Abstract
Longitudinal studies provide conflicting evidence regarding the impact of vitamin D deficiency on cognitive performance in older individuals. The present study aimed to investigate whether vitamin D deficiency is a risk factor for poorer trajectories in global cognition and specific cognitive domains over a six-year follow-up period. This cohort study analysed data from 2625 participants aged 50 years or older from the English Longitudinal Study of Ageing. Vitamin D [25-hydroxyvitamin D, 25(OH)D] was classified as sufficient (> 75 nmol/L), insufficient (> 30 and ≤ 75 nmol/L), or deficient (≤ 30 nmol/L). Cognitive performance was assessed using the verbal fluency test, word list test, temporal orientation test, and global cognition, standardised in z-scores by age group and educational level. Generalised linear mixed models, adjusted for sociodemographic, behavioural, and clinical variables, estimated rates of decline in global cognition and cognitive domains based on vitamin D status. Participants with vitamin D deficiency showed greater declines in the z-score of global cognition (− 0.037 standard deviations (SD) per year; 95% CI: − 0.069 to − 0.005) and executive function (− 0.038 SD per year; 95% CI: − 0.071 to − 0.004) compared to those with sufficient vitamin D levels. No significant differences were observed in the trajectories of memory or temporal orientation in relation to vitamin D status. Vitamin D deficiency is a risk factor for declines in global cognition and executive function in individuals aged 50 or older. The assessment and management of vitamin D levels may be crucial strategies for promoting cognitive health.
Supplementary Information
The online version contains supplementary material available at 10.1007/s11357-025-01800-9.
Keywords: Ageing, Cognition, Epidemiology, Executive function, Longitudinal study
Introduction
Cognitive impairment affects up to 23% of older adults worldwide, with an estimated 46% of cases progressing to dementia within three years [1]. Cognitive impairment is one of the primary causes of loss of autonomy and independence, negatively impacting the quality of life for those affected [2]. Given the social and economic repercussions of cognitive decline, which escalate with the ageing population, identifying its risk factors is crucial for developing effective prevention strategies [3].
Vitamin D deficiency, measured by serum concentrations of 25-hydroxyvitamin D [25(OH)D], is regarded as a significant public health issue, affecting approximately one billion individuals worldwide and associated with cognitive impairment functioning [4].
One of the primary causes of vitamin D deficiency is insufficient exposure to sunlight, from which 80 to 90% of the serum concentration of 25(OH)D is derived. Additional factors include low dietary vitamin D intake, diabetes mellitus, cardiovascular disease, and compromised immune function. Due to the ageing process, older adults experience physiological changes that may increase the risk of vitamin D deficiency. Age-related factors, such as a reduction in vitamin D receptors—potentially resulting from decreased skin thickness—a decline in renal hydroxylation, and impaired intestinal absorption, may contribute to lower circulating levels of vitamin D 25(OH)D [5].
Vitamin D plays a crucial role in regulating the bone, cardiovascular, endocrine, and nervous systems, as well as maintaining cognitive health [4]. The vitamin D receptors (VDRs) are located in brain regions essential for cognitive functioning, such as the prefrontal cortex, hippocampus, hypothalamus, and cerebellum. VDRs contribute to neuroprotection, exerting an anti-inflammatory effect and reducing oxidative stress [6]. Thus, it is plausible that deficient 25(OH)D levels may lead to neurocognitive impairment [7].
According to a meta-analysis by Harse et al., [8] studies have demonstrated an association between vitamin D deficiency and cognitive impairment in cross-sectional analyses involving older adults. However, longitudinal studies on this association offer conflicting evidence. While some studies have found an association between 25(OH)D concentrations and increased risk of cognitive decline [9–11], others have found no association over time [3, 12].
For example, Toffanello et al. [13] found that individuals with vitamin D deficiency or insufficiency were at greater risk of cognitive impairment in a four-year follow-up period than those with sufficient 25(OH)D levels. In contrast, Grajdean et al. [14] investigated the association between serum 25(OH)D levels and changes in memory performance and temporal orientation and found no significant decline over six years.
Furthermore, there is a lack of longitudinal studies that analyse the trajectory of global cognition and cognitive domains based on serum concentrations of 25(OH)D. Consequently, the present study aimed to investigate whether vitamin D deficiency is a risk factor for poorer trajectories of global cognitive performance and cognitive domains in individuals aged 50 or older over a period of six years.
Methods
Study sample
Data were extracted from the English Longitudinal Study of Ageing (ELSA), a longitudinal panel study involving a representative sample of community-dwelling men and women aged 50 and older [15]. The ELSA Study has been collecting information on sociodemographic, economic, and health-related characteristics since 2002. Nurses commenced visiting participants’ homes in 2004 to collect blood samples, take anthropometric measurements, and assess physical performance. Detailed information on the study and sampling procedures can be found in a previous publication [16].
As serum concentrations of 25-hydroxyvitamin D [25(OH)D] were first collected in wave 6 of the ELSA Study (2012), this wave was defined as the baseline for the present study. Among the 5659 participants in 2012, those with missing data on 25(OH)D concentrations, cognitive assessment test performance, or covariates, along with individuals showing signs of cognitive decline (z-score < − 1 SD), were excluded [17]. Thus, the final sample for the present study comprised 2625 individuals at baseline, and the participants were reassessed after four years (wave 8, 2016) and six years (wave 9, 2018) of follow-up. Figure 1 illustrates the flowchart of the sample selection process.
Fig. 1.
Flowchart of participants included in analyses
Cognitive function measures
The cognitive assessment tests used in the ELSA Study were developed and validated by the US Health and Retirement Study for epidemiological studies [18, 19], and have been used in longitudinal studies involving older adults in the UK [16, 20].
Memory was assessed using the word list test. The participant listened to a list of ten words and was asked to repeat them (immediate recall). After approximately 2 min, the participant was requested to repeat as many of the exact words as possible (delayed recall). The score was calculated based on the number of correctly recalled words from both sections of the test, ranging from 0 to 20 points [21].
Executive function was assessed using the verbal fluency test. The participant was asked to name the greatest number of animals in one minute. Names that fell within the animal category were awarded one point (repeated names were not counted), with higher scores indicating better performance on the test [22].
Temporal orientation was assessed using four questions regarding the day of the month, the day of the week, the month, and the year, awarding one point for each correct answer (total: 0 to 4 points) [15].
Global cognition comprised the total of the scores from the tests utilised to assess the three cognitive domains, with higher overall scores indicating better cognitive performance functioning.
The scores on the three cognitive tests were standardised as z-scores for each age group (50–59, 60–69, 70–79, and 80 years or older) and schooling (0–11, 12–13, and > 13 years, following the English classification), as these variables exert a strong influence on cognitive functioning [23, 24]. The mean and standard deviation (SD) of the scores for each age group and schooling level at baseline were used to calculate the z-scores for subsequent waves of the study, facilitating comparisons and the observation of cognitive decline over time [25].
Vitamin D assessment
Vitamin D levels were measured based on serum concentrations of 25(OH)D determined in blood samples collected after fasting during visits from nursing staff. The samples were analysed at the Royal Victoria Infirmary (Newcastle upon Tyne, UK) and conducted in duplicate using chemiluminescence (DiaSorin Liaison immunoassay), which has an analytical sensitivity of 7.5 nmol/L and a coefficient of variation ranging from 8.7% to 9.4%. The laboratory that performed the 25(OH)D analyses is part of the Vitamin D External Quality Assessment Schemes (DEQAS) [26].
Serum concentrations of 25(OH)D were classified as sufficient (> 75 nmol/L), insufficient (> 30 and ≤ 75 nmol/L), or deficient (≤ 30 nmol/L) [27, 28]. Common cut-off points used to define deficiency range from < 25 to < 64 nmol/L, definitions of insufficiency range from 25–50 to 65–74 nmol/L, and sufficiency is defined as > 75 nmol/L for optimal health, particularly in older adults [29–31]. Furthermore, large population cohorts have demonstrated a significant association between low 25(OH)D levels (usually < 30 nmol/L) and poorer cognitive performance compared to adequate 25(OH)D levels (> 75 nmol/L) [7].
Covariates
Variables that might confound the association between vitamin D deficiency and cognitive decline were utilised.
The sociodemographic variables included age (in years, centralised and fixed at baseline), sex, family wealth (divided into quintiles), and race (white or non-white).
The behavioural variables included smoking (non-smoker, ex-smoker or smoker), alcohol intake, measured by weekly frequency (rarely or never: up to once per week; frequently: two to six times per week; and daily) [32] based on previous longitudinal studies that analysed the association between alcohol intake and cognitive function [33, 34], and the practice of physical activity.
Given the potential for sun exposure related to physical activity on serum vitamin D levels and its possible influence on cognitive performance, physical activity level was included as one of the covariates [35]. This was assessed using the Physical Activity and Sedentary Behaviour Assessment Questionnaire, validated in the Health Survey for England [36].s The participants reported the frequency (more than once per week, once per week, one to three times per month, or hardly ever/never) with which they practised physical exercises of mild (vacuuming the home, washing clothes, and home repairs), moderate (gardening, washing the car, walking at a moderate pace, dancing, and stretching), and vigorous intensity (running, swimming, cycling, tennis, and aerobics weightlifting) [15]. The participants were classified as physically active (practised moderate or vigorous activity more than once per week) or physically inactive (practised vigorous or moderate activity up to once per week, one to three times per month, hardly ever or never; practised any mild physical activity) [37].
Clinical conditions were recorded based on self-reports of a medical diagnosis of myocardial infarction, heart failure, angina, asthma, stroke, systemic arterial hypertension, and diabetes mellitus. Depressive symptoms were investigated using the short version of the Center for Epidemiologic Studies Depression (CES-D) Scale, considering a cut-off point of ≥ 4 points [38]. Serum C-reactive protein levels were measured using the N Latex CRP mono immunoassay in the Dade Behring Nephelometer II Analyser, adhering to the quality control procedures outlined in the technical report of the Health Survey for England. C-reactive protein levels were analysed as continuous quantitative variables [39].
The anthropometric measure was waist circumference, taken twice with a flexible metric tape at the midpoint between the lowest rib and the upper edge of the iliac crest. The average of the two measurements was used. Abdominal obesity was recorded when the waist circumference exceeded 102 cm for men and 88 cm for women (National Institutes of Health, 1998) [40].
The biochemical measures included LDL cholesterol (considered high when ≥ 100 mg/dL) and HDL cholesterol (considered low when < 40 mg/dL for men and < 50 mg/dL for women) [41].
The season of the year in which blood was collected for the determination of serum concentrations of 25(OH)D was recorded as a control variable: spring (March to May), summer (June to August), autumn (September to November), and winter (December to February) [26]. The use of vitamin D supplementation and carbamazepine, an anticonvulsant that can lower 25(OH)D levels, was also considered [42].
Statistical analysis
The sample characteristics at baseline were expressed as mean, standard deviations and percentages. Differences in baseline characteristics according to vitamin D status were analysed using the chi-squared test, analysis of variance (ANOVA), and Tukey’s post-hoc test, with the significance level set at 5% (p < 0.05).
Generalised linear mixed models were employed to estimate trajectories of global cognition and specific cognitive domains as a function of vitamin D status (sufficient, insufficient, or deficient) over a six-year period. The models were adjusted for sociodemographic, behavioural, clinical, anthropometric, and biochemical variables. All variables were assessed repeatedly during each wave of follow-up, thus enabling a more robust temporal analysis of the associations. Such models facilitate the statistical modelling of time-dependent changes in the outcome variable and the strength of associations between variables, and are ideal for longitudinal studies involving repeated measures [43]. The STATA® SE statistical software, version 16.1 (StataCorp, College Station, TX, USA) was used for all analyses.
The decline rates in the cognitive domains and global cognition were compared using β coefficients and their 95% confidence intervals (CI). The intercept represents the differences in the average z-scores of global cognition and the cognitive domains between the groups with vitamin D insufficiency and deficiency and the group with sufficient vitamin D at baseline. The slope, i.e., time (in years), indicates the magnitude of the trajectory of change in global cognition and cognitive domains independently of the covariates (as if time itself were the determinant of decline). The time-vitamin D status interaction corresponds to the annual decline rate of global cognition and the cognitive domains in the groups with vitamin D insufficiency and deficiency compared to the group with vitamin D sufficiency.
Ethical aspects
The ELSA Study received approval from the London Multicentre Research and Ethics Committee (MREC 01/2/91). All participants signed a statement of informed consent.
Results
Among the 2625 participants at baseline, 2205 and 1989 were reassessed after four and six years, respectively. Approximately 84% of the sample participated in two waves, and 75.8% in three waves. At baseline, the average age of the sample was 70 years, and the prevalence of vitamin D deficiency was 21.9%. The sample characteristics according to vitamin D status are presented in Tables 1 and 2.
Table 1.
Sociodemographic and behavioural characteristics of 2625 participants at baseline according to serum 25(OH)D concentration, ELSA Study (2012)
| Total (n = 2625) |
Sufficiency (> 75 nmol/L) (n = 364) |
Insufficiency (> 30 and ≤ 75 nmol/L) (n = 1686) |
Deficiency (≤ 30 nmol/L) (n = 575) |
|
|---|---|---|---|---|
| Sociodemographic | ||||
| Age, years (SD) | 70.7 (7.4) | 70.8 (7.0) | 70.3 (7.1) | 72.0 (8.1)a,b |
| Sex (women) (%) | 57.4 | 54.1 | 55.8a | 64.0b |
| Race (non-white) (%) | 1.0 | 0.3 | 0.5 | 1.9a,b |
| Wealth (quintiles) (%) | ||||
| Highest quintile | 24.2 | 28.8 | 26.1 | 15.8a,b |
| 2nd quintile | 23.5 | 25.8 | 24.5 | 19.1 |
| 3rd quintile | 22.3 | 20.6 | 22.6 | 22.4 |
| 4th quintile | 16.6 | 13.5 | 15.8 | 20.9a,b |
| Lowest | 13.4 | 11.3 | 11.0 | 21.8a,b |
| Behavioural | ||||
| Alcohol intake (%) | ||||
| Rarely or never | 19.8 | 15.7 | 18.0 | 27.7a,b |
| Frequently | 39.9 | 42.3 | 40.5 | 36.5 |
| Daily | 34.5 | 39.0 | 36.5 | 26.1a,b |
| Did not answer | 5.8 | 3.0 | 5.0 | 9.7a,b |
| Smoking (%) | ||||
| Non-smoker | 36.5 | 36.8 | 37.7 | 32.9 |
| Ex-smoker | 55.4 | 58.0 | 55.8 | 52.5 |
| Smoker | 8.1 | 5.2 | 6.5 | 14.6a,b |
| Physical activity (inactive) (%) | 31.5 | 24.7 | 28.8 | 43.6a,b |
Data expressed as percentage, mean and standard deviation (SD); statistical significance p < 0.05
aSignificantly different from individuals with 25(OH)D sufficiency
bSignificantly different from individuals with 25(OH)D insufficiency
Table 2.
Clinical, anthropometric, and biochemical characteristics of 2625 participants at baseline according to serum 25(OH)D concentration, ELSA Study (2012)
| Total (n = 2625) |
Sufficiency (> 75 nmol/L) (n = 364) |
Insufficiency (> 30 and ≤ 75 nmol/L) (n = 1686) |
Deficiency (≤ 30 nmol/L) (n = 575) |
|
|---|---|---|---|---|
| Clinical conditions | ||||
| Depressive symptoms (%) | 9.0 | 6.9 | 8.2 | 12.9a,b |
| Hypertension (%) | 42.2 | 41.8 | 40.1 | 48.7a,b |
| Diabetes (%) | 10.3 | 7.7 | 10.4 | 11.8 |
| Myocardial infarction (%) | 5.5 | 5.5 | 5.3 | 6.1 |
| Angina (%) | 6.7 | 6.0 | 6.5 | 7.7 |
| Heart failure (%) | 0.5 | 0.3 | 0.5 | 0.5 |
| Stroke (%) | 4.3 | 4.4 | 3.4 | 6.8b |
| Asthma (%) | 10.9 | 8.5 | 10.6 | 13.2a |
| Use of carbamazepine (%) | 1.2 | 0.5 | 1.2 | 1.6 |
| Season of blood sample (%) | ||||
| Summer | 25.9 | 16.5 | 22.7a | 40.2a,b |
| Spring | 5.8 | 3.6 | 5.5 | 8.3a |
| Autumn | 42.6 | 38.7 | 27.1 | 14.1b |
| Winter | 25.7 | 41.2 | 44.7a | 37.4a,b |
| Vitamin D supplementation (%) | 5.3 | 13.2 | 4.9a | 1.6a,b |
| C-reactive protein mg/L (SD) | 3.3 (7.0) | 2.8 (5.4) | 3.1 (6.6) | 4.3 (8.9)a,b |
| Cognitive performance (SD) | ||||
| Temporal orientation z-score | 0.08 (0.9) | 0.11 (0.8) | 0.09 (0.9) | 0.06 (0.9) |
| Executive function z-score | 0.24 (0.8) | 0.15 (0.8) | 0.26 (0.8)a | 0.26 (0.8) |
| Memory z-score | 0.21 (0.9) | 0.32 (0.9) | 0.19 (0.9)a | 0.18 (0.9)a |
| Global cognition z-score | 0.28 (0.8) | 0.25 (0.7) | 0.29 (0.8) | 0.28 (0.8) |
| Serum 25(OH)D level, nmol/L (SD) | 50.4 (23.4) | 91.2 (17.3) | 51.2 (12.4)a | 22.4 (5.3)a,b |
| Anthropometric | ||||
| Waist circumference, (SD) | 95.8 (13.2) | 91.9 (12.5) | 95.6 (12.7)a | 98.5 (14.4)a,b |
| > 102 cm men > 88 cm women (%) | 52.7 | 38.5 | 52.1a | 63.7a,b |
| Biochemical | ||||
| HDL, (SD) | 65.1 (18.4) | 67.9 (17.8) | 65.3 (18.7)a | 62.6 (17.4)a,b |
| < 40 mg/dL men < 50 mg/dL women (%) | 9.9 | 6.3 | 9.7 | 12.7a,b |
| LDL (SD) | 121.5 (40.5) | 114.8 (37.9) | 123.1 (40.2)a | 121.3 (42.5)a |
| ≥ 100 mg/dL (%) | 69.5 | 64.8 | 71.2a | 67.5 |
Data expressed as percentage, mean and standard deviation (SD); statistical significance p < 0.05
aSignificantly different from individuals with 25(OH)D sufficiency
bSignificantly different from individuals with 25(OH)D insufficiency
Compared to participants who had sufficient vitamin D, those with vitamin D deficiency or insufficiency were older, predominantly non-white, possessed less family wealth, consumed less alcohol, smoked more, exhibited greater physical inactivity, experienced a higher frequency of depressive symptoms and hypertension, had lower rates of vitamin D supplementation, displayed elevated levels of C-reactive protein, presented with a larger waist circumference, and had lower HDL levels (Tables 1 and 2).
Compared to the participants with vitamin D insufficiency, those with vitamin D deficiency were older, predominantly female, more likely to be non-white, possessed less family wealth, consumed less alcohol, smoked more, were more physically inactive, exhibited a higher frequency of depressive symptoms, hypertension, and stroke, had a lower rate of vitamin D supplementation, higher levels of C-reactive protein, a larger waist circumference, and lower HDL levels (Tables 1 and 2).
Table 3 displays the estimated parameters for the generalised linear mixed models, including the intercept (baseline) and changes in global cognition and cognitive domains, during the six-year follow-up period as a function of vitamin D status. At the intercept, a significant difference was only found in the executive function domain. At baseline, a better executive function performance was observed among the participants with vitamin D insufficiency (0.103 SDs; 95% CI: 0.010 to 0.197; p < 0.05) and vitamin D deficiency (0.159 SDs; 95% CI: 0.043 to 0.274; p < 0.05) in comparison to those with vitamin D sufficiency.
Table 3.
Adjusted generalised linear mixed models for trajectory of cognitive performance in a six-year period according to serum 25(OH)D concentration in 2625 participants, ELSA Study (2012–2018)
| Estimated β coefficient (95% CI) | |
|---|---|
| (n = 2625) | |
| Model 1—global cognition | |
| Intercept | |
| 25(OH)D sufficiency | Reference |
| 25(OH)D insufficiency | 0.047 (− 0.038; 0.133) |
| 25(OH)D deficiency | 0.098 (− 0.007; 0.204) |
| Slope | |
| Time, years | 0.203 (0.103; 0.304)* |
| Time × 25(OH)D sufficiency | Reference |
| Time × 25(OH)D insufficiency | − 0.014 (− 0.040; 0.012) |
| Time × 25(OH)D deficiency | − 0.037 (− 0.069; − 0.005)* |
| Model 2—memory | |
| Intercept | |
| 25(OH)D sufficiency | Reference |
| 25(OH)D insufficiency | − 0.091 (− 0.186; 0.004) |
| 25(OH)D deficiency | − 0.048 (− 0.164; 0.067) |
| Slope | |
| Time, years | 0.002 (− 0.099; 0.105) |
| Time × 25(OH)D sufficiency | Reference |
| Time × 25(OH)D insufficiency | 0.010 (− 0.016; 0.037) |
| Time × 25(OH)D deficiency | − 0.015 (− 0.047; 0.017) |
| Model 3—executive function | |
| Intercept | |
| 25(OH)D sufficiency | Reference |
| 25(OH)D insufficiency | 0.103 (0.010; 0.197)* |
| 25(OH)D deficiency | 0.159 (0.043; 0.274)* |
| Slope | |
| Time, years | 0.242 (0.137; 0.347)* |
| Time × 25(OH)D sufficiency | Reference |
| Time × 25(OH)D insufficiency | − 0.025 (− 0.053; 0.002) |
| Time × 25(OH)D deficiency | − 0.038 (− 0.071; − 0.004)* |
| Model 4—temporal orientation | |
| Intercept | |
| 25(OH)D sufficiency | Reference |
| 25(OH)D insufficiency | − 0.049 (− 0.154; 0.055) |
| 25(OH)D deficiency | − 0.048 (− 0.175; 0.080) |
| Slope | |
| Time, years | 0.099 (− 0.030; 0.228) |
| Time × 25(OH)D sufficiency | Reference |
| Time × 25(OH)D insufficiency | 0.017 (− 0.015; 0.049) |
| Time × 25(OH)D deficiency | 0.009 (− 0.030; 0.048) |
Cognitive performance on each test is standardised by a z-score. Global cognition is calculated by the average z-scores of cognitive tests and subsequently standardised by the mean. Models adjusted by age, sex, race, wealth, alcohol intake, smoking, physical activity, depressive symptoms, hypertension, diabetes, myocardial infarction, angina, heart failure, stroke, asthma, waist circumference, HDL and LDL cholesterol, use of carbamazepine, season of the blood sample, vitamin D supplementation, and C-reactive protein
*p < 0.05
Over time, participants with vitamin D deficiency experienced a greater decline in global cognition (− 0.037 SDs per year; 95% CI: − 0.069 to − 0.005; p < 0.05) and executive function (−0.038 SDs per year; 95% CI: − 0.071 to − 0.004; p < 0.05) during the six-year follow-up period compared to those with vitamin D sufficiency (Table 3). Clinically, this represents a global cognition decline of − 0.470 SD and an executive function decline of − 0.350 SD over the six years. No significant decline rates were observed for memory or temporal orientation based on vitamin D status. Figure 2 and Table S1 illustrate the predicted annual mean values of the z-scores of global cognition and executive function for the groups with vitamin D sufficiency, insufficiency, and deficiency during the 6-year follow-up period.
Fig. 2.
Trajectory graphs of global cognition and executive function in 2625 individuals over six-year follow-up according to vitamin D status, ELSA Study, 2012–2018
Discussion
The main results of the present study demonstrated that individuals with deficient serum 25(OH)D concentrations are at a greater risk of experiencing a decline in global cognition and executive function over six years compared to those with adequate 25(OH)D concentrations.
Executive function performance at baseline was slightly higher among individuals with vitamin D insufficiency or deficiency compared to those with sufficient levels. Differences in cognitive performance at baseline are common in large observational cohorts. These differences may arise from the influence of factors accumulated throughout life, such as cognitive reserve and the quality of education [44]. Despite the better performance at baseline, participants with deficient 25(OH)D levels exhibited more pronounced cognitive decline over the six-year follow-up than those with sufficient 25(OH)D levels, emphasising the importance of longitudinal follow-up for determining changes over time.
Our findings align with data reported in previous studies. In a longitudinal study with a six-year follow-up involving 858 individuals aged 65 years or older, Llewellyn (2010) [9] found that the group with 25(OH)D deficiency (< 25 nmol/L) experienced a greater decline in global cognition and executive function compared to those with sufficient 25(OH)D (≥ 75 nmol/L) levels. Although the cutoff points for classifying vitamin D status differ, the results align with ours. However, the authors did not use scores standardised in z-scores stratified by age and education. In addition, their models were not controlled for variables influencing 25(OH)D, such as the use of carbamazepine.
Wilson et al. (2014) [10] investigated the longitudinal association between 25(OH)D levels and cognitive performance using data from 2777 participants. The authors found that individuals with 25(OH)D concentrations lower than < 50 nmol/L experienced a greater decline in global cognition (but not executive function) compared to those with sufficient 25(OH)D concentrations (≥ 75 nmol/L) over a 4-year follow-up period. However, the age of the participants was limited to between 70 and 79 years.
Palacios et al. (2020) [3] examined the association between serum 25(OH)D levels and cognitive decline in 967 individuals between 45 and 75 years of age. The authors found no differences in global cognition performance or any cognitive domain over a 2-year follow-up period. The methodological difference may have contributed to the different results, such as the representative sample of a specific population of Hispanics and Puerto Ricans residing in the United States, the non-use of reference values for categorising vitamin D, and the short follow-up period.
From the physiopathological standpoint, vitamin D is locally metabolised by neurons and microglia through VDRs expressed in the brain, which use its active form (1,25-hydroxyvitamin D) to regulate the proliferation, differentiation, and survival of neural cells. Moreover, vitamin D has an anti-inflammatory and neuroprotective role, reducing oxidative stress within neurons. Thus, low 25(OH)D concentrations in the organism can promote brain dysfunction and atrophy [45, 46].
No significant differences were observed in the trajectories of memory or temporal orientation based on vitamin D status in the present study. Analysing the impact of vitamin D deficiency on the trajectory of cognitive performance over a six-year follow-up period, Grajdean et al. (2022) [14] also found no significant associations with the decline in memory or temporal orientation. However, the authors did not explore the trajectories of global cognition or the executive function domain.
The association found between vitamin D deficiency and poorer trajectories of global cognition and executive function, but not memory or temporal orientation, may be explained by the hypothesis that vitamin D is linked to endocrine and cardiovascular factors, which contribute to a higher incidence of non-amnestic forms of cognitive impairment, such as impaired executive functions, processing speed, and language, rather than amnestic forms, which are characterised by impaired memory [47, 48].
Furthermore, a meta-analysis revealed that research has identified a stronger relationship between lower vitamin D concentrations and impaired executive function and processing speed compared to memory [49, 50]. While the hippocampus is associated with memory functioning, evidence suggests that the prefrontal cortex is more related to executive functions [46]. Therefore, the prefrontal cortex may play a greater role in cognitive impairment among individuals with vitamin D deficiency.
While pharmacological treatments for cognitive impairment and dementia have made progress in recent years, non-pharmacological and preventive strategies are still vital for promoting cognitive health, as some medications may not be available to all populations or have limited effectiveness in the early stages [51].
The study’s strengths include utilising a large, representative sample of the English population aged 50 and over, which facilitated the classification of individuals into groups based on their vitamin D status. Additionally, the statistical models were adjusted for a wide range of sociodemographic, behavioural, and clinical variables. Furthermore, the inclusion of data from three waves of the ELSA Study allowed for trajectory analysis over a sufficiently lengthy period. Lastly, the evaluation of cognitive domains and the application of a global cognition measure permitted the identification of the domains most affected by vitamin D deficiency.
This study has several limitations that should be considered, including the losses to follow-up, which are an inevitable source of bias in longitudinal studies. However, this did not prevent us from finding an association between vitamin D deficiency and cognitive decline over the six-year follow-up period. Moreover, generalised linear mixed models were employed, which accommodate all participants with at least one available observation, thereby minimising potential attrition bias. The ELSA Study included only community-dwelling individuals, which does not allow for estimates concerning institutionalised individuals, who tend to have more compromised cognitive functioning. The non-inclusion of creatinine, which was not measured in the ELSA Study, may also be considered a limitation, as elevated creatinine concentrations indicate kidney failure, potentially interfering with the metabolism of 25(OH)D [52]. The use of self-reported data for systemic arterial hypertension and diabetes mellitus may also be regarded as a limitation. However, evidence has demonstrated excellent reliability for such self-reported chronic conditions, particularly in older populations [53].
Based on the results of this study, vitamin D deficiency may be regarded as a risk factor for the decline in global cognition and executive function in individuals aged 50 years or older. This highlights the significance of implementing public health policies centred on the early identification and management of risk factors, as such strategies could promote cognitive health. Our findings may inform future intervention studies that could investigate whether sufficient levels of 25(OH)D are crucial in preventing cognitive decline and its subsequent negative effects. Furthermore, future randomised clinical trials should explore the impact of supplementation, particularly on global cognition and executive function, in individuals with vitamin D deficiency.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
The authors express gratitude to all collaborators and participants of the ELSA Study.
Author Contribution
AS and CO acquired the data. NCS and TSA conceived and designed the study. NCS and TSA analysed and interpreted the data. NCS and TSA wrote the paper. TBPS, MML, PST, MLBD, ROM, and CO revised the manuscript for important intellectual content and contributed significantly to the paper. All authors read and approved the final manuscript.
Funding
This work was supported by the Economic and Social Research Council (ESRC) (grant number ES/T008822/11). This study received support from the Brazilian fostering agencies Coordination for the Advancement of Higher Education Personnel (Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), code 001 grant number 88887.975898/2024–00), National Council of Scientific and Technological Development (Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), grant number 305338/2023–4), and State of São Paulo Research Assistance Foundation (Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP), grant numbers 2024/01918–6 and 2024/08019–7). The funders were not involved in the manuscript. Research reported in this publication was supported by the National Institute on Aging of the National Institutes of Health under Award Number R01AG017644. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. ELSA is funded by the NIHR Policy Research Programme (HEI) 98_1074_03. The opinions expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care.
Data availability
The data are available upon reasonable request. Data from the ELSA Study can be accessed from the UK Data Service for researchers who meet the criteria for accessing confidential data under the conditions of the End User Licence: https://discover.ukdataservice.ac.uk. The data can be accessed at https://www.elsa-project.ac.uk/accessing-elsa-data and requested via the site https://ukdataservice.ac.uk/help/.
Declarations
Ethics approval
Ethical approval was granted by the National Research Ethics Service (London Multicentre Research Ethics Committee [MREC/01/2/91]).
Consent to participate
All ELSA participants submitted a statement of informed consent.
Conflict of interest
The authors declare no competing interests.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Duchaine CS, Talbot D, Nafti M, et al. Vitamin D status, cognitive decline and incident dementia: the Canadian Study of Health and Aging. Can J Public Health. 2020;111(3):312–21. 10.17269/s41997-019-00290-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Prince M, Wimo A, Guerchet M., et al. World Alzheimer Report 2015. The global impact of dementia: an analysis of prevalence, incidence, cost and trends.. [Research Report] Alzheimer's Disease International. 2015. Aailable from: https://www.alzint.org/u/WorldAlzheimerReport2015.pdf
- 3.Palacios N, Scott T, Sahasrabudhe N, Gao X, Tucker KL. Serum vitamin D and cognition in a cohort of Boston-area Puerto Ricans. Nutr Neurosci. 2020;23(9):688–95. 10.1080/1028415X.2018.1545291. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Sultan S, Taimuri U, Basnan SA, et al. Low vitamin D and its association with cognitive impairment and dementia. J Aging Res. 2020;2020:1–10. 10.1155/2020/6097820. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Littlejohns TJ, Kos K, Henley WE, Kuma E, Llewellyn DJ. Vitamin D and dementia. J Prev Alzheimers Dis. Published online 2015:1–10. 10.14283/jpad.2015.68 [DOI] [PubMed]
- 6.Moretti R, Morelli ME, Caruso P. Vitamin D in neurological diseases: a rationale for a pathogenic impact. Int J Mol Sci. 2018;19(8):2245. 10.3390/ijms19082245. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Aspell N, Lawlor B, O’Sullivan M. Is there a role for vitamin D in supporting cognitive function as we age? Proc Nutr Soc. 2018;77(2):124–34. 10.1017/S0029665117004153. [DOI] [PubMed] [Google Scholar]
- 8.Harse JD, Marriott RJ, Zhu K, Murray K, Bucks RS. Vitamin D status and cognitive performance in community-dwelling adults: a dose-response meta-analysis of observational studies. Front Neuroendocrinol. 2023;70:101080. 10.1016/j.yfrne.2023.101080. [DOI] [PubMed] [Google Scholar]
- 9.Llewellyn DJ. Vitamin D and risk of cognitive decline in elderly persons. Arch Intern Med. 2010;170(13):1135. 10.1001/archinternmed.2010.173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wilson VK, Houston DK, Kilpatrick L, et al. Relationship between 25-hydroxyvitamin D and cognitive function in older adults: the health, aging and body composition study. J Am Geriatr Soc. 2014;62(4):636–41. 10.1111/jgs.12765. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Miller JW, Harvey DJ, Beckett LA, et al. Vitamin D status and rates of cognitive decline in a multiethnic cohort of older adults. JAMA Neurol. 2015;72(11):1295. 10.1001/jamaneurol.2015.2115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Laughlin GA, Kritz-Silverstein D, Bergstrom J, et al. Vitamin D insufficiency and cognitive function trajectories in older adults: the Rancho Bernardo Study Brayne. C, ed. J Alzheimers Dis. 2017;58(3):871–83. 10.3233/jad-161295. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Toffanello ED, Coin A, Perissinotto E, et al. Vitamin D deficiency predicts cognitive decline in older men and women: The Pro.V.A. Study. Neurology. 2014;83(24):2292–8. 10.1212/WNL.0000000000001080. [DOI] [PubMed] [Google Scholar]
- 14.Grajdean S, de Oliveira C, Cadar D. Vitamin D and changes in cognitive performance in mid and later life: findings from the English Longitudinal Study of Ageing (ELSA). medRxiv. 2022 Nov 14:2022-11. 10.1101/2022.11.14.22282290
- 15.Mindell J, Biddulph JP, Hirani V, et al. Cohort profile: the health survey for England. Int J Epidemiol. 2012;41(6):1585–93. 10.1093/ije/dyr199. [DOI] [PubMed] [Google Scholar]
- 16.Steptoe A, Breeze E, Banks J, Nazroo J. Cohort profile: the English longitudinal study of ageing. Int J Epidemiol. 2013;42(6):1640–8. 10.1093/ije/dys168. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.American Psychiatric Association: diagnostic and statistical manual of mental disorders - DSM-5, Fifth Edition. Arlington, VA, American Psychiatric Association, 2013. 10.1176/appi.books.9780890425596
- 18.Ofstedal MB, Fisher G. Documentation of cognitive functioning measures in the health and retirement study. Institute for Social Research, University of Michigan; 2005. 10.7826/ISR-UM.06.585031.001.05.0010.2005
- 19.Steffick D. Documentation of affective functioning measures in the health and retirement study. Institute for Social Research, University of Michigan; 2000. 10.7826/ISR-UM.06.585031.001.05.0005.2000
- 20.Cadar D, Abell J, Steptoe A. Cognitive impairment and dementia in older English adults: risk factors and diagnostic algorithms. In: Banks J, Nazroo J, Steptoe A, Zaninotto P. (eds.) The dynamics of ageing: Evidence from the English Longitudinal Study of Ageing 2002-2019 (wave 9). (pp. 62-92). The Institute for Fiscal Studies: London, UK. Available from: https://discovery.ucl.ac.uk/id/eprint/10120739/.
- 21.Banks J, Breeze E, Lessof C, Nazroo J. Retirement, health and relationships of the older population in England: The 2004 English Longitudinal Study Of Ageing (Wave 2). [Book]. Institute for Fiscal Studies: London; 2006. Available from: https://discovery.ucl.ac.uk/id/eprint/15351/.
- 22.Qiao H, Chen M, Li S, Li Y, Sun Y, Wu Y. Poor lung function accelerates cognitive decline in middle-aged and older adults: evidence from the English Longitudinal Study of Ageing. Arch Gerontol Geriatr. 2020;90:104129. 10.1016/j.archger.2020.104129. [DOI] [PubMed] [Google Scholar]
- 23.Seblova D, Berggren R, Lövdén M. Education and age-related decline in cognitive performance: systematic review and meta-analysis of longitudinal cohort studies. Ageing Res Rev. 2020;58:101005. 10.1016/j.arr.2019.101005. [DOI] [PubMed] [Google Scholar]
- 24.Cochar-Soares N, De Oliveira DC, Luiz MM, et al. Sex differences in the trajectories of cognitive decline and affected cognitive domains among older adults with controlled and uncontrolled glycemia. J Gerontol A Biol Sci Med Sci. Published online May 22, 2024:glae136. 10.1093/gerona/glae136. [DOI] [PMC free article] [PubMed]
- 25.Li C, Zhu Y, Ma Y, Hua R, Zhong B, Xie W. Association of cumulative blood pressure with cognitive decline, dementia, and mortality. J Am Coll Cardiol. 2022;79(14):1321–35. 10.1016/j.jacc.2022.01.045. [DOI] [PubMed] [Google Scholar]
- 26.Delinocente MLB, Luiz MM, De Oliveira DC, et al. Are serum 25-hydroxyvitamin D deficiency and insufficiency risk factors for the incidence of dynapenia? Calcif Tissue Int. 2022;111(6):571–9. 10.1007/s00223-022-01021-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Carvalho AC, Santos NC, Portugal-Nunes C, Castanho TC, Moreira P, Costa PS, Sousa N, Palha JA. 25-OH Vitamin D levels and cognitive performance: longitudinal assessment in a healthy aging cohort. Front Aging Neurosci. 2019;27(11):330. 10.3389/fnagi.2019.00330. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Wang TY, Wang HW, Jiang MY. Prevalence of vitamin D deficiency and associated risk of all-cause and cause-specific mortality among middle-aged and older adults in the United States. Front Nutrit. 2023;18(10):1163737. 10.3389/fnut.2023.1163737. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Dickens AP, Lang IA, Langa KM, Kos K, Llewellyn DJ. Vitamin D, cognitive dysfunction and dementia in older adults. CNS Drugs. 2011;25(8):629–39. 10.2165/11593080-000000000-00000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Holick MF, Binkley NC, Bischoff-Ferrari HA, et al. Evaluation, treatment, and prevention of vitamin d deficiency: an endocrine society clinical practice guideline. J Clin Endocrinol Metab. 2011;96(7):1911–30. 10.1210/jc.2011-0385. [DOI] [PubMed] [Google Scholar]
- 31.Amrein K, Scherkl M, Hoffmann M, et al. Vitamin D deficiency 2.0: an update on the current status worldwide. Eur J Clin Nutr. 2020;74(11):1498–513. 10.1038/s41430-020-0558-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Luiz MM, Máximo RDO, De Oliveira DC, et al. Could poor glycaemic control be a predictor of walking speed decline in older adults? Evidence from the English Longitudinal Study of Ageing. Diabetes Obes Metab. 2024;26(6):2349–58. 10.1111/dom.15549. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Wang M, Fan C, Han Y, et al. Associations of modifiable dementia risk factors with dementia and cognitive decline: evidence from three prospective cohorts. Front Public Health. 2025;13:1529969. 10.3389/fpubh.2025.1529969. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Listabarth S, Groemer M, Waldhoer T, et al. Cognitive decline and alcohol consumption in the aging population—a longitudinal analysis of the Survey of Health, Ageing and Retirement in Europe. Eur Psychiatry. 2022;65(1):e83. 10.1192/j.eurpsy.2022.2344. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Weuve J, Kang JH, Manson JE, Breteler MM, Ware JH, Grodstein F. Physical activity, including walking, and cognitive function in older women. Jama. 2004;292(12):1454–61. [DOI] [PubMed] [Google Scholar]
- 36.Craig R, Mindell J, Hirani V. Editors. [Health Survey for England 2008]. Physical activity and fitness: summary of key findings. The health and social care infromation centre. 2008;1-21. Available from: https://digital.nhs.uk/data-and-information/publications/statistical/health-survey-for-england/health-survey-for-england-2008-physical-activity-and-fitness.
- 37.Alexandre TDS, Scholes S, Santos JLF, De Oliveira C. Dynapenic abdominal obesity as a risk factor for worse trajectories of adl disability among older adults: the ELSA Cohort Study. J Gerontol Ser A. 2019;74(7):1112–8. 10.1093/gerona/gly182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Gallagher D, Kiss A, Lanctot K, Herrmann N. Depressive symptoms and cognitive decline: a longitudinal analysis of potentially modifiable risk factors in community dwelling older adults. J Affect Disord. 2016;190:235–40. 10.1016/j.jad.2015.09.046. [DOI] [PubMed] [Google Scholar]
- 39.Tofani PS, Máximo RDO, Cochar-Soares N, et al. Does the coexistence of pain and depressive symptoms accelerate cognitive decline? Aging Ment Health. 2025;29(2):334–42. 10.1080/13607863.2024.2392737. [DOI] [PubMed] [Google Scholar]
- 40.National Institutes of Health. Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults-The Evidence Report. Obes Res. 1998;2:51S-209S. [PubMed]
- 41.Fedder DO, Koro CE, L’Italien GJ. New National Cholesterol Education Program III Guidelines for primary prevention lipid-lowering drug therapy: projected impact on the size, sex, and age distribution of the treatment-eligible population. Circulation. 2002;105(2):152–6. 10.1161/hc0202.101971. [DOI] [PubMed] [Google Scholar]
- 42.LoPinto-Khoury C, Brennan L, Mintzer S. Impact of carbamazepine on vitamin D levels: a meta-analysis. Epilepsy Res. 2021;178:106829. 10.1016/j.eplepsyres.2021.106829. [DOI] [PubMed] [Google Scholar]
- 43.Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika. 1986;73(1):13–22. [Google Scholar]
- 44.Zammit AR, Yang J, Buchman AS, et al. Latent cognitive class at enrollment predicts future cognitive trajectories of decline in a community sample of older adults. J Alzheimers Dis. 2021;83(2):641–52. 10.3233/JAD-210484. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Arosio B, Rossi PD, Ferri E, Cesari M, Vitale G. Characterization of vitamin D status in older persons with cognitive impairment. Nutrients. 2022;14(6):1142. 10.3390/nu14061142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Roy NM. Impact of vitamin D on neurocognitive function in dementia, depression, schizophrenia and ADHD. Front Biosci. 2021;26(3):566–611. 10.2741/4908. [DOI] [PubMed] [Google Scholar]
- 47.Hooshmand B, Lökk J, Solomon A, et al. Vitamin D in relation to cognitive impairment, cerebrospinal fluid biomarkers, and brain volumes. J Gerontol Ser A. 2014;69(9):1132–8. 10.1093/gerona/glu022. [DOI] [PubMed] [Google Scholar]
- 48.Mengjie W, Feng J, Feng J. Research progress on the correlation between vitamin d and neurological disorders. Turk Neurosurg Published online. 2021. 10.5137/1019-5149.JTN.34816-21.1. [DOI] [PubMed] [Google Scholar]
- 49.Annweiler C, Montero-Odasso M, Llewellyn DJ, Richard-Devantoy S, Duque G, Beauchet O. Meta-analysis of memory and executive dysfunctions in relation to vitamin D. J Alzheimers Dis. 2013;37(1):147–71. 10.3233/JAD-130452. [DOI] [PubMed] [Google Scholar]
- 50.Gómez-Oliva R, Geribaldi-Doldán N, Domínguez-García S, et al. Vitamin D deficiency as a potential risk factor for accelerated aging, impaired hippocampal neurogenesis and cognitive decline: a role for Wnt/β-catenin signaling. Aging. 2020;12(13):13824–44. 10.18632/aging.103510. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Tripathi P, Lodhi A, Rai S, et al. Review of pharmacotherapeutic targets in Alzheimer’s disease and its management using traditional medicinal plants. Degener Neurol Neuromuscul Dis. 2024;14:47–74. 10.2147/dnnd.s452009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Agarwal R, Hynson JE, Hecht TJW, Light RP, Sinha AD. Short-term vitamin D receptor activation increases serum creatinine due to increased production with no effect on the glomerular filtration rate. Kidney Int. 2011;80(10):1073–9. 10.1038/ki.2011.207. [DOI] [PubMed] [Google Scholar]
- 53.Lee AJ, Sanchez D, Reyes-Dumeyer D, et al. Reliability and validity of self-reported vascular risk factors: hypertension, diabetes, and heart disease, in a multi-ethnic community based study of aging and dementia. J Alzheimer’s Dis. 2023;95(1):275–85. 10.3233/JAD-230374. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data are available upon reasonable request. Data from the ELSA Study can be accessed from the UK Data Service for researchers who meet the criteria for accessing confidential data under the conditions of the End User Licence: https://discover.ukdataservice.ac.uk. The data can be accessed at https://www.elsa-project.ac.uk/accessing-elsa-data and requested via the site https://ukdataservice.ac.uk/help/.