2 studies producing null results (24, 25) and 2 others finding inverse relations (26, 27). However, none of these studies focused specifically on long-term diet, and all employed a single measure of diet to estimate antioxidant intakes. In contrast, a major strength of our study is that we collected information on dietary antioxidant intake over 2 decades, allowing us to obtain stable estimates of long-term diet beginning in midlife—a particularly relevant exposure period during which dementia pathogenesis is thought to begin (11).
To our knowledge, the association between higher lycopene intake and slower cognitive decline has not been previously reported in a large, prospective study, although we found consistent relations between higher lycopene intake and cognition across all analyses. Lycopene is a strong antioxidant in the presence of oxygen free radicals (28), and 2 studies have identified a relation between higher levels of plasma lycopene and better cognitive function (29, 30). Growing evidence also indicates that lycopene has multiple microvascular benefits, including prevention of endothelial injury, reduction of inflammatory factors, and inhibition of smooth muscle cell proliferation (31). Furthermore, a recent meta-analysis demonstrated that lycopene significantly reduced serum total cholesterol, low density lipoprotein cholesterol, and systolic blood pressure in human intervention trials (32). Because strong evidence has linked vascular factors and cognitive decline (33), these benefits might represent mechanisms by which lycopene could help to mitigate cognitive decline. Nonetheless, among the various carotenoids that we considered, we did not have any specific a priori hypothesis regarding lycopene, and this finding should be interpreted cautiously.
Several important limitations of this study should be discussed. First, the dietary information used in our analysis was reported by the participants, and therefore nondifferential misclassification of our exposures may have caused us to underestimate the associations of interest. Although this could possibly explain our null findings for most of the antioxidant nutrients, our averaging of multiple dietary assessments helped to reduce random measurement error (34). In addition, we have previously established our ability to detect important diet-cognition associations for other self-reported nutrients in this cohort (35, 36). Second, this was an observational study, and therefore residual confounding could explain our findings, although we would expect such confounding to create rather than obscure associations. In particular, physical activity levels increased across increasing quintiles of carotenoid intake, and although we controlled for physical activity levels and other indicators of healthy lifestyle, we cannot entirely rule out residual confounding by these factors. Third, residual confounding might particularly influence analyses of average cognition due to factors that persist over time (e.g., education) and affect the cognitive level a person achieves more than a person's cognitive trajectory in later life. However, we carefully adjusted for many potentially confounding variables in all analyses, and for the most part these adjustments did not meaningfully change our results for average cognition, which suggests that residual confounding, or unknown confounding factors, would be unlikely to completely explain our findings. In addition, our conclusions regarding a potential association of carotenoid intake and cognition were based on associations observed for both cognitive decline and average cognition, which provides some reassurance that our results are valid. Nonetheless, residual confounding cannot be ruled out entirely, and therefore our findings should be interpreted with caution.
In summary, this study found that long-term consumption of vitamins E and C was not consistently related to cognition in this cohort. However, there appeared to be a relation between higher intake of lycopene and slower cognitive decline, as well as strong associations between greater carotenoid intake and better overall cognition. Future studies should further investigate the potential association between carotenoids and cognitive decline, especially long-term patterns of intake, which are most likely to influence cognitive health.
ACKNOWLEDGMENTS
Author affiliations: Channing Laboratory, Department of Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts (Elizabeth E. Devore, Jae H. Kang, Meir J. Stampfer, Francine Grodstein); Department of Epidemiology, Harvard School of Public Health, Boston, Massachusetts (Meir J. Stampfer, Francine Grodstein); and Department of Nutrition, Harvard School of Public Health, Boston, Massachusetts (Meir J. Stampfer).
The National Institutes of Health provides funding for the Nurses’ Health Study (grant P01 CA87969) and supported Dr. Devore's research on this project (grant F32 AG031633).
Conflict of interest: none declared.
REFERENCES
- 1.Behl C. Amyloid beta-protein toxicity and oxidative stress in Alzheimer's disease. Cell Tissue Res. 1997;290(3):471–480. doi: 10.1007/s004410050955. [DOI] [PubMed] [Google Scholar]
- 2.Christen Y. Oxidative stress and Alzheimer disease. Am J Clin Nutr. 2000;71(2):621S–629S. doi: 10.1093/ajcn/71.2.621s. [DOI] [PubMed] [Google Scholar]
- 3.Socci DJ, Crandall BM, Arendash GW. Chronic antioxidant treatment improves the cognitive performance of aged rats. Brain Res. 1995;693(1-2):88–94. doi: 10.1016/0006-8993(95)00707-w. [DOI] [PubMed] [Google Scholar]
- 4.Heart Protection Study Collaborative Group. MRC/BHF Heart Protection Study of antioxidant vitamin supplementation in 20,536 high-risk individuals: a randomised placebo-controlled trial. Lancet. 2002;360(9326):23–33. doi: 10.1016/S0140-6736(02)09328-5. [DOI] [PubMed] [Google Scholar]
- 5.Yaffe K, Clemons TE, McBee WL, et al. Impact of antioxidants, zinc, and copper on cognition in the elderly: a randomized, controlled trial. Neurology. 2004;63(9):1705–1707. doi: 10.1212/01.wnl.0000142969.19465.8f. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Petersen RC, Thomas RG, Grundman M, et al. Vitamin E and donepezil for the treatment of mild cognitive impairment. N Engl J Med. 2005;352(23):2379–2388. doi: 10.1056/NEJMoa050151. [DOI] [PubMed] [Google Scholar]
- 7.Kang JH, Cook N, Manson J, et al. A randomized trial of vitamin E supplementation and cognitive function in women. Arch Intern Med. 2006;166(22):2462–2468. doi: 10.1001/archinte.166.22.2462. [DOI] [PubMed] [Google Scholar]
- 8.Kang JH, Cook NR, Manson JE, et al. Vitamin E, vitamin C, beta-carotene, and cognitive function among women with or at risk of cardiovascular disease: the Women's Antioxidant and Cardiovascular Study. Circulation. 2009;119(21):2772–2780. doi: 10.1161/CIRCULATIONAHA.108.816900. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Grodstein F, Kang JH, Glynn RJ, et al. A randomized trial of beta-carotene supplementation and cognitive function in men: the Physicians’ Health Study II. Arch Intern Med. 2007;167(20):2184–2190. doi: 10.1001/archinte.167.20.2184. [DOI] [PubMed] [Google Scholar]
- 10.Grodstein F, Chen J, Willett WC. High-dose antioxidant supplements and cognitive function in community-dwelling elderly women. Am J Clin Nutr. 2003;77(4):975–984. doi: 10.1093/ajcn/77.4.975. [DOI] [PubMed] [Google Scholar]
- 11.Launer LJ. The epidemiologic study of dementia: a life-long quest? Neurobiol Aging. 2005;26(3):335–340. doi: 10.1016/j.neurobiolaging.2004.03.016. [DOI] [PubMed] [Google Scholar]
- 12.Willett WC, Sampson L, Stampfer MJ, et al. Reproducibility and validity of a semiquantitative food frequency questionnaire. Am J Epidemiol. 1985;122(1):51–65. doi: 10.1093/oxfordjournals.aje.a114086. [DOI] [PubMed] [Google Scholar]
- 13.Willett WC, Howe GR, Kushi LH. Adjustment for total energy intake in epidemiologic studies. Am J Clin Nutr. 1997;65(4 suppl):1220S–1228S. doi: 10.1093/ajcn/65.4.1220S. [DOI] [PubMed] [Google Scholar]
- 14.Salvini S, Hunter DJ, Sampson L, et al. Food-based validation of a dietary questionnaire: the effects of week-to-week variation in food consumption. Int J Epidemiol. 1989;18(4):858–867. doi: 10.1093/ije/18.4.858. [DOI] [PubMed] [Google Scholar]
- 15.Brandt J, Spencer M, Folstein M. The Telephone Interview for Cognitive Status. Neuropsychol Behav Neurol. 1988;1:111–117. [Google Scholar]
- 16.Albert M, Smith LA, Scherr PA, et al. Use of brief cognitive tests to identify individuals in the community with clinically diagnosed Alzheimer's disease. Int J Neurosci. 1991;57(3-4):167–178. doi: 10.3109/00207459109150691. [DOI] [PubMed] [Google Scholar]
- 17.Morris JC, Heyman A, Mohs RC, et al. The Consortium to Establish a Registry for Alzheimer's Disease (CERAD). Part I. Clinical and neuropsychological assessment of Alzheimer's disease. Neurology. 1989;39(9):1159–1165. doi: 10.1212/wnl.39.9.1159. [DOI] [PubMed] [Google Scholar]
- 18.Monsch AU, Bondi MW, Butters N, et al. Comparisons of verbal fluency tasks in the detection of dementia of the Alzheimer type. Arch Neurol. 1992;49(12):1253–1258. doi: 10.1001/archneur.1992.00530360051017. [DOI] [PubMed] [Google Scholar]
- 19.Baddeley AD, Bressi S, Della Sala S, et al. The decline of working memory in Alzheimer's disease. A longitudinal study. Brain. 1991;114(6):2521–2542. doi: 10.1093/brain/114.6.2521. [DOI] [PubMed] [Google Scholar]
- 20.Locascio JJ, Growdon JH, Corkin S. Cognitive test performance in detecting, staging, and tracking Alzheimer's disease. Arch Neurol. 1995;52(11):1087–1099. doi: 10.1001/archneur.1995.00540350081020. [DOI] [PubMed] [Google Scholar]
- 21.Small BJ, Fratiglioni L, Backman L. Canaries in a coal mine: cognitive markers of preclinical Alzheimer disease. Arch Gen Psychiatry. 2001;58(9):859–860. doi: 10.1001/archpsyc.58.9.859. [DOI] [PubMed] [Google Scholar]
- 22.Welsh K, Butters N, Hughes J, et al. Detection of abnormal memory decline in mild cases of Alzheimer's disease using CERAD neuropsychological measures. Arch Neurol. 1991;48(3):278–281. doi: 10.1001/archneur.1991.00530150046016. [DOI] [PubMed] [Google Scholar]
- 23.Berson EL, Rosner B, Sandberg MA, et al. Clinical trial of lutein in patients with retinitis pigmentosa receiving vitamin A. Arch Ophthalmol. 2010;128(4):403–411. doi: 10.1001/archophthalmol.2010.32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Kalmijn S, Feskens EJ, Launer LJ, et al. Polyunsaturated fatty acids, antioxidants, and cognitive function in very old men. Am J Epidemiol. 1997;145(1):33–41. doi: 10.1093/oxfordjournals.aje.a009029. [DOI] [PubMed] [Google Scholar]
- 25.Dunn JE, Weintraub S, Stoddard AM, et al. Serum alpha-tocopherol, concurrent and past vitamin E intake, and mild cognitive impairment. Neurology. 2007;68(9):670–676. doi: 10.1212/01.wnl.0000255940.13116.86. [DOI] [PubMed] [Google Scholar]
- 26.Morris MC, Evans DA, Bienias JL, et al. Vitamin E and cognitive decline in older persons. Arch Neurol. 2002;59(7):1125–1132. doi: 10.1001/archneur.59.7.1125. [DOI] [PubMed] [Google Scholar]
- 27.Wengreen HJ, Munger RG, Corcoran CD, et al. Antioxidant intake and cognitive function of elderly men and women: the Cache County Study. J Nutr Health Aging. 2007;11(3):230–237. [PubMed] [Google Scholar]
- 28.Di Mascio P, Kaiser S, Sies H. Lycopene as the most efficient biological carotenoid singlet oxygen quencher. Arch Biochem Biophys. 1989;274(2):532–538. doi: 10.1016/0003-9861(89)90467-0. [DOI] [PubMed] [Google Scholar]
- 29.Akbaraly NT, Faure H, Gourlet V, et al. Plasma carotenoid levels and cognitive performance in an elderly population: results of the EVA Study. J Gerontol A Biol Sci Med Sci. 2007;62(3):308–316. doi: 10.1093/gerona/62.3.308. [DOI] [PubMed] [Google Scholar]
- 30.Polidori MC, Pratico D, Mangialasche F, et al. High fruit and vegetable intake is positively correlated with antioxidant status and cognitive performance in healthy subjects. J Alzheimers Dis. 2009;17(4):921–927. doi: 10.3233/JAD-2009-1114. [DOI] [PubMed] [Google Scholar]
- 31.Palozza P, Parrone N, Simone RE, et al. Lycopene in atherosclerosis prevention: an integrated scheme of the potential mechanisms of action from cell culture studies. Arch Biochem Biophys. 2010;504(1):26–33. doi: 10.1016/j.abb.2010.06.031. [DOI] [PubMed] [Google Scholar]
- 32.Ried K, Fakler P. Protective effect of lycopene on serum cholesterol and blood pressure: meta-analyses of intervention trials. Maturitas. 2011;68(4):299–310. doi: 10.1016/j.maturitas.2010.11.018. [DOI] [PubMed] [Google Scholar]
- 33.Grodstein F. Cardiovascular risk factors and cognitive function. Alzheimers Dement. 2007;3(2 suppl):S16–S22. doi: 10.1016/j.jalz.2007.01.001. [DOI] [PubMed] [Google Scholar]
- 34.Willett WC. Nutritional Epidemiology. 2nd ed. New York, NY: Oxford University Press; 1998. [Google Scholar]
- 35.Kang JH, Ascherio A, Grodstein F. Fruit and vegetable consumption and cognitive decline in aging women. Ann Neurol. 2005;57(5):713–720. doi: 10.1002/ana.20476. [DOI] [PubMed] [Google Scholar]
- 36.Devore EE, Stampfer MJ, Breteler MM, et al. Dietary fat intake and cognitive decline in women with type 2 diabetes. Diabetes Care. 2009;32(4):635–640. doi: 10.2337/dc08-1741. [DOI] [PMC free article] [PubMed] [Google Scholar]