A total of four studies were identified to match our eligibility criteria [15–18]. Three studies [16–18] were randomized controlled trials (RCTs), while one [15] was a case-control study. Three studies were conducted in Spain [15–17] while one was conducted in Malaysia [18]. The general information of the included studies is summarized in Table 2.
Table 2
General information of included studies. RCT: Randomized Controlled Trials.
Authors, year | Study location | Study design | Title | Objectives | Participants | Intervention | Control Group | Cognitive Assessments | Findings |
Yang et al. (2015) [15] | Spain | Case‒control | Coconut oil: alternative nonpharmacological treatment against Alzheimer's disease | To assess coconut oil's effects on Alzheimer's dementia and factors including sex and diabetes | 44 patients with Alzheimer’s disease out of a population of 458 patients evaluated | Administration of 40 ml/day of extra virgin coconut oil to the patients in the case group for 21 consecutive days | Control group on the same feeding pattern as the case group during the study period, but without administration of oil | Mini-Examen Cognitivo (MEC), which is a validated translation into Spanish of the Folstein Minimental Test (MMSE) | Improvement in cognitive abilities, with differences according to sex and disease severity |
La Rubia Ortí Je et al. (2018) [16] | Spain | RCT | Improvement of Main Cognitive Functions in Patients with Alzheimer’s Disease after Treatment with Coconut Oil Enriched Mediterranean Diet: A Pilot Study | To look for improvements in people with AD's primary cognitive function after adopting a Mediterranean diet high in coconut oil | 44 patients with Alzheimer’s disease | Coconut Oil Enriched Mediterranean diet | Control group following a standard low fat diet | Seven Minute Screen (7-MS) | Improvements in episodic, temporal orientation, and semantic memory in AD patients after the diet |
De La Rubia Ortí et al. (2017) [17] | Spain | RCT | How does coconut oil affect cognitive performance in alzheimer patients? | To investigate coconut oil's cognitive effects on Alzheimer's sufferers in many cognitive domains | 44 Patients with Alzheimer’s disease | 40 ml of extra virgin coconut oil daily for 21 consecutive days | Control group did not receive any oil during the 21-day intervention period. | Cognitive Test, Mini-mental State Examination | Improvement in cognitive abilities, with varying intensity depending on the cognitive area |
Chan et al. (2017) [18] | Malaysia | RCT | Effect of cold pressed coconut oil on cognition and behavior among patients with Alzheimer’s disease - A pilot intervention study | To investigate the potential benefits of a dietary intervention for improving cognitive function in AD patients | 40 individuals out of 99 screened | Coconut oil enriched Mediterranean diet | The control group received a placebo consisting of water plus coconut essence. | Cognition score was assessed using MMSE and CDT. | The study found no significant difference in the mean scores in all the parameters assessed between the intervention and control group at baseline, at 3 months and 6 months. |
META-ANALYSIS:
A total of four studies [15–18] were identified according to the eligibility criteria mentioned. The goal was to make use of the various cognitive assessment scores to compare the improvement in memory and cognition of Alzheimer’s disease (AD) patients. Three of the four studies mentioned cognition scores. Out of the 3 remaining studies, one study did not provide comprehensive data for pooling results in the final table but rather stated the significant improvement of pre- and posttest clock drawing test (CDT) scores in the control group (mean CDT − 0.78571, 95% CI CDT: −1.50824 − 0.06319; P = 0.035) [18].
Two studies made use of and reported cognitive assessment scales in a comprehensive manner. One study [15] made use of the MEC (Mini-Examen Cognoscitivo) scale, which is a Spanish translation of the Mini-Mental State Examination (MMSE), while the other [16] evaluated individuals on the basis of the 7-minute screen test. This test uses four individual measures: Benton’s test for temporal awareness, the CDT for visuospatial memory, the Categorical Verbal Fluency test, and the Free and Cued Selective Reminding Test for episodic memory. For the lack of sufficient studies with results that can be pooled, the approach adopted for statistical analysis involved evaluating the results of each of these studies, comparing the percentage improvement in results, and reaching a conclusion about their statistical significance using the paired samples T test.
To derive meaningful results from the two studies, we used the standardized mean differences (SMD) of the two cognitive test scores, namely, the MMSE and the 7-minute screen (7MS). The total score for the MMSE was calculated out of 30 [Hu Yang 2015], with the standard deviation provided for the demographic within the study, allowing for comparison between intervention and control groups.
The 7-Minute Screen scores (7MS) are a combination of four cognitive test scores, as mentioned previously. The total score for the 7MS is calculated in our study of interest [16] out of 156, with a higher score denoting better performance. Since the study did not provide the scores for the entire sample but rather provided individual test scores for the four tests, along with the standard error of the mean (SEM) for each observation, the individual test scores and the SEM were converted into 7MS scores in total, along with the standard deviations calculated from the SEM for each observation.
A forest plot comparing the intervention and control group scores on the MMSE and 7MS was created using the inverse variance method of statistical analysis using the random effects model. Confidence intervals were set at 95%. The mean scores for the intervention and control samples were added, along with the standard deviations, and the standardized mean difference (SMD), also known as Cohen’s d for effect size, was used to compare the two scores. The reason for using SMD in this situation was the use of different scoring systems in the two studies. By comparing the standardized values, we were able to draw meaningful conclusions from the combination of these two studies.
The resultant forest plot is as follows:
Heterogeneity tests of the included studies yielded Tau2 = 0.00, which indicates that there is little variability in the true effects of coconut oil on cognitive scores across the studies included in the meta-analysis. The chi-square test is not statistically significant (Chi2 = 0.08, df = 1, p = 0.78), which suggests that there is no significant heterogeneity among the studies' effect sizes. I2 was also 0%, indicating no observed heterogeneity among the studies. This means that all the studies show consistent results regarding the effect of coconut oil on cognitive scores. The overall effect of coconut oil on Alzheimer's disease was found to be statistically significant (Z = 2.52, p = 0.01), which implies that there is a positive impact of coconut oil on cognitive scores when compared to the control group.
Both studies enrolled a total of 44 patients; consequently, both had similar weights in the final score. The pooled standardized mean difference was 0.55 (95% CI, 0.12–0.97), which showed a medium effect size (according to Cohen’s d interpretation for effect size). With a Cohen's d of 0.55, 70.9% of patients in the intervention group will have a score greater than the mean of those in the control group (Cohen's U3), and the chance that a person picked at random from the intervention group consuming the coconut oil product/coconut oil-rich meal will have a higher score than a person picked at random from the control group (probability of superiority) is 65.1%. Moreover, to have one more favorable outcome in the intervention group compared to the control group, we need to treat 5.4 people on average [19].
MMSE Score:
The maximum score for the MMSE is 30. A score of 25 or higher is classified as normal. If the score is below 24, the result is usually considered abnormal, indicating possible cognitive impairment. For the MMSE scores reported in the study by Hu Yang [15], a 38.92% improvement was observed in the population taking the coconut oil product compared to the placebo, which showed minimal improvement. A paired sample t test was conducted to determine the effect of the coconut product. A significant difference between the MMSE score before intervention (M = 11.61; SD = 6.85) and the MMSE score after intervention (M = 16.13; SD = 7.59) [t (21) = 3.009, p < 0.05] was observed.
Table 3
Mean MMSE scores (case vs control) [15]
Experiment or Placebo | Pretest MMSE | Post-test MMSE |
Experiment | 11.61 | 16.31 |
Placebo | 11.42 | 11.56 |
Similarly, improvements were seen when the averages of the MMSE scores were compared on the basis of their genders, with women showing slightly greater improvement in scores after receiving the treatment (39.70%) compared to men (36.99%). However, the difference in mean scores for men and women was not statistically significant. (t = 1.3192, df = 20, p = 0.202).
Table 4
Mean MMSE scores for the male population (case vs control) [15]
Experiment or Placebo | Pretest MMSE | Post-test MMSE |
Experiment | 14.60 | 20.00 |
Placebo | 11.67 | 12.00 |
Table 5
Mean MMSE scores for the female population (case vs control) [15]
Experiment or Placebo | Pretest MMSE | Post-test MMSE |
Experiment | 10.73 | 14.99 |
Placebo | 11.38 | 11.49 |
CDT scores for Visuospatial Memory:
Mean scores for the CDT were reported as part of the 7-minute screening in the study by Jose Enrique [17]. The patients were scored out of a maximum of 7 points, wherein a higher score was associated with better cognition and visuospatial memory. We analyzed the improvements in performance of those with moderate disease in a descriptive manner while incorporating the entire sample size (i.e., moderate and severe disease patients) in the forest plot analysis.
Table 6
Mean CDT scores for the male population [16]
Experiment or Placebo | Pretest CDT | Post-test CDT |
Experiment | 0.66 | 1.33 |
Placebo | 0.00 | 0.00 |
In contrast, coconut oil-based product administration seemed to worsen the CDT scores in females. This observation showed a 50% decrease in the mean CDT scores in women before and after taking the intervention, as shown below.
Table 7
Mean CDT scores for the female population [16]
Experiment or Placebo | Pretest CDT | Post-test CDT |
Experiment | 1.50 | 0.75 |
Placebo | 1.00 | 1.00 |
Benton’s temporal awareness test:
The scoring for this test is done on the basis of errors from accurate judgment of time, wherein a score of 0 means most errors made for the scoring system adopted in the study by de la Rubia Ortí et al. [16], and a maximum score of 113. A higher score indicates a good prognosis. Evidently, from Fig. 9 and Table 7, the intervention provided a greater benefit than the control, with a 142% score improvement after the intervention. The improvements for the control group were not as substantial.
Table 8
Mean Benton’s Test scores for the male population [16]
Experiment or Placebo | Pretest Benton’s | Post-test Benton’s |
Experiment | 17.33 | 42.00 |
Placebo | 7.00 | 12.00 |
Similarly, in the female population, the group taking the intervention showed an improvement of 37.4% over the pretest scores, whereas the control group exhibited a decrease in the scores (34.9%).
Table 9
Mean Benton’s Test scores for the female population [16]
Experiment or Placebo | Pretest score | Post-test score |
Experiment | 56.75 | 78.00 |
Placebo | 77.33 | 50.33 |
Verbal Fluency Test:
The verbal fluency test assessed the patient’s capacity for fluent speech and semantic memory, with maximum scores of 20 and 1 point, respectively, for each word spoken by the patient. Among the male population, there was no observable improvement in the scores before and after the intervention, while the control group showed a slight improvement in posttest scores.
Table 10
Mean Verbal Fluency Test scores for the male population [16]
Experiment or Placebo | Pretest score | Post-test score |
Experiment | 9.33 | 9.33 |
Placebo | 5.00 | 7.00 |
Among the female patients, the intervention showed an improvement compared to pretest scores (33.3% improvement), whereas the control group experienced a decline in performance. Comparison of posttest mean scores was carried out using the independent T test between the intervention and the control groups. The results, however, showed a statistically insignificant difference in means (t = 0.7793, df = 8, p = 0.45).
Table 11
Mean Verbal Fluency Test scores for the female population [16]
Experiment or Placebo | Pretest score | Post-test score |
Experiment | 7.50 | 9.50 |
Placebo | 6.66 | 6.16 |
Episodic Memory Test:
A test for the episodic memory of patients was carried out using the Free and Cued Selective Reminding Test, with a maximum score of 16. The better the patient’s recall of figures that are presented to him, the better his score is. In the male demographic, the scores showed an improvement for both the intervention and control groups, and surprisingly, a greater improvement was seen in the control group (150% increase in scores) compared to the intervention group (41.3%). However, there was only one patient in the control group, which reduced the reliability of this observation. The study itself reported a statistically significant difference in the pretest and posttest scores; however, only the p value provided was p < 0.01.
Table 12
Mean Episodic memory Test scores for the male population [16]
Experiment or Placebo | Pretest score | Post-test score |
Experiment | 5.66 | 8.00 |
Placebo | 4.00 | 10.00 |
The scoring in women seemed to continue the trend from other tests, where the women receiving the coconut product showed an improvement in the test scores (21.2%), while the women in the control group showed a decline in their performance. This observation, however, when verified by the Independent sample’s T test, did not prove a statistically significant difference between the mean values (t = 0.5446, df = 8, p = 0.600).
Table 13
Mean Episodic memory Test scores for the female population [16]
Experiment or Placebo | Pretest score | Post-test score |
Experiment | 8.25 | 10.00 |
Placebo | 7.50 | 6.00 |