Effect of the replacement of dietary vegetable oils with a low dose of extravirgin olive oil in the Mediterranean Diet on cognitive functions in the elderly

Individuals with cognitive impairments and dementia increase in prevalence exponentially with age, with trends worldwide likely to worsen in ensuing decades [1]. These clinical conditions are associated with high overall costs representing a severe burden to society which need to plan wisely to allocate appropriate resources to meet the demands of the disease. Thus, all non-pharmacologic measures are also important. Several observational studies and randomized controlled trials have shown the protective role of the Mediterranean Diet (MeDiet) against cognitive decline and in decreasing the risk of developing Alzheimer disease (AD) [2‐7].

MeDiet can improve cognitive performances in the elderly [2] and the beneficial effects have been attributed to its high monounsaturated fatty acid (MUFA) content [6] which makes MedDiet a healthy dietary pattern regardless of its high fat content [8]. However, the main concern in the elderly is the change in healthy dietary habits, a reduced interest in food intake and calorie counting, especially in individuals with some degree of cognitive decline [9].

A research suggests that food quality, but not quantity, is a key determinant in order to achieve and maintain a healthy status [10]. Furthermore, there are uncertainties on the notion that increasing MUFA intake per se is effective in improving cognitive functions [11]. In this study, we hypothesized that a change in the quality of vegetable oils would improve the cognitive performances in the elderly better than the quantity. A specific associations between extravirgin olive oil (OO) consumption, and cognitive functions have not still been formally examined, especially at low doses and in the short term.

Thus, our aim was to investigate whether the replacement of all vegetable oils with a lower amount of extravirgin OO, in the contest of a MedDiet, would improve cognitive performances, in the short-term, better than a control MedDiet in a population of elderly individuals.

This study included participants recruited into the study entitled: Effect of the MedDiet on cognitive function in the elderly” carried-out between February 2013 to August 2016, funded by Italian Ministry of Health and whose protocol was approved by the local ethics committee at the “Mater Domini” University Hospital in Catanzaro, Italy (projects codes 2011.48).

The participants were from a Mediterranean area (Calabria region, southern Italy) and were invited to participate in the study by newspapers advertisements. All subjects were white, community-dwelling individuals aged ≥ 65 years and had an MMSE score greater than 20 [12]. Participants were literate and were not suffering from any debilitating diseases (like stage 2–5 chronic kidney disease, end stage liver failure, cancer, congestive heart failure) as ascertained from their medical history, a physical and neurological examination and laboratory tests. They had no previous history of cardiovascular disease (CVD) or thyroid dysfunction or excessive alcohol consumption and did not take any dietary supplements, psychotropic drugs. All participants underwent a neuropsychological assessment conducted by an expert neurologist using a medical assessment and the following neuropsychological tests: the Mini Mental State Examination (MMSE) [13‐16] and the Alzheimer’s Disease Assessment Scale-Cognitive sub-scale (ADAS-cog) [17‐19].

Then 180 participants were randomly assigned to one of the two dietary treatment groups for 1 year (allocation ratio 1:2): (1) a MedDiet in which all vegetable oils (including olive oil, high-oleic safflower oil, high-oleic sunflower oil, canola oil and hydrogenated vegetable oils) were substituted by extravirgin OO at dose of 20–30 g per day, (2) a control MedDiet alone (Fig. 1).

To improve adherence, in the MedDiet plus extravirgin OO group, extravirgin OO was given for free to participants and provided every 3 months by local oils producers (5 L of extravirgin OO for 3 months by Opipari and Torchia Companies, Calabria; extravirgin OO main characteristics: acidity < 0.8%, polyphenol content 280 ppm).

Participants received intensive oral and written recommendations to increase adherence to the MedDiet, reduce all types of fat but energy restriction was not advised for any of the intervention groups. We performed a longitudinal evaluation lasting 12 months with the ADAS-Cog as the main outcome [18].

The investigation conforms to the principles outlined in the Declaration of Helsinki [20]. Written informed consent was obtained from participants.

Data are reported as mean ± SD. Since an effect size of ADAS-Cog change (ES = mean ADAS-cog difference/baseline SD) of 0.8 has been considered as a large clinically relevant change [32], with 80% power on a two-sided level of significance, a minimum of 25 subjects for each group are required.

A total of 110 individuals completed the study (55 participants/group, Fig. 1). Table 1 shows the baseline characteristics of the study population according with the randomization. The two groups were comparable for age, gender, BMI and calories intake. The mean age was 70 ± 4 years, the MMSE score was 24 ± 1 in both groups (p = 0.63 and 0.72 respectively) and ADAS-cog score was 14 ± 4 and 15 ± 5 in the MedDiet and MedDiet plus extravirgin OO groups, respectively (p = 0.14). Furthermore, all other baseline characteristics were not significantly different between groups.

The baseline participants’ nutritional intake is showed in Table 2. Energy intake was not significantly different between groups. Only alcohol and cheese consumption were significantly different between groups (alcohol: 9 g ± 12 and 5 g ± 8 in the MedDiet and MedDiet plus extravirgin OO groups, respectively, p = 0.037; cheese: 48 g ± 39 and 66 g ± 50 in the MedDiet and MedDiet plus extravirgin OO groups, respectively p = 0.038 respectively).

The findings of the present study show a higher reduction of ADAS-cog scores (improved test) after 1 year in the elderly of the MedDiet plus low dose of extravirgin OO group than that observed with a MedDiet alone (− 3.0 ± 0.4 Vs. − 1.6 ± 0.4 respectively; Table 4). This result suggests that, in the short term and despite a low dose, extravirgin OO could be involved in improving the ADAS-cog scores whereas a MedDiet alone slightly preserve cognitive functions. In this study, the two groups had a similar calories and food groups intake at baseline as well as at the second time point (Tables 1, 2, 3), except for alcohol and cheese (at baseline) and carbohydrates, fruit and milk intake (after 1 year). After adjusting the ADAS-cog score variation for these food groups, a higher improvement of ADAS-cog scores was, again, observed in participants of the MedDiet plus low dose of extravirgin OO than in participants at MedDiet alone.

The beneficial effects of the MedDiet on cognitive functions have been attributed to its high MUFA content [6]. However, several evidences raised uncertainty on this association. In the EPIC-Greek cohort, a population of adults aged 60 years old or more evaluated for a period lasting 8 years, MUFA were not associated with MMSE score [11]. A finding from a cross-sectional study reported that an increased consumption of extravirgin OO were independently related to better cognitive functions [33]. However, a specific associations between extravirgin OO intake and cognitive functions or dementia in the short term and with low doses have not been examined yet. Several interventions studies suggest a protective role for OO or extravirgin OO on brain functions over a long period of time. In this regard, one study carried out in a population of community-dwelling individuals at high vascular risk, demonstrated that a MedDiet supplemented with EVOO (1 L/week) was associated with a better global cognitive performance after 6.5 years of follow-up compared with a control low-fat diet [2]. In this study, mean MMSE and Clock Drawing Test (CDT) scores were significantly higher for participants allocated to the MedDiet supplemented with extravirgin OO group in comparison with the control group. However ADAS-cog score was not assessed. In the EPIC—Greek study it has been found that olive oil consumption was weakly positively associated with MMSE score, whereas the association between extravirgin OO and cognition was no tested [11]. Thus, our investigation is original from the point of view of the duration of the study, quality of the OO (only extravirgin) and doses tested. Our results are consistent with that found by Violi et al. [34] Compared with baseline, in healthy subjects of both gender given a Mediterranean-type lunch containing only 10 mg of extravirgin OO, a significant less increase of blood glucose, LDL-Cholesterol and ox-LDL and a more marked increase of blood insulin were detected [34]. It is well accepted that cardiovascular risk factors such as diabetes, dyslipidemia, hypertension and conditions such as insulin-resistance are associated with the development of cognitive impairments [35, 36]. Thus, we assumed that extravirgin OO, even at low dose, may positively affect cognitive functions. Furthermore, in November 2004, the US Federal Drug Administration (FDA) allowed a claim on olive oil labels concerning “the benefits on the risk of coronary heart disease of eating about 23 g of olive oil daily, due to the MUFA in olive oil” [37]. In our study we did find any significant effect on the lipids (data not shown). Nevertheless, since ox-LDL and insulin were not assessed, we cannot rule out these biological effects. In addition, 1 year may be not sufficient to identify cognitive improvements related to reduction of cardiovascular risk factors. We hypothesized that extravirgin OO supplementation rapidly influence cognitive function tests, even at low doses [2, 34‐36] whereas the MedDiet may prevent cognitive decline over a long period of time and have a beneficial effect during the long prodromal phase of dementia [38].

EVOO is the best quality oil, produced by mechanically pressing ripe olives, and retains most of its lipophilic components (including alpha-tocopherol, beta-carotene, and phenolic flavonoid compounds) with strong antioxidant properties [39]. In contrast, common olive oil is a mixture of virgin and refined oil, with fewer antioxidant and anti-inflammatory compounds [38]. An industrial solvent extraction of oil from plants involves the majority of the vegetal oils. Thus, while many vegetable oils may contain MUFAs, the industrialized processing of these oils makes them a less desirable choice for health. For these reasons, it is plausible that a low dose (~ 26 g) of extravirgin OO may have different effects on central nervous system from other vegetal oils, as also suggested by some experimental studies. Compared with virgin OO, refined OO is less protective on oxidative damage to lipids, free radical generation and inflammatory activity [40, 41]. Oleuropein is generally the most prominent phenolic compound in olive cultivars, thus, the neuroprotective effects may be mediated by oleuropein [42].

Of course, our study was not designed to investigate which component of OO (oleic acid or phenols) underlie the association with cognitive functions. Since it has been demonstrated that food quality is a key determinant in order to maintain a healthy status [10], our study suggests that a dietary pattern including extravirgin OO, even at low dose, have a protective action on cognitive functions. Several researches have provided evidence of diet-induced changes in cerebral neurotransmitters [43, 44]. Thus, our study may reveal a new approach to prevent the cognitive decline by changing the quality of vegetable oils in the diet.

In this study, as expected, both groups showed an improvement. Our control MedDiet diet improved cognitive functions in line with previous epidemiological and intervention studies [8], possibly through its effect on oxidative stress and inflammation in the brain [32, 39]. However, a higher reduction of ADAS-cog scores was found in the participants of the MedDiet plus extravirgin OO group than that observed with a MedDiet alone (Table 5).

In our population, no indication about calories restriction was given. However, energy intake decreased by about 300 kcal over 1 year. We hypothesize an age-related reduction in energy intake, that is largely a physiologic effect of healthy aging [9]. Furthermore, physical factors such as poor dentition or age-associated changes in taste and smell may limit the type and quantity of food eaten in older people [9].

In this investigation some weaknesses and strengths must be pointed out. It has been suggested that a mean ADAS-Cog change in individuals judged to have clinically relevant change was over three points with an effect size (ES = mean ADAS-cog difference/baseline SD) of more than 0.5 for a minimal clinically relevant change, with values for those not undergoing a clinically significant change being smaller (0.2–0.4) [32]. Thus, in our study in both groups the change seems to be clinically relevant. However, at present, a cut-off point on the ADAS cog that accurately classifies patients in respect of their clinical response is not universally accepted, especially with a dietary intervention. Consequently, caution must be exercised when interpreting our results on the basis of ADAS-cog score change.

To assess how close the food intakes of our population groups were from the Mediterranean Dietary pattern we used MAI, by dividing the sum of the percentage of total energy from typical Mediterranean food groups by the sum of the percentage of total energy from non-typical Mediterranean food groups [30]. The lack of a significant change in the MAI after 1 year was not a surprise since it has been demonstrated that, even though adherence to Mediterranean Diet did not change over time, the consumption of some foods can change toward an healthier food pattern [45]. In addition, our findings are in line with that demonstrated by Fidanza et al. in the elderly [30].

The exclusion of the elderly affected by CVD makes these results not generalizable to a community-dwelling older population but these observations are applicable only to a population with similar characteristics. However, we restricted to individuals without CVD and 1 year of follow-up to reduce the number of potential intermediary events altering the association between the intervention and the studied outcomes. Longer follow-up in the elderly with CVD may be associated with more intermediary events potentially biasing interpretation of the results [46]. In this regard, the rate of drop out during follow-up was of 62% in the MedDiet group. There are no universally agreed criteria for acceptable follow-up rates in nutritional randomized control trials (RCTs) or cohort studies. However, in RCTs, typically investigating drugs, a cut-off of 80% is used in Evidence-Based Medicine (EBM) ‘‘Levels of Evidence’’ to separate ‘‘high’’- and ‘‘low’’-quality randomised trials [47, 48].

They are important points to mention in interpreting the credibility of these findings. Protocol deviations are very common in interventional research. In this study the overall drop-out rate was 53% and it can have implications on the validity and generalizability of research findings. Of course the method of handling data may lead to different conclusions and it is possible that, in this study, the causal relationship cannot be completely established. However, we explored differences in the characteristics of the participants who completed the study versus those who withdrew (Additional file 1: Table S1) and did not differ significantly, suggesting that the drop-out rate did not impact the final results. We performed a common types of sensitivity analyses to assess the robustness of the results in which participants who violate the protocol (missing appointments; discontinued intervention) were excluded from the analysis. Furthermore, while the overall drop-out rate was 53%, it did not affect power because 25 participants each group were required.

Finally, since it is difficult to disentangle OO from the other components of the MedDiet, and aimed to improve adherence to the protocol, we designed this randomized study in which, in the MedDiet plus extravirgin OO group, extravirgin OO was given for free to participants by local oils producers. We believe this is a strength of the study. Further research is needed to establish whether a Mediterranean Diet supplemented with a low dose of extravirgin OO can prevent dementia.

In this study, we demonstrated that the higher improvement of cognitive functions scores, in a short-term, was observed in individuals of the MedDiet plus low dose (~ 26 g) of extravirgin OO rather than MedDiet alone. In the elderly, a change in the quality of vegetable oils would improve the cognitive functions better than the quantity. Additional studies are required to confirm this results.