Skip to main content
NCBI home page
Alzheimer's & Dementia logoLink to Alzheimer's & Dementia
. 2024 Jun 12;20(7):4594–4601. doi: 10.1002/alz.13884

Concurrent nutrient deficiencies are associated with dementia incidence

Annick P M van Soest 1,✉, Lisette C P G M de Groot 1, Renger F Witkamp 1, Debora Melo van Lent 2,3,4, Sudha Seshadri 2,3,4, Ondine van de Rest 1
PMCID: PMC11247665  PMID: 38865433

Abstract

INTRODUCTION

While observational research suggests a protective role for nutrition in brain aging, intervention studies remain inconclusive. This failing translation from observational to interventional research may result from overlooking nutrient interactions.

METHODS

We developed a nutrient status index capturing the number of suboptimal statuses of omega‐3 fatty acids, homocysteine, and vitamin D (range 0 to 3). We associated this index with dementia incidence in a subsample (age ≥ 50 years) of the Framingham Heart Study Offspring cohort.

RESULTS

Among 968 participants, 79 developed dementia over 15.5 years (median follow‐up). Each point increase in nutrient status index was associated with a 50% higher risk of dementia (hazard ratio [HR] = 1.50; 95% confidence interval [CI] = 1.16, 1.96). Participants with three high‐risk statuses had a four‐fold increased risk of dementia compared to participants without high‐risk status (HR = 4.68; 95% CI = 1.69, 12.94).

DISCUSSION

Concurrent nutrient deficiencies are associated with the risk of dementia. The potential of optimizing nutritional status to lower dementia risk warrants further study.

Highlights

  • Nutrition and dementia research calls for multiple‐nutrient approaches.

  • We studied combined suboptimal statuses of omega‐3 polyunsaturated fatty acids, homocysteine, and vitamin D.

  • Suboptimal status of the three nutrients was associated with dementia risk.

  • The risk estimate was larger than for other factors (ie, diabetes, apolipoprotein E ε4 carrier).

  • Future studies should assess the effect of improving nutrient status on dementia risk.

Keywords: 25‐hydroxyvitamin D, aging, Alzheimer's disease, apolipoprotein E, biomarkers, B vitamins, elderly, nutrition, older adults, polyunsaturated fatty acids, prevention

1. INTRODUCTION

The rapidly increasing prevalence of dementia due to population aging, in combination with the enormous social impact and economic costs of dementia, demonstrate the urgent need for action. In the absence of curative treatment, the interest in preventive strategies is increasing. 1

Preclinical research has indicated that several nutritional factors have the potential to modulate brain aging. 2 Nutrients of interest include omega‐3 polyunsaturated fatty acids (n‐3 PUFAs), B vitamins, vitamin D, antioxidants, and polyphenols, which are mostly assumed to be effective due to their antioxidant, anti‐inflammatory, and vascular health‐promoting properties. 2 Epidemiological research into the association between nutrient intake or status of single nutrients and brain aging generally confirms these preclinical findings. 2 However, clinical trials involving single‐nutrient supplementation have yielded mainly negative results. 2 This raises the question of which factors are responsible for the failure to translate findings from preclinical to clinical research.

A first explanation for the lack of effect of single‐nutrient supplementation in slowing brain aging is that nutrients are part of interacting processes with other nutrients quickly becoming limiting. Indeed, mechanisms underlying nutrition and brain aging are considered multifactorial, 3 and evidence for dietary patterns is stronger than for single nutrients. 2 A second explanation may be the lack of considering baseline nutrient status in the setup of clinical trials. In the majority of trials, participants are selected irrespective of their baseline nutrient status, while only individuals with a nutrient deficiency will likely benefit from nutrient supplementation. 4 This is supported by a secondary analysis of the VITACOG trial, in which the effect of B‐vitamin supplementation on brain atrophy was dependent on baseline homocysteine levels. 5

A better understanding of the cumulative beneficial effects of nutrients and baseline nutrient status may advance the field. However, the literature on multiple nutritional deficiencies in relation to brain aging is limited, with only two longitudinal studies having explored this topic. Here, it was demonstrated that a concurrent nutrient deficiency of n‐3 PUFAs, B vitamins, and vitamin D was associated with steep rates of cognitive decline, 6 and combined suboptimal statuses of n‐3 PUFAs, carotenoids, and vitamin D were strongly associated with an increased risk of dementia. 7 More longitudinal research is needed to reveal the complex interactions between multiple‐nutrient suboptimal statuses and cognitive aging. Specifically, the association between a combined suboptimal status of B vitamins, vitamin D, and n‐3 PUFAs with dementia incidence has not been investigated up till now. Therefore, we developed a nutrient status index including three nutrient biomarkers – homocysteine (as marker of vitamins B6 and B12 and folate status), vitamin D, and n‐3 PUFAs – and associated this index with dementia incidence in the Framingham Heart Study (FHS) Offspring cohort, a prospective community‐based cohort.

2. METHODS

2.1. Study design and population

The FHS is an ongoing prospective community‐based cohort of residents of the city of Framingham, Massachusetts, USA. In 1948, the Original cohort was established to gain insight into the factors contributing to cardiovascular disease. 8 The Offspring cohort was established in 1971 as a second‐generation cohort, including children of the Original cohort and their spouses. A total of 5124 participants have been enrolled in the Offspring cohort. To date, these participants have been studied over 10 examination cycles, about once every 4 years. 9 The study was approved by the Institutional Review Board of Boston University Medical Center, and all participants gave written informed consent.

For this study, we included data from participants aged ≥50 years, free of dementia, with available blood biomarker data on homocysteine, 25‐hydroxyvitamin D, and n‐3 PUFAs. We set the study baseline at exam 7, as biomarker data were measured at this time point. Among the 5124 participants in the FHS Offspring cohort at exam 7, 1525 participants had available data on all three biomarkers. All these participants were free of dementia at baseline. Data from 557 participants were excluded due to being younger than 50 years (n = 130) or missing covariate data (n = 427; of which n = 169 education; n = 25 apolipoprotein E (APOE) carrier status; n = 59 physical activity; n = 8 smoking; n = 107 alcohol intake; and n = 59 depression). Thus, our analysis included data of 968 participants.

2.2. Laboratory measurements

Fasting serum, plasma, and red blood cell samples were collected and stored at −80°C until testing. In the samples collected at exam 7, plasma total homocysteine concentration was measured by high‐performance liquid chromatography with fluorometric detection, 10 and serum 25‐hydroxyvitamin D concentrations were determined by radioimmunoassay (DiaSorin, Stillwater, Minnesota, USA). 11 In the samples collected at exam 8, the fatty acid composition of red blood cell membranes was determined by gas chromatography according to the methods described by Tan and colleagues. 12 The omega‐3 index was calculated using the sum of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) and was expressed as a weight percentage of total fatty acids.

While it would be preferred to have fatty acid composition data also from exam 7, in line with the other biomarker data, we are confident that data from exam 8 are also valid. The fatty acids were measured in red blood cells, which is preferred over measurement in serum or plasma, as red blood cell fatty acid composition is more biologically stable 13 and reflects dietary fatty acid intake over a longer time span (up to ∼120 days). 14 Even though the time between exams 7 and 8 is longer than this time interval, we assume that the red blood cell measurement still provides a reliable and stable representation of the n‐3 PUFA status, as dietary patterns (and, thus, n‐3 PUFA intake) in the elderly are reasonably stable over time, 15 and other factors that may influence variation (eg, geographic and genetic reasons) also have remained stable.

2.3. Ascertainment of incident dementia

Our outcome of interest was incidence of all‐cause dementia, assessed through December 2018. Extensive explanation of the diagnostic procedures used was published previously. 16 In short, participants were continuously screened for cognitive decline. They were flagged for being at risk when they experienced a decline in routinely administered Mini–Mental State Examination performance, when participants, family members, or outside medical records reported subjective cognitive decline or when participants were referred for further screening by FHS staff or physicians. Subsequently, flagged participants underwent additional neuropsychological examination. A neurologist evaluated possible cognitive impairment or dementia and referred participants for dementia review. Dementia diagnosis was made by consensus of at least one neurologist and one neuropsychologist and was based on criteria from the Diagnostic and Statistical Manual of Mental Disorders, 4th edition. 17 If a participant passed away or was lost to follow‐up, the review panel reviewed medical records up to the date of death/loss to follow‐up to determine whether the participant may have had cognitive decline.

RESEARCH IN CONTEXT

  1. Systematic review: A literature review using PubMed regarding associations between concurrent nutrient deficiencies and brain aging yielded only two longitudinal studies. These studies show associations between combined suboptimal status of multiple nutrients with cognitive decline and dementia risk, with relatively large effect sizes. This necessitates further research.

  2. Interpretation: We demonstrated that suboptimal status of one or more critical nutrients (omega‐3 polyunsaturated fatty acids, homocysteine, and vitamin D) was associated with dementia risk, with each additional suboptimal nutrient status elevating this risk. Individuals with suboptimal status for all three nutrients showed a four‐fold higher risk compared to those with optimal levels for all, confirming the potential of multiple‐nutrient strategies for the first time in a US population.

  3. Future directions: Our findings can serve as the basis for designing future intervention trials in dementia prevention, where suboptimal nutrient statuses are selectively optimized, such as through personalized multinutrient supplementation or a brain‐healthy dietary intervention.

2.4. Covariates

Data for all covariates were collected at study baseline (exam 7). Information on age, sex, education level (no high school degree, high school degree, some college, or college degree), smoking status (never, former, or current) was obtained via medical questionnaires. APOE genotype was determined as described previously 18 and classified into carriers and non‐carriers of at least one ε4 allele. Physical activity was self‐reported and measured by the physical activity index. 19 Alcohol consumption was estimated from a food frequency questionnaire and classified as non‐excessive or excessive (< or ≥21 units per week). Hypertension was defined as systolic blood pressure >140 mmHg and/or use of antihypertensive medication, and diabetes was defined as random blood glucose ≥200 mg/dL or fasting blood glucose ≥126 mg/dL or on anti‐diabetic medication. Finally, depression was defined as a score of ≥16 on the Center for Epidemiologic Studies Depression Scale (CES‐D). 20

2.5. Construction of nutrient status index

To construct the nutrient status index as used in our study, we combined the approaches of Bowman and colleagues 6 and Neuffer and colleagues. 7 This nutrient status index indicates the number of high‐risk statuses for three nutrients: homocysteine (as marker of B vitamin status), vitamin D, and n‐3 PUFAs. These nutrient biomarkers were selected a priori on the basis of having a plausible mechanism of action in preventing dementia, having proof from observational studies of the beneficial associations between the nutrient biomarker and dementia risk, 2 and being readily available in the FHS Offspring cohort. The cut‐off for what level is high risk was based on our own data a posteriori. We did this by visualizing the dose‐response relationships between each nutrient and the risk of dementia, using penalized splines in a Cox proportional hazard model. Each model used age at exam 7 (delayed entry) and age at time of event or censoring (age as time scale), nutrient status winsorized at the 2.5th and 97.5th percentiles, and the covariates sex, education, and APOE ε4 carrier status. After visualization of the dose‐response relationships, we set cut‐offs based on graphical inspection of the curves where the splines crossed y = 0.

Figure 1 shows the dose‐response associations between the individual nutrient statuses with dementia. For homocysteine, the cut‐off was set at 8 μmol/L. As homocysteine level was positively associated with dementia risk, participants with homocysteine status ≥8 μmol/L were classified in the high‐risk category, and those with status < 8 μmol/L were classified as low risk. The cut‐off for vitamin D (measured as 25‐hydroxyvitamin D) was set at 15 ng/mL (37.5 nmol/L). For vitamin D levels between 5 and 25 ng/mL (12.5‐62.5 nmol/L), there was an inverse association between vitamin D status and dementia incidence. To this end, participants with vitamin D levels ≤15 ng/mL were classified as high risk and participants with a status >15 ng/mL as low risk. Even though the line y = 0 also crosses the spline at vitamin D level 28 ng/mL, we did not set another cut‐off because very few participants had vitamin D levels of ≥28 ng/mL, and this observation cannot be explained from a physiological perspective. For omega‐3 PUFAs, the cut‐off was set at an omega‐3 index of 5%. As the omega‐3 index was inversely associated with dementia incidence, participants with an omega‐3 index ≤5% were classified as high risk and participants with status > 5% as low risk. Again, while y = 0 also crosses the spline at omega‐3 index 8.5%, we did not set a second cut‐off for reasons given earlier.

FIGURE 1.

FIGURE 1

Open in a new tab

Dose‐response relationships between nutrient status of homocysteine, vitamin D, and omega‐3 polyunsaturated fatty acids and dementia risk, used to set cut‐offs for optimal (low risk) and suboptimal (high risk) status to construct the nutrient status index. EPA, eicosapentaenoic acid; DHA, docosahexaenoic acid.

The nutrient status index captures the number of high‐risk statuses (range 0 to 3). In other words, we assigned a score of 0 if a participant fell into the low‐risk category and a score of 1 if the participant was at high risk. The ultimate nutrient status index sums the values for homocysteine, vitamin D, and n‐3 PUFA status.

2.6. Statistical analysis

For the comparison of baseline characteristics, participants were grouped according to the number of high‐risk nutrient statuses. Baseline characteristics for these groups were compared using ANOVA or Kruskal–Wallis test for continuous variables and chi‐squared for categorical variables.

For the main analyses, we examined the longitudinal association between the nutrient status index and dementia incidence. We performed multivariate‐adjusted Cox proportional hazard models and modeled delayed entry and age as time scale as a function of the nutrient status index (categorical and continuous). Model 1 was adjusted for the covariates age, sex, education, and APOE ε4 carrier status, and Model 2 was additionally adjusted for physical activity, smoking, alcohol intake, hypertension, diabetes, and depression. The proportional hazard assumption was met. Results are presented as adjusted hazard ratios (HRs) accompanied by 95% confidence intervals (CIs). The HRs represent the difference in dementia risk compared to participants without a high‐risk status (categorical) and the change in dementia risk by each unit increase in the nutrient status index (continuous).

For a sensitivity analysis, we evaluated the robustness of the nutrient status index. We tested the effect of adjusting the cut‐offs by 10% by adopting the same cut‐offs as in a previous article, 6 changing the definition of omega‐3 PUFA status by including docosapentaenoic acid (DPA) and by changing the age cut‐off at ≥60 years. The nutrient status indices thus obtained were again associated with dementia incidence similarly to the primary analysis. Additionally, to investigate whether sex or APOE ε4 carrier status modified the association, we tested for interactions with these two variables.

A p‐value < 0.05 and < 0.10 were considered statistically significant for the main analyses and for the tests for interactions, respectively. All analyses were performed using RStudio version 1.1.463. 21

3. RESULTS

3.1. Participant characteristics

Baseline characteristics of the 968 participants are presented in Table 1 and the prevalence of high‐risk status in Table S1. The participants were on average 61.4 ± 7.6 years, and 48% were male. A total of 22% of participants were carriers of at least one APOE ε4 allele. The nutrient status index ranged from 0 (low risk for all nutrients) to 3 (high risk for all nutrients), with the majority of participants (40%) having one high‐risk nutrient status. Participants with a higher number of high‐risk statuses were more likely to be male, had on average a higher body mass index, and were more often current or former smoker.

TABLE 1.

Characteristics of the Framingham Heart Study population per number of high‐risk statuses.

Number of high‐risk statuses
Overall (n = 968) 0 (n = 232) 1 (n = 391) 2 (n = 268) 3 (n = 77) p‐value
Dementia cases n (%) 79 (8%) 7 (3%) 38 (10%) 25 (9%) 9 (12%)
Age (years) 61.4 ± 7.6 60.8 ± 7.1 61.4 ± 7.8 62.1 ± 7.7 61.1 ± 7.8 .33
Sex, n (%) <.001
Male 461 (48%) 75 (32%) 185 (47%) 159 (59%) 42 (55%)
Female 507 (52%) 157 (68%) 206 (53%) 109 (41%) 35 (54%)
APOE ε4 carrier n (%) 215 (22%) 50 (22%) 85 (22%) 60 (22%) 20 (26%) .85
Level of education, n (%) <.001
High school non‐graduate 34 (4%) 4 (2%) 9 (2%) 18 (7%) 3 (4%)
High school graduate 313 (32%) 66 (28%) 126 (32%) 94 (35%) 27 (35%)
Some college 238 (25%) 70 (30%) 81 (21%) 67 (25%) 20 (26%)
College graduate 383 (40%) 92 (40%) 175 (45%) 89 (33%) 27 (35%)
BMI (kg/m2) 28.1 ± 5.1 26.4 ± 4.3 28.3 ± 5.1 28.9 ± 5.1 29.8 ± 5.4 <.001
Physical activity (PAI score) 38.2 ± 6.4 38.7 ± 5.9 37.7 ± 6.3 38.6 ± 7.1 37.9 ± 5.9 .18
Smoking behavior, n (%) .02
Current smoker 30 (3%) 3 (1%) 9 (2%) 13 (5%) 5 (6%)
Former smoker 563 (58%) 131 (56%) 229 (59%) 151 (56%) 53 (69%)
Never‐smoker 374 (39%) 98 (42%) 153 (39%) 104 (39%) 19 (25%)
Systolic blood pressure (mmHg) 126 ± 17 124 ± 17 127 ± 17 127 ± 17 127 ± 16 .07
Use of anti‐hypertensives n (%) 337 (35%) 70 (30%) 148 (38%) 94 (35%) 25 (32%) .26
Depression (CES‐D) 3 [0 – 6] 2 [1 – 7] 3 [0 – 6] 2 [0 – 6] 3 [1 – 7] .42
Cardiovascular disease, n (%) 114 (12%) 20 (9%) 53 (14%) 33 (12%) 8 (10%) .30
Diabetes, n (%) 84 (9%) 13 (6%) 31 (8%) 32 (12%) 8 (10%) .07
Omega‐3 index (wt%) 5.6 ± 1.7 6.8 ± 1.5 5.7 ± 1.7 4.7 ± 1.2 4.0 ± 0.7 <.001
Plasma homocysteine (μmol/L) 8.4 ± 3.8 6.4 ± 1.1 8.1 ± 2.3 9.5 ± 2.5 11.7 ± 9.9 <.001
Serum 25‐hydroxyvitamin D (ng/mL) 19.9 ± 7.7 23.7 ± 6.7 20.7 ± 6.8 18.1 ± 7.9 10.7 ± 2.8 <.001
Open in a new tab

Note: Data are mean ± standard deviation, median [IQR] or number (%).

Abbreviations: APOE, Apolipoprotein E; BMI: body mass index; CES‐D, Center for Epidemiologic Studies Depression Scale; PAI, Physical Activity Index.

3.2. Association nutrient status index with dementia incidence

Among the 968 participants, 79 developed dementia over a median follow‐up of 15.5 [12.9, 19.0] years. In multivariable‐adjusted models, the nutrient status index was associated with dementia incidence (Table 2). Each point increase in nutrient status index was associated with a 50% higher risk of dementia (HR = 1.50; 95% CI = 1.16, 1.96). Moreover, participants with three high‐risk statuses had a four‐fold increased risk of dementia compared to participants without a high‐risk status (HR = 4.64; 95% CI = 1.68, 12.83).

TABLE 2.

Risk of dementia by multinutrient status index.

Crude model Model 1 Model 2
Number of high‐risk statuses
0 (lowest risk) Reference Reference Reference
1 2.98 [1.33, 6.70] 0.008 2.79 [1.23, 6.33] 0.014 2.89 [1.27, 6.58] 0.012
2 3.03 [1.31, 7.00] 0.010 3.09 [1.32, 7.24] 0.009 3.28 [1.38, 7.80] 0.007
3 (highest risk) 4.30 [1.60, 11.56] 0.004 4.70 [1.74, 12.69] 0.002 4.68 [1.69, 12.94] 0.003
Continuous 1.43 [1.11, 1.83] 0.005 1.48 [1.15, 1.93] 0.002 1.50 [1.16, 1.96] 0.002
Open in a new tab

Note: Data are hazard ratio [95% confidence interval] p‐value. Model 1: adjusted for age, sex, education, and APOE ε4 carrier status; Model 2: additionally adjusted for physical activity, smoking, alcohol consumption, diabetes, hypertension, and depression. In the continuous analysis, data are shown per point increment in nutrient status index.

3.3. Sensitivity analyses

We performed sensitivity analyses to assess the robustness of the association between the nutrient status index with dementia incidence. Overall, the association was robust to changes in nutrient status cut‐offs (Table S2). Varying the cut‐offs by 10% and adopting the same cut‐offs as Bowman and colleagues 6 did not alter the results. Similarly, results were robust to variations in study population and components. Changing the age cut‐off to ≥60 years or adapting the definition of the omega‐3 index by including DPA in addition to EPA and DHA did not change the results.

Subsequently, we tested for interactions to investigate whether sex and APOE ε4 carrier status modified the association between the nutrient status index and dementia incidence (Table S3). Interestingly, APOE ε4 carrier status appeared to influence the association (p interaction = 0.01). The nutrient status index (continuous) was positively associated with dementia incidence in carriers (HRper point increase = 2.05, 95% CI = 1.23, 2.44), but not in non‐carriers (HRper point increase = 1.11, 95% CI = 0.77, 1.59). There was no significant overall interaction between sex and nutrient status index (p = .23).

To put our found effect size for the nutrient status index into context, we assessed how many known risk factors increased the risk of developing dementia in our sample. The risk of dementia was doubled for current smokers (HR = 1.97; 95% CI = 0.53, 7.32) and for individuals with diabetes (HR = 2.18; 95% CI = 1.14, 4.17) and tripled for carriers of the APOE ε4 allele (HR = 3.11; 95% CI = 1.92, 5.05) (Table S4).

4. DISCUSSION

Using the nutrient status index developed for our study, we found that individuals with a higher index, that is, having suboptimal status of homocysteine (as a marker of B vitamins), vitamin D, and n‐3 PUFAs, had a higher risk of developing dementia compared to those with a lower index. Remarkably, a suboptimal status of all three nutrients was associated with a four‐fold increased risk of dementia compared to individuals without suboptimal statuses. In addition, APOE ε4 carrier status appeared to influence the association between nutrient status index and dementia incidence, with the association only evident in APOE ε4 carriers.

The effect size we observed was substantial: a four‐fold increased risk of developing dementia in individuals with combined suboptimal status of n‐3 PUFAs, vitamin D, and homocysteine. This effect size is large in comparison with other risk factors of dementia. In our sample, being a current smoker or having diabetes doubled the risk, and being a carrier of at least one APOE ε4 allele tripled the risk of dementia.

Previous research complements our results and the large effect sizes. To our knowledge, the association between multiple‐nutrient suboptimal status and brain aging has been investigated in two other studies. In a secondary analysis of the Bordeaux Three‐City (3C) study, Neuffer and colleagues developed a nutrient status index composed of n‐3 PUFAs (EPA+DHA+DPA), carotenoids, and 25(OH)D. Similarly to our approach, they set nutrient cut‐offs based on their own data (n3‐PUFA at 3 and 4.5%, carotenoids at 100 and 200 μg/mmol, and vitamin D at 8 and 26 ng/mL). A higher nutrient status index (ie, more nutrient suboptimal statuses) was associated with a higher risk to develop dementia. The 13% of participants with the highest nutrient status index had a four‐fold increased chance of developing dementia compared to the 21% with the lowest index scores. 7 Additionally, Bowman and colleagues investigated the role of combined deficiencies in n‐3 PUFAs (EPA + DHA), 25(OH)D, and homocysteine on cognitive decline in a secondary analysis of the French Multi‐domain Alzheimer's Prevention Trial (MAPT), with cut‐offs for nutrient deficiencies set a priori (n‐3 PUFA 4.82%, Hcy 14 μmol/L, vitamin D 20 ng/mL). Individuals without nutrient deficiencies of these nutrients showed cognitive improvements over 3 years, while each additional nutrient deficiency led to an incremental faster rate of cognitive decline. 6

While this earlier research also demonstrated associations between multiple‐nutrient suboptimal statuses and brain aging, direct comparison between results is complicated by differences in components and cut‐offs.

Regarding components, in the 3C study data were available on carotenoid but not on homocysteine status. Carotenoids are also nutrients of prime interest in relation to the aging brain. These nutrients reduce oxidative stress, a mechanism involved in the pathogenesis of dementia. 22 Additionally, higher carotenoid status has been associated with lower odds of dementia. 23 It is a limitation of the current study that we did not have data available on carotenoid status or on other antioxidant nutrient statuses like vitamin C or E. Further research on multinutrient suboptimal statuses should consider incorporating antioxidant nutrients, alongside n‐3 PUFAs, homocysteine, and vitamin D status.

With respect to cut‐offs, our homocysteine cut‐off (8 μmol/L) was lower compared to the French MAPT trial (14 μmol/L), which was anticipated because of folate fortification in the United States. Our low cut‐off is likely not applicable in countries where foods are not fortified. Vitamin D cut‐offs among studies varied, ranging from 15 ng/mL in FHS, 20 ng/mL in MAPT, to 8 and 26 ng/mL in 3C. Our cut‐off is lower than the World Health Organization guidelines of 20 ng/mL (50 nmol/L). However, this cut‐off was set for bone health rather than brain health. Our n‐3 PUFA cut‐off (5%) was slightly higher compared to the 3C study (3 and 4.5%) and MAPT trial (4.82%) but still low compared to the target range of 8% to 11%. 24 While increasing the cut‐off to this target range could provide even stronger protective associations, the baseline omega‐3 index of our study population was too low to ensure sufficient contrast. All in all, optimal nutrient cut‐offs may be population‐specific, and therefore, it can be seen as a limitation that we set cut‐offs based on our own data. Nevertheless, the nutrient status index applied in our study was robust to variations in cut‐offs, as demonstrated in the sensitivity analyses, and this adds to the robustness of our findings. In addition, despite methodological differences between our and previous studies, results are remarkably consistent.

The observation that there is an association between multiple suboptimal nutrient statuses and dementia risk is biologically plausible. Preclinical studies stress that brain aging depends on multiple, dynamically interacting mechanisms, with nutrients playing distinctive roles. Vitamin D, among others, promotes healthy brain aging by suppressing amyloid beta deposition, regulating calcium homeostasis, and reducing oxidative stress and inflammation. 25 Omega‐3 fatty acids also possess anti‐inflammatory and antioxidant properties, as well as vascular health‐promoting effects. In addition, these fatty acids serve as building blocks for neuronal tissue. 26 Homocysteine has been shown to negatively impact the aging brain by impairing vascular functioning and increasing tau phosphorylation and via inhibition of methylation reactions. 27 Considering the multifactorial nature of dementia, it is conceivable that these single effects of the nutrients targeting different mechanisms of action have additive effects.

In addition to these additive effects, it is possible that nutrients act in a synergistic manner. This has been hypothesized for homocysteine and n‐3 PUFAs, as a consequence of the regulatory role of homocysteine in the transport of n‐3 PUFAs to the brain. 28 DHA is transported with the help of phosphatidylcholine (PC), the formation of which is dependent on homocysteine levels. Elevated homocysteine levels decrease the activity of phosphatidylethanolamine N‐methyltransferase, the enzyme responsible for the conversion of phosphatidylethanolamine to PC. Consequently, this results in low transport of n‐3 PUFAs to the brain. 28 Indeed, the synergistic effects between homocysteine and n‐3 PUFAs have been confirmed in secondary analyses of B vitamin 29 , 30 , 31 and n‐3 PUFA 32 supplementation trials.

For further research, we strongly encourage researchers with access to data on multiple‐nutrient statuses and brain aging outcomes to further investigate the potential of multinutrient suboptimal statuses. This will give more insight into optimal nutrient status cut‐offs. Additionally, these results can be the basis for the design of clinical trials, in which the nutrient status index can be used to select participants at nutritional risk of dementia. 4 Participants may then undergo nutrient supplementation to correct suboptimal status. Instead of nutrient supplementation, participants could undergo a diet intervention targeted at improving general dietary intake, as suboptimal status of the three nutrients investigated in this research could also be a proxy for general suboptimal nutritional status.

Another topic that deserves further investigation is the interaction with the APOE ε4 genotype. We observed an interaction with this genotype, with strong associations between multinutrient suboptimal statuses and dementia in carriers, but not in non‐carriers of the APOE ε4 allele. This interaction has not been investigated in the two previous articles on multinutrient suboptimal statuses. 6 , 7 However, a large body of literature is available on the interaction between APOE genotype and n‐3 PUFAs in relation to brain aging. According to this literature, APOE ε4 carriers seem more susceptible to the benefits of omega‐3 PUFAs in preclinical stages, while benefits in the clinical stages are limited to non‐carriers. 33 The mechanistic rationale on why APOE ε4 carriers may need more n‐3 PUFAs during the preclinical stage is that they experience accelerated DHA catabolism and less efficient transport of DHA both across the blood brain barrier and within the brain. These processes occur before the onset of neurodegeneration. 33 At baseline, our study population was likely in the preclinical stage of dementia, considering the relatively young age (≥50 years). Literature on vitamin D and homocysteine in relation to the APOE genotype is limited, but a pattern similar to that of n‐3 PUFAs has been demonstrated for other preventive strategies to lower dementia risk, with APOE ε4 carriers benefitting more in preclinical stages. 34 However, it is important to emphasize that these results come from a subgroup analysis, so interpretation is limited. Further research is required to confirm the interaction with the APOE ε4 genotype.

In conclusion, in our community‐based sample, concurrent suboptimal status of n‐3 PUFAs, homocysteine, and vitamin D is associated with the risk of dementia. The results support earlier observations that multiple‐nutrient suboptimal status is highly detrimental for brain aging, suggesting that nutrition is a key modifiable risk factor for dementia. Further research is needed to optimize nutrient status cut‐offs and to study the potential of optimizing nutritional status to lower dementia risk.

CONFLICT OF INTEREST STATEMENT

Dr. Melo van Lent is chair of the Alzheimer's Association ISTAART Nutrition Metabolism and Dementia Professional Interest Area. All other authors report no conflicts of interest. Author disclosures are available in the supporting information.

CONSENT STATEMENT

All participants provided written informed consent. The study protocol was approved by the Institutional Review Board at Boston University Medical Center.

Supporting information

Supporting Information

ALZ-20-4594-s001.docx (22.4KB, docx)

Supporting Information

ALZ-20-4594-s002.pdf (870.9KB, pdf)

ACKNOWLEDGMENTS

The authors gratefully acknowledge all study participants for their time. The FHS was supported by the National Heart, Lung, and Blood Institute (contracts N01‐ HC‐25195, HHSN268201500001I, and 75N92019D00031), the National Institute on Aging (NIA) (R01 AG054076, R01 AG049607, R01 AG033193, U01 AG049505, U01 AG052409, U01 AG058589, RF1 AG059421), and grants from the National Institute of Neurological Disorders and Stroke (NS017950 and UH2 NS100605). Dr. Melo van Lent received funding from the NIH‐NIA (R03 AG067062‐01) and an Alzheimer's Association Research Fellowship grant (AARF‐22‐918316) and is supported by NIH‐NIA grants 1RF1AG059421 and 1P30 AG066546‐01A1. Dr. Sudha Seshadri is partially supported by the South Texas Alzheimer's Disease Center (1P30AG066546‐01A1) and the Bill and Rebecca Reed Endowment for Precision Therapies and Palliative Care. Dr. Seshadri is also supported by 1RF1AG059421 and by an endowment from the Barker Foundation as the Robert R. Barker Distinguished University Professor of Neurology, Psychiatry and Cellular and Integrative Physiology. The funding agencies had no role in study design, in the collection, analysis, and interpretation of data, in the writing of the report, or in the decision to submit the article for publication.

van Soest APM, de Groot LCPGM, Witkamp RF, van Lent DM, Seshadri S, van de Rest O. Concurrent nutrient deficiencies are associated with dementia incidence. Alzheimer's Dement. 2024;20:4594–4601. 10.1002/alz.13884

REFERENCES

  • 1. World Health Organization . Dementia: a public health priority. World Health Organization; 2019. [Google Scholar]
  • 2. Scarmeas N, Anastasiou CA, Yannakoulia M. Nutrition and prevention of cognitive impairment. Lancet Neurol. 2018;17(11):1006‐1015. doi: 10.1016/S1474-4422(18)30338-7 [DOI] [PubMed] [Google Scholar]
  • 3. Yassine HN, Samieri C, Livingston G, et al. Nutrition state of science and dementia prevention: recommendations of the Nutrition for Dementia Prevention Working Group. Lancet Healthy Longev. 2022;3(7):e501‐e512. doi: 10.1016/S2666-7568(22)00120-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Morris MC, Tangney CC. A potential design flaw of randomized trials of vitamin supplements. JAMA. 2011;305(13):1348‐1349. doi: 10.1001/jama.2011.383 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Smith AD, Smith SM, de Jager CA, et al. Homocysteine‐lowering by b vitamins slows the rate of accelerated brain atrophy in mild cognitive impairment: a randomized controlled trial. PLoS ONE. 2010;5(9):1‐10. doi: 10.1371/journal.pone.0012244 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Bowman GL, Dodge HH, Guyonnet S, et al. A blood‐based nutritional risk index explains cognitive enhancement and decline in the multidomain Alzheimer prevention trial. Alzheimers Dement (N Y). 2019;5:953‐963. doi: 10.1016/j.trci.2019.11.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Neuffer J, Gourru M, Thomas A, et al. A biological index to screen multi‐micronutrient deficiencies associated with the risk to develop dementia in older persons from the community. J Alzheimers Dis. 2022;85(1):331‐342. doi: 10.3233/jad-215011 [DOI] [PubMed] [Google Scholar]
  • 8. Dawber TR, Meadors GF, Moore Jr FE. Epidemiological approaches to heart disease: the Framingham study. Am J Public Health. 1951;41(3):279‐281. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Feinleib M, Kannel WB, Garrison RJ, McNamara PM, Castelli WP. The Framingham offspring study. design and preliminary data. Prev Med. 1975;4(4):518‐525. doi: 10.1016/0091-7435(75)90037-7 [DOI] [PubMed] [Google Scholar]
  • 10. Araki A, Sako Y. Determination of free and total homocysteine in human plasma by high‐performance liquid chromatography with fluorescence detection. J Chromatogr. 1987;422(C):43‐52. doi: 10.1016/0378-4347(87)80438-3 [DOI] [PubMed] [Google Scholar]
  • 11. Wang TJ, Pencina MJ, Booth SL, et al. Vitamin D deficiency and risk of cardiovascular disease. Circulation. 2008;117(4):503‐511. doi: 10.1161/CIRCULATIONAHA.107.706127 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Tan Z, Harris W, Beiser A, et al. Red blood cell omega‐3 fatty acid levels and markers of accelerated brain aging. Neurology. 2012;78(9):658‐664. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Harris WS, Thomas RM. Biological variability of blood omega‐3 biomarkers. Clin Biochem. 2010;43(3):338‐340. doi: 10.1016/j.clinbiochem.2009.08.016 [DOI] [PubMed] [Google Scholar]
  • 14. Arab L. Biomarkers of fat and fatty acid intake. J Nutr. 2003;133:925S‐932S. doi: 10.1093/jn/133.3.925s [DOI] [PubMed] [Google Scholar]
  • 15. Jankovic N, Steppel MT, Kampman E, et al. Stability of dietary patterns assessed with reduced rank regression; The Zutphen Elderly Study. Nutr J. 2014;13(1):30. doi: 10.1186/1475-2891-13-30 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Satizabal CL, Beiser AS, Chouraki V, Chêne G, Dufouil C, Seshadri S. Incidence of dementia over three decades in the Framingham heart study. N Engl J Me. 2016;374(6):523‐532. doi: 10.1056/NEJMoa1504327 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Bell CC. DSM‐IV: diagnostic and statistical manual of mental disorders. Jama. 1994;272(10):828‐829. [Google Scholar]
  • 18. Hixson JE, Vernier DT. Restriction isotyping of human apolipoprotein E by gene amplification and cleavage with HhaI. J Lipid Res. 1990;31(3):545‐548. [PubMed] [Google Scholar]
  • 19. Kannel WB, Sorlie P. Some health benefits of physical activity: the framingham study. Arch Intern Med. 1979;139(8):857‐861. doi: 10.1001/archinte.1979.03630450011006 [DOI] [PubMed] [Google Scholar]
  • 20. Radloff LS. The CES‐D scale: a self‐report depression scale for research in the general population. Appl Psychol Meas. 1977;1(3):385‐401. doi: 10.1177/014662167700100306 [DOI] [Google Scholar]
  • 21. R Core Team . R: A language and enironment for statistical computing. R Foundation for Statistical Computing. https://www.R‐project.org/ [Google Scholar]
  • 22. Mecocci P, Boccardi V, Cecchetti R, et al. A long journey into aging, brain aging, and Alzheimer's disease following the oxidative stress tracks. J Alzheimers Dis. 2018;62(3):1319‐1335. doi: 10.3233/JAD-170732 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Wang L, Zhao T, Zhu X, Jiang Q. Low blood carotenoid status in dementia and mild cognitive impairment: a systematic review and meta‐analysis. BMC Geriatr. 2023;23(1):195. doi: 10.1186/s12877-023-03900-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. von Schacky C, Kuipers RS, Pijl H, Muskiet FAJ, Grobbee DE. Omega‐3 fatty acids in heart disease—why accurately measured levels matter. Neth Heart J. 2023;31(11):415‐423. doi: 10.1007/s12471-023-01759-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Annweiler C. Vitamin D in dementia prevention. Ann N Y Acad Sci. 2016;1367(1):57‐63. doi: 10.1111/nyas.13058 [DOI] [PubMed] [Google Scholar]
  • 26. Dyall SC. Long‐chain omega‐3 fatty acids and the brain: a review of the independent and shared effects of EPA, DPA and DHA. Front Aging Neurosci. 2015;7(APR):52. doi: 10.3389/fnagi.2015.00052 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Smith AD, Refsum H. Homocysteine, B vitamins, and cognitive impairment. Annu Rev Nutr. 2016;36:211‐239. doi: 10.1146/annurev-nutr-071715-050947 [DOI] [PubMed] [Google Scholar]
  • 28. Selley ML. A metabolic link between S‐adenosylhomocysteine and polyunsaturated fatty acid metabolism in Alzheimer's disease. Neurobiol Aging. 2007;28(12):1834‐1839. doi: 10.1016/j.neurobiolaging.2006.08.003 [DOI] [PubMed] [Google Scholar]
  • 29. van Soest APM, van de Rest O, Witkamp RF, Cederholm T, de Groot LCPGM. DHA status influences effects of B‐vitamin supplementation on cognitive ageing: a post‐hoc analysis of the B‐proof trial. Eur J Nutr. 2022;61(7):3731‐3739. doi: 10.1007/s00394-022-02924-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Oulhaj A, Jernerén F, Refsum H, Smith AD, de Jager CA. Omega‐3 fatty acid status enhances the prevention of cognitive decline by B vitamins in mild cognitive impairment. J Alzheimers Dis. 2016;50(2):547‐557. doi: 10.3233/JAD-150777 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Jernerén F, Elshorbagy AK, Oulhaj A, Smith SM, Refsum H, Smith AD. Brain atrophy in cognitively impaired elderly: the importance of long‐chain ω‐3 fatty acids and B vitamin status in a randomized controlled trial. Am J Clin Nutr. 2015;102(1):215‐221. doi: 10.3945/ajcn.114.103283 [DOI] [PubMed] [Google Scholar]
  • 32. Jernerén F, Cederholm T, Refsum H, et al. Homocysteine status modifies the treatment effect of omega‐3 fatty acids on cognition in a randomized clinical trial in mild to moderate Alzheimer's disease: the OmegAD study. J Alzheimers Dis. 2019;69(1):189‐197. doi: 10.3233/JAD-181148 [DOI] [PubMed] [Google Scholar]
  • 33. Yassine HN, Braskie MN, Mack WJ, et al. Association of docosahexaenoic acid supplementation with Alzheimer disease stage in apolipoprotein e ϵ4 carriers: a review. JAMA Neurology. 2017;74(3):339‐347. doi: 10.1001/jamaneurol.2016.4899 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. Angelopoulou E, Paudel YN, Papageorgiou SG, Piperi C. APOE genotype and Alzheimer's disease: the influence of lifestyle and environmental factors. ACS Chem Neurosci. 2021;12(15):2749‐2764. doi: 10.1021/acschemneuro.1c00295 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting Information

ALZ-20-4594-s001.docx (22.4KB, docx)

Supporting Information

ALZ-20-4594-s002.pdf (870.9KB, pdf)

Articles from Alzheimer's & Dementia are provided here courtesy of Wiley

RESOURCES