Two-Year Changes in Physical Activity and Concurrent Changes in Cognitive Function in a Cohort of Adults with Metabolic Syndrome

Abstract

Background:

It has been proposed that physical activity (PA) could prevent cognitive decline.

Objective:

To evaluate the association between changes in PA and changes in cognitive function in a cohort of adults with metabolic syndrome.

Methods:

Longitudinal observational study including 5,500 adults (mean age 65 years, SD = 5; women = 49.3% ) with metabolic syndrome. Participants underwent physical activity measurements and cognitive evaluation at baseline and at two-years of follow-up. PA was quantified using the Minnesota questionnaire-shortened version. Cognitive function was evaluated using a battery of tests: Mini-Mental Test Examination, Clock Drawing Test, Trail Making Test A and B, Verbal Fluency Test, and Digit Span. The primary outcome was two-year change in cognition, measured through the Global Composite Score (GCS) of all neuropsychological tests. Multivariable-adjusted linear regression models were fitted with baseline PA and their changes as the main exposures and changes in cognitive function as the outcome.

Results:

No significant association was found between PA levels (or their changes) in the GCS of cognitive function. A greater increase in PA levels was associated with a more favorable two-year change in the Trail Making Test A (Q4 versus Q1: b = – 2.24s, 95% CI –4.36 to –0.12s; p-trend = 0.020). No significant association was found for other neuropsychological test.

Conclusion:

Our results do not support an association between increases in PA and the evolution of the global cognitive function at two-year in an intervention trial which included PA promotion in one of its two randomized arms, but they suggested a possible beneficial effect of PA on attentional function in older adults.

Keywords

Alzheimer’s disease cognitive function dementia metabolic syndrome physical activity prevention

INTRODUCTION

The global number of individuals who lived with dementia was 43.8 million in 2016, a figure which more than doubled from 1990 [1]. A variety of neuropathologies underlie dementia syndromes, but Alzheimer’s disease (AD) accounts for 60–80% of all cases [2]. AD develops over a long preclinical period of several decades and mild cognitive impairment (MCI) is its prodromal stage [3]. Despite the declining incidence rate of dementia in Europe and North America over the past 25 years [4], probably due to improvement in living conditions, education, and healthcare [5], several studies suggest that the number of people affected with dementia will increase to 66 million by 2030 [6] and to 152.8 million by 2050 [1].

Therefore, dementia will constitute an increasing challenge to health-care systems worldwide until breakthroughs are made in prevention or curative treatment.

Thus, identifying risk factors which may contribute to its prevention, is of utmost importance and some of them are potentially modifiable [7]. A Lancet Commission Report suggested a new model for modifiable risk factors and found that around 40% of dementia cases are attributable to a combination of twelve risk factors: lower educational level (only to a maximun of age 11-12 years), midlife hypertension, midlife obesity, hearing loss, late-life depression, diabetes, physical inactivity, smoking, social isolation, excessive alcohol consumption, head injury, and air pollution [8].

Available evidence for the effect of lifestyle changes on cognitive decline is controversial although there is a growing evidence that physical activity could prevent dementia [9]. The World Health Organization (WHO) recommends PA to adults with normal cognitive function with the aim of reducing their risk of cognitive decline, with a moderate quality of evidence. However, in the case of adults with mild cognitive impairment, the quality of the available evidence is low [10]. A meta-analysis of randomized clinical trials (RCTs) [11] reported that exercise did not improve cognition in healthy older adults, and potential benefits were only limited to specific cognitive domains. Data from RCTs are largely inconsistent, particularly for older adults who are at a greater risk of cognitive decline [12], but in recent years, increasing evidence suggests that preventive lifestyle interventions such as increases in PA levels might result in better outcomes in these at-risk populations [13]. Therefore, we aimed to assess the association between 2-year changes in PA and concurrent changes in cognitive function within the frame of the PREDIMED-Plus trial, in the context of a lifestyle intervention. This work will contribute to a limited body of longitudinal studies on PA and cognition in adults with risk of cognitive decline.

MATERIALS AND METHODS

Study population

We conducted an observational study within the PREDIMED-Plus trial, a 6 year-multicenter, randomized, parallel-group clinical trial for primary prevention of cardiovascular disease through lifestyle modification. Participants enrolled in this study were adults (55–75 years old men and 60–75 years old women) with overweight/obesity (body-mass index (BMI)> = 27 and < 40 Kg/m²) who met the criteria of metabolic syndrome according to the International Diabetes Federation and the American Heart Association [14], had preserved cognitive function to understand and give consent, and were not institutionalized. Participants were randomly assigned in a 1 : 1 ratio to one of two groups: an intensive weight-loss intervention group, based on an energy-reduced Mediterranean Diet (MedDiet), individualized PA promotion, and behavioral support; or a control group, which included promotion of an unrestricted-energy MedDiet. The study design, methods and cohort profile have been previously reported in detail elsewhere [15, 16].

From October 2013 to December 2016, 6874 participants were recruited from 23 Spanish centers. After exclusions, the final analytical sample was 5500 participants (Fig. 1).

Fig. 1

Flow-chart of participants. ^aPA, physical activity. ^bResiduals with PA as dependent variable and age, sex, RAPA^c and chair test as independent variable to identify outliers. ^cRapid Assessment of Physical Activity (RAPA).

Measurements

At baseline, a general questionnaire was used to collect information regarding socio-demographic variables, smoking status, medical history, use of medication and highest attained educational level. Afterwards, these variables were assessed according to two-year changes in PA.

Exposure

For PA, all participants completed the Rapid Assessment of PA (RAPA) [17] and the Minnesota-REGICOR Short Physical Activity questionnaire [18] at baseline and yearly thereafter. The main construct of this questionnaire covers all four dimensions of PA: type of activity, frequency, duration, and intensity and lists 6 types of activities [19]. The Metabolic equivalents (METs) were assigned to each activity according to the Compendium of Physical Activities [20]. A MET is defined as the resting metabolic rate, that is, the amount of oxygen consumed at rest, sitting quietly in a chair, approximately 3.5 ml o₂/kg/min (1.2 kcal/min for a 70 kg person) [21]. Subsequently, the number of METs for each activity was multiplied by the minutes dedicated per week, obtaining the total of METs-min/week. Total energy expenditure in PA was quantified by adding the total METs-min/day of all the activities carried out by each participant.

In addition, the 30-s chair-stand-test was repeated with the same frequency in order to evaluate their physical fitness [22].

Outcome

Neuropsychological evaluation of participants consisted of a battery of six cognitive tests, which were administered at 2 time points: before randomization (at the third screening visit) and after two years of follow-up. The Mini-Mental State Evaluation (MMSE) [23] and the Clock-Drawing Test (CDT) [24, 25] were used to assess global cognitive performance. In addition, cognitive performance on specific cognitive domains was assessed by the following tests: the Trail Making Test A (TMT-A) and Trail Making Test B (TMT-B) to measure speed of attention and divided attention; the Digit Span test to assess attention and working memory; and the Semantic and Phonemic verbal fluency tasks test to evaluate verbal ability and executive control (further information in the Supplementary Material).

We created a Global Composite Score (GCS) based on the sum of the Z scores from the individual tests, with higher scores denoting better performance. The Z score of TMTA and TMTB was calculated with the inverse score.

All instruments included in the cognitive battery have been standardized for the Spanish population within the age range of the study [25 –27]. Trained study personnel administered face-to-face these tests.

We excluded 68 participants because their PA levels were beyond percentile 99; 120 participants because their residual levels of baseline PA were< –1.96 SD or > 1.96 SD in a model which included age, sex, RAPA score and performance in the chair-test; 763 participants because of missing values in PA at year 2; 59 participants because their PA levels at year 2 were beyond percentile 99; 188 participants because their residual values of PA at year 2 were< –1.96 SD or > 1.96 SD in a model which included age, sex, RAPA score and performance in the chair-test; 10 participants because of inconsistent values in self-reported PA and RAPA questionnaires; and 166 participants because they were not of Spanish nationality. We further excluded 166 participants only for the MMSE, because their baseline values were < 24 and also participants with missing information in the baseline or in the 2-year cognitive tests: 329 for the MMSE, 344 for the CDT, 203 for the TMT-A, 217 for the TMT-B, 198 for the Phonologic fluency and Semantic fluency, 954 for Digit forward and 955 for Digit backward.

Other covariates

A 17-item diet questionnaire was used to assess adherence to the energy-reduced MedDiet [28] and a 143-item food frequency questionnaire [29] to measure energy intake. In addition, for the psychological evaluation, the Beck Depression Inventory (BDI-II) was completed by the participants in the screening period prior to actual enrolment [30]. Furthermore, regarding sleep, participants reported their average daily sleeping time.

Statistical analysis

Participants were grouped into quartiles according to their baseline PA (range, METs-min/wk. Q1:≤867.13; Q2:>867.13 to≤ 1876.46; Q3:>1876.46 to≤3356.64; Q4 > 3356.64 in the main analyses) and two-year changes in PA (range, METs-min/wk. Q1:≤– 614.22; Q2:> – 614.22 to≤391.61; Q3:>391.61 to≤1678.32;Q4:>1678.32 in the main analyses). The first quartile was considered as the reference category.

To analyze the association between baseline PA or 2-year changes in PA and changes in cognitive function, we conducted multiple linear regression models. The primary outcome was the change in the GCS of the neuropsychological battery from baseline to 2-years of follow-up. Two-year changes in the performance on each individual cognitive test were considered as secondary outcomes. Crude and multivariate models adjusted for several potential confounders were conducted as follows: Model 1 was adjusted for intervention group, sex, age, smoking habit (former/current), educational level (primary or less/secondary/university studies), civil status (single or religious/married/widowed/separated or divorced), diabetes, hypertension, BMI, BDI-II, and recruitment center; Model 2 was additionally adjusted for total energy intake kcal/d, alcohol intake g/d, adherence to the energy-restricted Mediterranean diet 0–17, sleeping hours/d and use of medications (statins, antihypertensive medication, antidiabetic agents, aspirin) and vitamins or minerals; Model 3 adjusted for weight (quartiles), adherence to an energy-reduced Mediterranean diet (quartiles) and sleeping hours changes (quartiles).

As supplementary analyses, we assessed how changes in blood pressure are associated with changes in cognitive function. We also evaluated how changes in physical activity are associated with changes in systolic and diastolic blood pressure, at baseline and at two years. We adjusted our analysis for changes in systolic and diastolic blood pressure by creating a third multivariable model. In addition, we further adjusted our models for changes in weight (quartiles), adherence to an energy-restricted Mediterranean diet (quartiles) and sleeping hours changes (quartiles).

The p for trend across different categories of baseline PA or 2-year changes in PA was calculated by assigning the median value to each category and treating the resulting variable as continuous.

In addition, we also conducted sensitivity analyses including only those subjects with a baseline MMSE score≥27 and addressing the likelihood of an MMSE score < 27 after 2 years of follow up with logistic regression models including the same potential confounders as in the aforementioned linear regression models. Additionally, we calculated the residuals for the change in the GCS according to quartiles of baseline and 2-year changes in PA. Also, we created an optimal PA value and calculated changes in the GCS according to it. Optimal PA was defined as staying in the 4th quartile of PA along follow-up or increasing the PA in at least 1 quartile over the two-year follow-up period. Finally, we tested the interaction between quartiles of PA or optimal PA and sex, intervention group and number of metabolic syndrome criteria on the GCS with the likelihood ratio test.

Statistical analyses were performed with STATA version 14.0 (STATA Corp., TX, USA). All p values were two tailed and a p value < 0.05 was considered as statistically significant.

The database used for this analysis was 202101111708_PREDIMEDplus_2021-01-11.

RESULTS

Baseline characteristics of participants according to quartiles of baseline PA are shown in Table 1. Out of 5500 participants, 50.7% were male and 49.3% female. The mean age of participants at recruitment was 65 (SD 5) years. Compared to subjects with a low score (≤848.4) of METs-min/week, those with a high score (>3,356.6) were more likely to be male, to be married, to refrain from smoking, to not suffer diabetes, to have a lower BMI and weight, a higher adherence to the energy-reduced MedDiet, lower levels of depression and lower use of medications (antihypertensive medication, insulin, antidiabetic agents, and aspirin), vitamins, or minerals.

Table 1

Baseline characteristics of the participants

	Quartiles of baseline physical activity
Range, METs-min/wk	≤848.5	>848.5–≤1867.4	>1867.4–≤3356.6	>3356.6
n	1,379	1,371	1,426	1,324
Women, %	57.8	52.2	49.2	37.6
Age, years	65 (5)	65 (5)	65 (5)	65 (5)
Smoking habit, %
Former	37.2	41.4	41.9	49.6
Current	14.7	10.5	13.0	11.4
Educational level, %
Primary or less	51.2	50.3	51.3	48.3
Secondary	27.8	28.4	28.5	30.1
University studies	21.0	21.3	20.2	21.6
Marital status, %
Single or religious	5.3	4.8	4.3	4.8
Married	74.9	74.8	77.8	80.4
Widowed	11.6	11.7	11.9	8.0
Separated or divorced	8.2	8.7	6.0	6.8
Body mass index, kg/m²	33.3 (3.6)	32.6 (3.5)	32.2 (3.3)	31.9 (3.2)
Weight, kg	87.7 (13.5)	85.9 (12.9)	85.4 (13.2)	86.2 (12.6)
Self-reported diabetes at baseline, %	34.0	30.9	28.9	28.1
Systolic blood pressure	138.5 (17.5)	138.9 (16.5)	140.4 (16.6)	140.5 (16.7)
Diastolic blood pressure	80.4 (9.9)	80.6 (9.8)	81 (9.7)	81 (10)
Hypertension at baseline, %	83.8	84.8	84.9	83.2
Beck’s depression inventory-II	10.1 (8.3)	8.6 (7.2)	8.0 (7.0)	7.0 (6.6)
Total energy intake, kcal/d	2,379 (645)	2,427 (625)	2,400 (601)	2,446 (629)
Adherence to an energy-reduced Mediterranean diet (0–17 point)^a	8.0 (2.6)	8.2 (2.6)	8.6 (2.7)	9.1(2.6)
Alcohol intake, g/d	10 (15)	11 (15)	11 (15)	13 (17)
Sleeping, h/d	7.0 (1.3)	7.0 (1.2)	7.0 (1.2)	7.1 (1.2)
Use of medications %
Statins	47.4	48.9	47.8	47.5
Antihypertensive	78.4	76.9	80	77.6
Insulin	6.1	4.0	4.1	4.3
Antidiabetics	25.7	22.8	21.6	22.1
Aspirin	17.7	14.3	14.2	14.4
Use of vitamins or minerals %	14.1	12.7	11.9	9.8

^aScore validated by Schröder et al. [28].

Supplementary Table 1 shows the characteristics of participants according to 2-year changes in PA (median 2-year changes in PA 392 METs-min/week). Compared to those who showed a lesser increase in their former PA (≤–656.41 METs-min/week), participants who increased their PA the most (>1,663.8 METs-min/week) were more likely to be male, to refrain from smoking, to not suffer hypertension, to have lower depression levels and lower use of medications, vitamins, or minerals.

After adjustment for potential confounders, we did not find a significant association between baseline PA or two-year changes in PA and the evolution of overall cognition as measured by the GCS after 2-year follow-up (Table 2). No substantial changes were observed when we further adjusted for weight (quartiles), adherence to an energy-restricted Mediterranean diet (quartiles) and sleeping hours changes (quartiles). We neither found significant associations when we only included subjects with baseline MMSE score≥27 and addressed the likelihood of showing an MMSE score < 27 after 2 years of follow-up (results not shown). However, on specific cognitive domains, a greater increase in PA levels was associated with more favorable two-year changes in TMT-A test performance (Q4 versus Q1: b = –2.24s, 95% CI –4.36 to –0.12s; p-trend = 0.020) (Supplementary Table 3). No significant association was found between baseline or two-year changes in PA and changes in performance in the rest of the tests (Supplementary Tables 2 and 3). We also did not find a significant association between residual changes in the GCS according to quartiles of baseline or two-year changes in PA (Table 3).

Table 2

Multivariable-adjusted mean differences in two-year changes in the Global Composite Score according to quartiles of baseline physical activity or to quartiles of 2-year change in physical activity

Differences in the Global Composite Score according to quartiles of baseline physical activity
Range, METs-min/wk	≤867.13	>867.13 to≤1,878.79	>1,878.79 to≤3,356.64	>3,356.64
n	1,068	1,060	1,090	1,038	p for trend
Crude model	0 (ref)	–0.08 (–0.31 to 0.14)	0.10 (–0.12 to 0.33)	0.12 (–0.11 to 0.35)	0.135
Multivariable model 1	0 (ref)	–0.08 (–0.30 to 0.15)	0.13 (–0.10 to 0.36)	0.10 (–0.13 to 0.34)	0.197
Multivariable model 2	0 (ref)	–0.08 (–0.31 to 0.15)	0.13 (–0.10 to 0.36)	0.10 (–0.14 to 0.34)	0.192
Multivariable model 3	0 (ref)	–0.09 (–0.32 to 0.14	0.12 (–0.11 to 0.35)	0.09 (–0.15 to 0.33)	0.231
Differences in the Global Composite Score according to quartiles of 2-year change in physical activity
Range, METs-min/wk	≤–614.22	>–614.22 to ≤ 391.61	>391.61 to≤1,678.32	>1,678.32
n	1,064	1,067	1,064	1,061	p for trend
Crude model	0 (ref)	0.01 (–0.22 to 0.24)	–0.11 (–0.33 to 0.12)	0.01 (–0.22 to 0.24)	0.935
Multivariable model 1	0 (ref)	0.07 (–0.16 to 0.30)	–0.04 (–0.27 to 0.18)	0.03 (–0.20 to 0.26)	0.958
Multivariable model 2	0 (ref)	0.06 (–0.17 to 0.29)	–0.04 (–0.27 to 0.19)	0.03 (–0.20 to 0.26)	0.943
Multivariable model 3	0 (ref)	0.07 (–0.16 to 0.29)	–0.04 (–0.27 to 0.19)	0.04 (–0.20 to 0.27)	0.894

Multivariable model 1: adjusted for intervention group, sex, age (quartiles), smoking habit (3 categories), educational level (3 categories), civil status (4 categories), diabetes, hypertension, body-mass index (quartiles), Beck’s depression inventory (quartiles), and recruitment center. Multivariable model 2: additionally adjusted for total energy intake (quartiles), alcohol intake (quartiles), adherence to the Mediterranean diet (quartiles) and sleeping hours (quartiles), use of medications (statins, antihypertensive medication, antidiabetic agents, aspirin), and vitamins and minerals use. Multivariable model 3: additionally adjusted for changes in weight (quartiles), adherence to an energy-restricted Mediterranean diet (quartiles) and sleeping hours changes (quartiles).

Table 3

Residual changes in two-year changes in the Global Cognitive Score according to quartiles of baseline physical activity or to quartiles of 2-year change in physical activity

Residual changes in the Global Cognitive Score according to quartiles of baseline physical activity
Range, METs-min/wk	≤867.13	>867.13 to≤1,878.79	>1,878.79 to≤3,356.64	>3,356.64
n	1,068	1,060	1,090	1,038	p for trend
Crude model	–0.03 (–0.18 to 0.11)	–0.05 (–0.19 to 0.10)	0.04 (–0.11 to 0.18)	0.04 (–0.10 to 0.19)	0.372
Multivariable model 1	–0.00 (–0.14 to 0.14)	–0.08 (–0.22 to 0.06)	0.07 (–0.07 to 0.21)	0.01 (–0.13 to 0.16)	0.580
Multivariable model 2	0.00 (–0.14 to 0.15)	–0.09 (–0.23 to 0.05)	0.01 (–0.07 to 0.21)	0.01 (–0.13 to 0.16)	0.603
Residuals changes in the Global Cognitive Score according to Quartiles of 2-year change in physical activity
Range, METs-min/wk	≤–614.22	> –614.22 to≤391.61	>391.61 to≤1,678.32	>1,678.32
n	1,064	1,067	1,064	1,061	p for trend
Crude model	–0.04 (–0.18 to 0.11)	–0.01 (–0.15 to 0.14)	–0.04 (–0.18 to 0.11)	0.08 (–0.07 to 0.22)	0.311
Multivariable model 1	–0.03 (–0.17 to 0.11)	0.03 (–0.11 to 0.17)	–0.02 (–0.15 to 0.12)	0.02 (–0.12 to 0.16)	0.666
Multivariable model 2	–0.03 (–0.17 to 0.11)	0.02 (–0.12 to 0.16)	–0.01 (–0.15 to 0.13)	0.02 (–0.12 to 0.16)	0.662

Likewise, we found no association between an optimal PA at two-years of follow-up and overall cognitive function changes (Table 4).

Table 4

Changes in the Global Cognition Score according to an optimal physical activity over follow up^a

	Suboptimal physical activity over follow-up	Optimal physical activity over follow up
Crude model	–0.07 (–0.18 to 0.04)	–0.07 (–0.19 to 0.06)
Multivariable model 1	–0.06 (–0.17 to 0.04)	–0.08 (–0.20 to 0.05)
Multivariable model 2	–0.07 (–0.17 to 0.04)	–0.07 (–0.20 to 0.05)

^aOptimal physical activity: staying in the Q4 along the two year or improving physical activity in at least one quartile. Multivariable model 1: adjusted for intervention group, sex, age (quartiles), smoking habit (3 categories), educational level (3 categories), civil status (4 categories), diabetes, hypertension, body-mass index (quartiles), Beck’s depression inventory (quartiles), and recruitment center. Multivariable model 2: additionally adjusted for total energy intake (quartiles), alcohol intake (quartiles), adherence to the Mediterranean diet (quartiles) and sleeping hours (quartiles), use of medications (statins, antihypertensive medication, antidiabetic agents, aspirin), and vitamins and minerals use.

No significant interactions were observed between PA and sex or intervention group on the GCS after 2-years of follow-up.

We did not find any significant association between physical activity and cognitive function, additionally adjusting for systolic and diastolic blood pressure (Supplementary Table 4). We also found no significant association between changes in systolic and diastolic blood pressure and changes in cognitive function (Supplementary Table 5), nor between physical activity and systolic/diastolic blood pressure (Supplementary Table 6).

DISCUSSION

In this longitudinal assessment on two-year changes in cognitive function according to levels of PA, we found no significant association between PA levels or their 2-year changes and 2-year changes in global cognition. However, we observed that higher levels of daily PA were independently associated with more favorable two-year changes in attention, as measured with the TMT-A. This test is a measure of sustained attention that requires visual scanning and visuomotor and graphomotor processing speed as main cognitive functions [31]. The task in TMT-B evaluated attention set-shifting and involves greater executive complexity. Therefore, our results suggest that PA may be a positive factor for good attentional functioning in older adults. Some authors have argued that prefrontal cortex, which is of high relevance for attention and executive functioning, might be particularly sensitive to neurophysiological changes induced by exercise [32].

The association between changes in PA and cognitive function is still controversial and overall, scientific evidence that PA reduces dementia risk is inconsistent. Recent guidelines, such as the WHO and Lancet Commission, point to PA as a protective factor against dementia, but in combination with other factors [8] or in populations without cognitive impairment [10]. Most epidemiological studies on the association between PA and cognitive outcomes have been carried out in populations of healthy older adults, and the results cannot be generalized to other populations because individuals with different risk profiles might show different results from similar lifestyle interventions. Our study included older adults without major cognitive impairment or cognitive complaints, but with overweight/obesity and at least three criteria of metabolic syndrome. This syndrome has long been known to be associated with an increased risk of CVD and type 2 diabetes and, in recent years, evidence has accumulated for its association with neurological conditions, such as cognitive impairment and dementia [33]. Reiner et al. summarized existing evidence for the relationship between PA and weight gain, obesity, type 2 diabetes, and dementia [34]. In recent years, increasing evidence shows that individuals with an elevated risk of dementia are those who may obtain greater health benefits from lifestyle interventions [13 , 36]. Therefore, preventive interventions among high-risk participants might result in stronger associations.

Cross-sectional studies have suggested that older adults who reported greater participation in PA showed better cognitive function [35, 37]. These results do not fully coincide with ours, but we must consider that the cross-sectional design does not allow any causal inference to be made from the observed associations, since reverse causality cannot be excluded.

There are several recent prospective cohort studies that indicate that a lack of PA is associated with greater decline in cognitive function and increased risk of dementia [38 –42]. However, a large prospective cohort study with 27 years of follow-up found no evidence of a neuroprotective effect of PA [43] and a recent study found that, among carriers of the APOE ɛ4 allele, an active lifestyle was not associated with beneficial changes in cognitive function [44]. These studies had an observational design and included a homogeneous sample; therefore, they cannot draw conclusions on cause and effect, and cannot be generalized to other populations.

In the past few years, three large multidomain trials, the Finnish Geriatric Intervention Study to Prevent Cognitive Impairment and Disability FINGER [45, 46], the Multidomain Alzheimer's Disease (MAPT) [47], and PreDIVA [48] have been completed. The FINGER trial showed that a two-year multidomain lifestyle intervention was able to benefit cognition in elderly people without baseline cognitive decline, but with an elevated risk of dementia and cardiovascular risk factors [45]. The FINGER trial showed significant intervention effects on a primary outcome (overall cognition based on 14 tests), and in consistency with our study, effects were beneficial on executive functioning (main cognitive secondary outcome). The MAPT trial included people with memory complaints (non-demented), aged 70 years or older, and observed that a multidomain intervention had no significant effects in slowing cognitive decline over 3 years [47]. The PreDIVA study assessed a 6-year multidomain vascular care intervention on dementia prevention among participants aged 70–78 years without baseline dementia. The results did not show an overall decrease of dementia incidence, which was the primary outcome, but in additional analyses, the intervention had a protective effect for non-AD dementia [48]. Our results only slightly differ from the FINGER and PreDIVA studies because we have not found an association between PA and overall cognition measured by the GCS. These inconsistencies could be related to differences in the population of the study and tools used to measure cognitive function and PA and the length of follow-up. In the FINGER study, inclusion criteria were CAIDE Dementia Risk Score≥6 and cognitive performance at the mean level or slightly lower than expected for age. In the preDIVA trial the only exclusion criteria were dementia and other disorders. In contrast, our study excluded participants who did not understand, could not give informed consent, or had communication problems and therefore might have cognitive impairment. Therefore, data from current longitudinal studies included different population profiles and the results cannot be generalized. Furthermore, when interpreting the results of these studies in relation to ours, we must bear in mind that they assessed the effect of a multidomain intervention, and we have only assessed the potential relationship between PA and changes in cognition. Our results have been analyzed as a cohort adjusting for intervention group, but we have not yet analyzed the effect of the multicomponent intervention.

A recent randomized controlled trial [49] with a 4-year follow-up period, investigated the independent and combined effects of resistance and aerobic exercise and dietary interventions on cognition in a general population sample of middle-aged and older individuals. In that study, the combination of regular at least moderate-intensity aerobic exercise with a healthy diet showed a trend toward improved global cognition during 4 years. Overall, PA alone was not associated with cognition. Therefore, their results on PA were consistent with ours. However, the results are difficult to compare because they analyzed a smaller population and without elevated cardiovascular risk. In addition, the follow-up was longer, and they used cognitive tests which are hardly comparable with ours. Additionally, the lack of any effect of PA alone could be explained by the fact that the cohort was already physically active at baseline.

We acknowledge that our study may have some limitations. First, this is a population with overweight/obesity and metabolic syndrome, which may influence their level of PA and their room for change in PA over a two-year period. Second, the inclusion criteria of the trial may limit the generalizability of the results to the general population. Third, information on PA was self-reported, which may lead to some potential misclassification that most probably will be non-differential. Fourth, follow-up lasted for two years, and this may be seen as a not sufficiently long period as to observe changes in cognition and exclude reverse causality. Moreover, no in-person scheduled exercise sessions were carried out and it is possible that this fact might lead to long-term differences.

However, our study also has some strengths, such as the large sample size, longitudinal follow-up, as well as face to face cognitive evaluation with a diverse battery of global and domain-specific normalized cognitive tests. In addition, this study has an extensive and detailed information of participants that has allowed us to adjust for possible confounding factors.

In conclusion, our results did not support an association between PA (or its changes) and the evolution of global cognitive function at two years. However, we found a possible beneficial effect of PA on attention function, which should be addressed in future studies with a longer follow-up period.

Footnotes

ACKNOWLEDGMENTS

We thank all the PREDIMED-Plus participants for their continued cooperation and participation. We also thank the other PREDIMED-Plus personnel for their outstanding support, and the personnel of all associated primary care centers for their exceptional effort. Authors want to thank the invaluable contribution of Centros de Investigación Biomédica en Red: Obesidad y Nutrición (CIBEROBN), Centros de Investigación Biomédica en Red: Epidemiología y Salud Pública (CIBERESP) and Centros de Investigación Biomédica en Red: Diabetes y Enfermedades Metabólicas asociadas (CIBERDEM), which are initiatives of Instituto de Salud Carlos III (ISCIII), Madrid, Spain. Food companies, Hojiblanca and Patrimonio Comunal Olivarero, donated extra-virgin olive oil and Almond Board of California, American Pistachio Growers and Paramount Farms donated nuts.

FUNDING

The PREDIMED-Plus trial was supported by the European Research Council Advanced Research Grand 2014-1019 (340918) granted to Miguel Ángel Martínez-González and by the official funding agency for biomedical research of the Spanish government, ISCIII, through the Fondo de Investigación para la Salud (FIS), which is co-funded by the European Regional Development Fund (five coordinated FIS projects led by Jordi Salas-Salvadó and Josep Vidal, including the following projects: PI13/00673, PI13/00492, PI13/00272, PI13/01123, PI13/00462, PI13/00233, PI13/02184, PI13/00728, PI13/01090, PI13/01056, PI14/01722, PI14/0147, PI14/00636, PI14/00972, PI14/00618, PI14/00696, PI14/01206, PI14/01919, PI14/00853, PI14/01374, PI16/00473, PI16/00662, PI16/01873, PI16/01094, PI16/00501, PI16/00533, PI16/00381, PI16/00366, PI16/01522, PI16/01120, PI17/00764, PI17/01183, PI17/00855, PI17/01347, PI17/00525, PI17/01827, PI17/00532, PI17/00215, PI17/01441, PI17/00508, PI17/01732, PI17/00926, PI19/00957, PI19/00386, PI19/00309, PI19/01032, PI19/00576, PI19/00017, PI19/01226, PI19/00781, PI19/01560, PI19/01332, PI20/01802, PI20/00138, PI20/01532, PI20/00456, PI20/00339, PI20/00557, PI20/00886, PI20/01158), a Especial Action Project for PREDIMED-Plus granted to Jordi Salas-Salvadó, the Recercaixa grant led by Jordi Salas-Salvadó (2013ACUP00194), grants from the Consejería de Salud de la Junta de Andalucía (PI0458/2013; PS0358/2016, PI0137/2018), the PROMETEO/2017/017 grant from the Generalitat Valenciana, the SEMERGEN grant and FEDER funds (CB06/03). Jordi Salas-Salvadó, gratefully acknowledges the financial support by ICREA under the ICREA Academia programme. None of the funding sources took part in the design, collection, analysis or interpretation of the data, or in the decision to submit the manuscript for publication.

CONFLICT OF INTEREST

The authors have no conflict of interest to report.

DATA AVAILABILITY

Data collaboration for PREDIMED-Plus study is guided by the Data Sharing and Management guide. We follow a controlled data collaboration model, using anonymized (de-identified) study data only, for collaborating with approved researchers. Requests are considered by the PREDIMED-Plus Steering Committee. Please see for detail.

The supplementary material is available in the electronic version of this article: .

References

GBD 2019 Dementia Forecasting Collaborators (2022) Estimation of the global prevalence of dementia in 2019 and forecasted prevalence in 2050: An analysis for the Global Burden of Disease Study 2019. Lancet Public Health 7, e105–e125.

GBD 2016 Dementia Collaborators (2016) Global, regional, and national burden of Alzheimer’s disease and other dementias, 1990–2016: A systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol 18, 88–106.

First WHO Ministerial Conference on Global Action Against Dementia. Meeting report, WHO Headquarters, Geneva, Switzerland, 16–17 March 2015. https://apps.who.int/iris/handle/10665/179537. Last updated October 7, 2015. Accessed on September 8, 2022.

Wolters

, Chibnik

, Waziry

, Anderson

, Claudine

, Beiser

, Bis

, Blacker

, Bos

, Brayne

, Dartigues

, Darweesh

SKL

, Davis-Plourde

, de Wolf

, Debette

, Dufouil

, Fornage

, Goudsmit

, Grasset

, Gudnason

, Hadjichrysanthou

, Helmer

, Ikram

, JE , Kern

, Kuller

, Launer

, Lopez

, Matthews

, McRae-McKee

, Meirelles

, Mosley

, Pase

, Psaty

, Satizabal

, Seshadri

, Skoog

, Stephan

BCM

, Wetterberg

, Wong

, Zettergren

, Hofman

(2020) A Twenty-seven-year time trends in dementia incidence in Europe and the United States: The Alzheimer Cohorts Consortium. Neurology 95, e519–531.

, Beiser

, Breteler

MMB

, Fratiglioni

, Helmer

, Hendrie

, Honda

, Ikram

, Langa

, Lobo

, Matthews

, Ohara

, Pérès

, Qiu

, Seshadri

, Sjölund

, Skoog

, Brayne

(2017) The changing prevalence and incidence of dementia over time-current evidence. Nat Rev Neurol 13, 327–339.

Prince

, Wimo

, Guerchet

, Ali

, Wu YT

(2015) World Alzheimer Report 2015. The Global Impact of Dementia: An Analysis of Prevalence, Incidence, Cost and Trends. Alzheimer's Disease International, London.

Norton

, Matthews

, Barnes

, Yaffe

(2014) Potential for primary prevention of Alzheimer’s disease: An analysis of population-based data. Lancet Neurol 13, 788–794.

Livingston

, Huntley

, Sommerlad

, Ames

, Ballard

, Banerjee

, Brayne

, Burns

, Cohen-Mansfield

, Cooper

, Costafreda

, Dias

, Fox

, Gitlin

, Howard

, Kales

, Kivimäki

, Larson

, Ogunniyi

, Orgeta

, Ritchie

, Rockwood

, Sampson

, Samus

, Schneider

, Selbæk

, Teri

, Mukadam

(2020) Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet 396, 413–446.

Liu

, Zhang

, Wang

, Li

, Chang

, Jia

(2022) Effect of physical exercise on cognitive function of Alzheimer’s disease patients: A systematic review and meta-analysis of randomized controlled trial. Front Psychiatry 13, 927128.

10.

World Health Organization. Risk Reduction of Cognitive Decline and Dementia. https://www.who.int/publications/i/item/9789241550543. Last updated January 1, 2019, Accessed on November 6, 2022.

11.

Young

, Angevaren

, Rusted

, Tabet

(2015) Aerobic exercise to improve cognitive function in older people without known cognitive impairment. Cochrane Database Syst Rev, CD005381.

12.

Kelly

, Loughrey

, Lawlor

, Robertson

, Walsh

, Brennan

(2014) The impact of exercise on the cognitive functioning of healthy older adults: A systematic review and meta-analysis. Ageing Res Rev 16, 12–31.

13.

Kivipelto

, Mangialasche

, Ngandu

(2018) Lifestyle interventions to prevent cognitive impairment, dementia and Alzheimer disease. Nat Rev Neurol 14, 653–666.

14.

Alberti

, Eckel

, Grundy

, Zimmet

, Cleeman

, Donato

, Fruchart

, James

WPT

, Loria

, Smith

(2009) Harmonizing the metabolic syndrome: A joint interim statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity. Circulation 120, 1640–1645.

15.

Martínez-González

, Buil-Cosiales

, Corella

, Bulló

, Fitó

, Vioque

, Romaguera

, Martínez

, Wärnberg

, López-Miranda

, Estruch

, Bueno-Cavanillas

, Arós

, Tur

, Tinahones

, Serra-Majem

, Martín

, Lapetra

, Vázquez

, Pintó

, Vidal

, Daimiel

, Delgado-Rodríguez

, Matía

, Ros

, Fernández-Aranda

, Botella

, Portillo

, Lamuela-Raventós

, Marcos

, Sáez

, Gómez-Gracia

, Ruiz-Canela

, Toledo

, Alvarez-Alvarez

, Díez-Espino

, Sorlí

, Basora

, Castañer

, Schröder

, Navarrete-Muñoz

, Zulet

, Angeles García-Rios

, Salas-Salvadó

, and the PREDIMED-Plus Study Investigators (2019) Cohort profile: Design and methods of the PREDIMED-Plus randomized trial. Int J Epidemiol 48, 387–388o.

16.

The Predimed-Plus Trial. Research Plan. Effect of an energy-restricted mediterranean diet, physical activity and behavioral treatment on the primary prevention of cardiovascular disease. https://www.predimedplus.com/wp-content/uploads/2018/11/Protocolo-PREDIMED-Plus_Eng.pdf. Last updated January 1, 2014, Accessed on May 19, 2022.

17.

Topolski

, LoGerfo

, Patrick

, Williams

, Walwick

, Patrick

(2006) The Rapid Assessment of Physical Activity (RAPA) among older adults. Prev Chronic Dis 3, 1–8.

18.

Molina

, Sarmiento

, Peñafiel

, Donaire

, Garcia-Aymerich

, Gomez

, Ble

, Ruiz

, Frances

, Schröder

, Marrugat

, Elosua

(2017) Validation of the regicor short physical activity questionnaire for the adult population. PLoS One 12, e0168148.

19.

Schröder

, Cárdenas-Fuentes

, Martínez-González

, Corella

, Vioque

, Romaguera

, Alfredo Martínez

, Tinahones

, Miranda

, López Estruch

, Bueno-Cavanillas

, Arós

, Marcos

, Tur

, Warnberg

, Serra-Majem

, Martín

, Vázquez

, Lapetra

, Pintó

, Vidal

, Daimiel

, Gaforio

, Matía-Martín

, Ros

, Castañer

, Lassale

, Ruiz-Canela

, Asensio

, Basora

, Torres-Collado

, Garcia-Rios

, Abete

, Toledo

, Buil-Cosiales

, Bullo

, Goday

, Fitó

, Salas-Salvadó

(2018) Effectiveness of the physical activity intervention program in the PREDIMED-Plus study: A randomized controlled trial. Int J Behav Nutr Phys Act 15, 110.

20.

Ainsworth

, Haskell

, Herrmann

, Meckes

, Bassett

Jr. , Tudor-Locke

, Greer

, Vezina

, Whitt Glover

, Leon

(2011) 2011 Compendium of Physical Activities: A second update of codes and MET values. Med Sci Sports Exerc 43, 1575–1581.

21.

Jetté

, Sidney

KBG

(1990) Metabolic equivalents (METS) in exercise testing, exercise prescription, and evaluation of functional capacity. Clin Cardiol 13, 5555–5565.

22.

Jones

, Rikli

, Beam

(1999) A 30-s chair-stand test as a measure of lower body strength in community-residing older adults. Res Q Exerc Sport 70, 13–19.

23.

Folstein

, Folstein SE

(1975) . Mini-mental state. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12, 189–198.

24.

Freedman

, Leach

, Kaplan

, Winocur

, Shulman

, Delis

(1994), Clock Drawing a Neuropsychological Analysis. Oxford University Press.

25.

del Ser Quijano

, García de Yébenes

, Sánchez Sánchez

, Frades Payo

, Rodriguez Laso

, Bartolomé Martinez

, Otero Puime

(2004) Cognitive assessment in the elderly. Normative data of a Spanish population sample older than 70 years. Med Clin 122, 727–740.

26.

Blesa

, Pujol

, Aguilar

, Santacruz

, Bertran-Serra

, Hernández

, Sol

, Peña-Casanova

, Soler

, Zabay

, Riera

, Castellví

, Torner

, Charques

, Toirán

, Manero

, Böhm

, Espel

, Martí

, Meza

, Crespo

(2011) Clinical validity of the “mini-mental state” for Spanish speaking communities. Neuropsychologia 39, 1150–1157.

27.

Peña-Casanova

, Blesa

, Aguilar

, Gramunt-Fombuena

, Gómez-Ansón

, Oliva

, Molinuevo

, Robles

, Barquero

, Antúnez

, Martínez-Parra

, Frank-García

, Fernández

, Alfonso

, Sol

(2009) Spanish multicenter normative studies (NEURONORMA project): Methods and sample characteristics. Arch Clin Neuropsychol 24, 307–319.

28.

Schröder

, Zomeño

, Martínez-González

, Salas-Salvadó

, Corella

, Vioque

, Romaguera

, Martínez

, Tinahones

, Miranda

, López Estruch

, Bueno-Cavanillas

, Alonso Gómez

, Tur

, Warnberg

, Serra-Majem

, Martín

, Vázquez

, Lapetra

, Pintó

, Vidal

, Daimiel

, Gaforio

, Matía-Martín

, Ros

, Lassale

, Ruiz-Canela

, Babio

, Sorlí

, García-Arellano

, Díaz-López

, Fitó

, Castañer

(2021) Validity of the energy-restricted Mediterranean Diet Adherence Screener. Clin Nutr 40, 4971–4979.

29.

Fernández-Ballart

, Piñol

, Zazpe

, Corella

, Carrasco

, Toledo

, Perez-Bauer

, Martínez-González

MÁ

, Salas-Salvadó

, Martín-Moreno

(2010) Relative validity of a semi-quantitative food-frequency questionnaire in an elderly Mediterranean population of Spain. Br J Nutr 103, 1808–1816.

30.

Sanz

, Perdigon

, Vazquez

(2003) The Spanish adaptation of Beck’s Depression Inventory-II (BDI-II): 2. Psychometric properties in the general population. Clín Salud 14, 249–280.

31.

Llinàs-Reglà

, Vilalta-Franch

, López-Pousa

, Calvó-Perxas

, Torrents Rodas

, Garre-Olmo

(2017) The Trail Making Test: Association with other neuropsychological measures and normative values for adults aged 55 years and older from a Spanish-speaking population-based sample. Assessment 24, 183–196.

32.

Mehren

, Luque

, Brandes

, Lam

, Thiel

, Philipsen

, Özyurt

(2019) Intensity-dependent effects of acute exercise on executive function. Neural Plast 2019, 8608317.

33.

Kim

, Feldman

(2015) Insulin resistance as a key link for the increased risk of cognitive impairment in the metabolic syndrome. Exp Mol Med 47, e149.

34.

Reiner

, Niermann

, Jekauc

, Woll

(2013) Long-term health benefits of physical activity – A systematic review of longitudinal studies. BMC Public Health 13, 813.

35.

Coll-Padrós

, León

, Valech

, Ros

, Vidal

, Estruch

, Fitó

, Salas-Salvadó

, Corella

, Molinuevo

, Rami

(2019) Physical activity is associated with better global cognition and frontal function in overweight/obese older adults with metabolic syndrome. Eur Rev Aging Phys Act 16, 23.

36.

Erickson

, Hillman

, Stillman

, Ballard

, Bloodgood

, Conroy

, Macko

, Marquez

, Petruzzello

, Powell

(2019) Physical activity, cognition, and brain outcomes: A review of the 2018 physical activity guidelines. Med Sci Sports Exerc 51, 1242–1251.

37.

Daimiel

, Martínez-González

, Corella , Salas-Salvadó

, Schröder

, Vioque

, Romaguera

, Martínez

, Wärnberg

, Lopez-Miranda

, Estruch

, Cano-Ibáñez

, Alonso-Gómez

, Tur

, Tinahones

, Serra-Majem

, Micó-Pérez

, Lapetra

, Galdón

, Pintó

, Vidal

, Micó

, Colmenarejo

, Gaforio

, Matía

, Ros

, Buil-Cosiales

, Vázquez-Ruiz

, Sorlí

, Graniel

, Cuenca-Royo

, Gisbert-Sellés

, Galmes-Panades

, Zulet

, García-Ríos

, Díaz-López

, de la Torre

, Galilea-Zabalza

, Ordovás

(2020) Physical fitness and physical activity association with cognitive function and quality of life: Baseline cross-sectional analysis of the PREDIMED-Plus trial. Sci Rep 10, 3472.

38.

Palta

, Sharrett

, Deal

, Evenson

, Gabriel

, Folsom

, Gross

, Windham

, Knopman

, Mosley

, Heiss

(2019) Leisure-time physical activity sustained since midlife and preservation of cognitive function: The Atherosclerosis Risk in Communities Study. Alzheimers Dement 15, 273–281.

39.

Hörder

, Johansson

, Guo

, Grimby

, Kern

, Östling

, Skoog

(2018) Midlife cardiovascular fitness and dementia. Neurology 90, e1298–305.

40.

Northey

, Cherbuin

, Pumpa

, Smee

, Rattray

(2018) Exercise interventions for cognitive function in adults older than 50: A systematic review with metaanalysis. Br J Sports Med 52, 154–160.

41.

Zheng

, Xia

, Zhou

, Tao

, Chen

(2016) Aerobic exercise ameliorates cognitive function in older adults with mild cognitive impairment: A systematic review and metaanalysis of randomised controlled trials. Br J Sports Med 50, 1443–1450.

42.

Kurudamannil

, Reddy

(2019) Can healthy diets, regular exercise, and better lifestyle delay the progression of dementia in elderly individuals? J Alzheimers Dis 72, s37–58.

43.

Sabia

, Dugravot

, Dartigues

, Abell

, Elbaz

, Kivimaki

, Singh-Manoux

(2017) Physical activity, cognitive decline, and risk of dementia: 28 year follow-up of Whitehall II cohort study. BMJ 357, j2709.

44.

Fernández-Matarrubia

, Goni

, Rognoni

, Razquin

, Fernández-Lázaro

, Bes-Rastrollo

, Martínez-González

, Toledo

(2021) An active lifestyle is associated with better cognitive function over time in APOE ɛ4 non-carriers. J Alzheimers Dis 79, 1257–1268.

45.

Kivipelto

, Solomon

, Ahtiluoto

, Ngandu

, Lehtisalo

, Antikainen

, Bäckman

, Hänninen

, Jula

, Laatikainen

, Lindström

, Mangialasche

, Nissinen

, Paajanen

, Pajala

, Pel

, Soininen

(2013) The Finnish Geriatric Intervention Study to Prevent Cognitive Impairment and Disability (FINGER): Study design and progress. Alzheimers Dement 6, 657–665.

46.

Ngandu

, Lehtisalo

, Solomon

, Levälahti

, Ahtiluoto

, Antikainen

, Bäckman

, Hänninen

, Jula

, Laatikainen

, Lindström

, Mangialasche

, Paajanen

, Pajala

, Peltonen

, Rauramaa

(2015) A 2 year multidomain intervention of diet, exercise, cognitive training, and vascular risk monitoring versus control to prevent cognitive decline in at-risk elderly people (FINGER): A randomised controlled trial. Lancet 385, 2255–2263.

47.

Andrieu

, Guyonnet

, Coley

, Cantet

, Bonnefoy

, Bordes

, Bories

, Cufi

M-N

, Dantoine

, Dartigues

J-F

, Desclaux

, Gabelle

, Gasnier

, Pesce

, Sudres

, Touchon

, Robert

, Rouaud

, Legrand

, Payoux

, Caubere

J-P

, Weiner

, Carrie

, Ousset

P-J

, Vellas

, MAPT Study

Group

(2017) Effect of long-term omega 3 polyunsaturated fatty acid supplementation with or without multidomain intervention on cognitive function in elderly adults with memory complaints (MAPT): A randomised, placebo-controlled trial. Lancet Neurol 16,, 377–389.

48.

van Charante

EPM

, Richard

, Eurelings

, van Dalen

, Ligthart

, van Bussel

, Hoevenaar-Blom

, Vermeulen

, van Gool

(2016) Effectiveness of a 6-year multidomain vascular care intervention to prevent dementia (preDIVA): A cluster-randomised controlled trial. Lancet 388, 797–805.

49.

Komulainen

, Tuomilehto

, Savonen

, Männikkö

, Hassinen

, Lakka

, Hänninen

, Kiviniemi

, Jacobs

, Kivipelto

, Rauramaa

(2021) Exercise, diet, and cognition in a 4-year randomized controlled trial: Dose-Responses to Exercise Training (DR’s EXTRA). Am J Clin Nutr 113, 1428–1439.