Mediterranean diet and treating diabetes and depression in old age may reduce dementia risk

ABSTRACT FROM: Cooper C, Sommerlad A, Lyketsos CG, et al. Modifiable predictors of dementia in mild cognitive impairment: a systematic review and meta-analysis. Am J Psychiatry 2015;172:323–34.

What is already known on this topic

The long ‘preclinical’ phase including mild cognitive impairment-stage (MCI) in the course of Alzheimer's disease (AD) provide the opportunity for preventive interventions to reduce the risk for conversion to dementia or rather slow disease progression.¹ Currently, neither psychopharmacological nor non-pharmacological strategies turned out to prevent patients with MCI from converting to AD. There is increasing evidence that the risk for conversion from MCI to AD could be reduced by modifying life style and consequent treatment of especially the metabolic syndrome. Further, depressive symptoms in old age are frequently associated with cognitive impairment and have been reported as an important risk factor for an early manifestation of dementia.² However, we know little about how to treat or predict dementia outcomes in people with this condition.

Methods of the study

Cooper and colleagues reviewed longitudinal study findings on a variety of potentially modifiable risk factors for the conversion from MCI stage to dementia. The authors conducted an online literature search in PubMed and Web of Knowledge without any limits for language or date of publication (last update 5 June 2014). Sixty two studies were included for qualitative analysis and at least 30 studies for meta-analysis. Quality assessment included an independent evaluation of all eligible studies by two authors based on the following three predefined criteria: (1) representative sample with a response rate of at least 60%; (2) minimal followed up period 1 year and at least 70% followed up; (3) objective diagnostic criteria for MCI and dementia. The authors assigned ‘grade 1 evidence’ for results of numerous high-quality studies (studies meeting all three criteria) and ‘grade 2 evidence’ for results of single high-quality studies or other studies. For data analysis meta-analyses using a random effect model were conducted and unadjusted pooled ORs calculated.

What this paper adds

• Patients with mild cognitive impairment-stage (MCI) with diabetes or the metabolic syndrome have an increased risk for developing all cases of dementia including Alzheimer's disease. In seven reviewed studies the unadjusted pooled OR for conversion to dementia in patients with and without diabetes was 1.65 (95% CI 1.12 to 2.43).

• There is inconsistent evidence for depression and other neuropsychiatric symptoms as predictors for conversion from any MCI type to any dementia. However, meta-analysis of four studies showed an association between the presence of neuropsychiatric symptoms and the conversion to dementia (pooled OR of 3.11, 95% CI 1.38 to 7.02). Further, epidemiological studies did find higher depression scores in patients who developed dementia even though clinical studies did not confirm these findings.

• There is grade 2 evidence that following a Mediterranean diet or not suffering from folate deficiency decreases the risk for conversion from MCI to any dementia.

Limitations

• As the authors stated, applied diagnostic criteria of dementia and MCI are known to be a possible factor of inaccuracy. For instance, some included studies used the 1984 published clinical criteria for the diagnosis of Alzheimer's disease (NINCDS-ADRA) while others already use more recent criteria such as the criteria of McKhann et al³ which take already biomarkers for diagnosing probable dementia into account.

• Many included studies have not differentiated between primary neurodegenerative forms of dementia (such as Alzheimer’s dementia) and secondary dementia subtypes (such as different types of vascular dementia). Since the pathogenesis and clinical presentation varies between the numerous subtypes of dementia, this missing information would have been relevant.

What next in research

Further studies should focus on the active and controlled modification of the presented modifiable predictors by designing interventional longitudinal studies. Especially the use of uniform applied diagnostic criteria for MCI as well as the different types of dementia could significantly raise the quality and clinical benefit of such studies.

Do these results change your practices and why?

Yes. Results of this review underline the strong relationship between modifiable risk factors in particular diabetes, neuropsychiatric symptoms and lower serum folate levels—and the development of dementia in late life and should be considered in the assessment and treatment of patients with cognitive impairment and dementia. Further on, this review emphasised the negative impact of psychiatric diseases, such as alcohol dependency and depression, on the development of any dementia. Consequently, the assessment of cardiovascular comorbidities should be incorporated in the routine diagnostic evaluation of cognitive impairment and dementia. Vice versa, internal medicine specialists should be actively encouraged to screen for signs of cognitive impairment in their patients and whenever possible initiate a cognitive testing on an special memory unit.