To date, no effective treatment is available for prevention or cure of dementia. In order to reduce the burden of Alzheimer's disease (AD) and dementia, it is crucial to understand the mechanisms underlying the disease and to identify potentially modifiable risk factors. Accumulating evidence suggests that multiple vascular and environmental risk factors are associated with late-life cognitive impairment, however consistent findings in epidemiological studies for the association with dementia exist only for age, family history and ApoE4 genotype, and all these are not modifiable. This inconsistency may be attributed to differences in population sample, definitions of variables and study design. In exploring risk factors associated with dementia it is notably important to use a prospective design with long follow-up periods, since this disease has a long sub-clinical phase. In FHS, we had the opportunity to explore various potential risk factors for AD and dementia in a large population-based cohort with long and intensive follow-ups.
One example is a study in which the relationship between levels of Plasma Phosphatidylcholine (PC) Docosahexaenoic Acid (DHA), an Omega-3 fatty-acid and risk of dementia and AD was assessed[14
]. In this study, plasma samples were obtained from 899 dementia-free participants of the Original cohort (median age was 79), and plasma PC DHA were measured. Then, these participants were followed-up for the development of AD or dementia over a mean duration of 9.1 years. In addition, dietary data on DHA and fish intake was available in a subsample. In this study, subjects with baseline plasma PC DHA levels in the upper quartile experienced a significant lower risk of dementia (Relative Risk =0.53; 95% confidence interval, 0.29-0.97) compared to participants with levels in the lower 3 quartiles, independent of age, sex, ApoE4 genotype, plasma homocysteine concentration, and education level. In addition, a correlation between plasma PC DHA content and fish intake was found (both with p value<0.001), however fish intake accounted for less than half of the variability in DHA levels. This study was the first to use a prospective design in assessing the predictive value of plasma PC DHA content in the occurrence of dementia and AD, and it set the background for subsequent research, including clinical trials and gene expression studies. Recently, further exploration of these findings in the Offspring cohort revealed lower brain volumes and a “vascular” pattern of cognitive impairment among dementia free persons with lower red blood cell DHA levels[15
Diabetes is another example of a modifiable risk factor that was explored within the Original cohort of the FHS. Although diabetes is known to be a risk factor for cognitive impairment and dementia as reported in numerous population-based studies, it remains controversial in relation to AD; for instance it has been suggested that the association observed is with the vascular disease that co-exists in persons with clinical AD[16
]. The purpose of the study by Akomolafe et al. [17
] was not just to explore whereas such an association exists, but also to define a sub-population in which this relationship might be more apparent. Thus, 2210 dementia-free participants (mean age 70 years) were followed for a mean duration of 12.7 years. The Relative Risk (RR) of AD comparing diabetic patients to non-diabetics was not statistically significant (1.15; 95% Confidence interval, 0.65-2.05). However, among participants with neither the ApoE4 allele nor elevated plasma homocysteine levels, the RR (95% Confidence interval) was 2.98 (1.06-8.39). Moreover, the effect was strongest among younger individuals (<75 years) (RR=4.77; 95% confidence interval, 1.28-17.72). This study suggested that diabetes should be considered as a risk factor for AD, and emphasized the need to further explore the predictive value of diabetes and other risk factors in specific subgroups such as in those with an initial low risk for developing AD.
Though the pathological mechanisms that relate diabetes, DHA levels and other risk factors to the development of dementia are not clear, strong evidence supports the role of inflammation as a common link. Diabetes is associated with low-grade inflammation [18
] and Omega-3 fatty acids have a variety of anti-inflammatory effects [19
], and they influence the expression of many genes which are involved in inflammation [20
]. Yet, clinical observations on the potential role of inflammation in AD have yielded inconsistent results. In a study by Tan et al., serum cytokines and spontaneous production of peripheral blood mononuclear cell (PBMC) cytokines were related to the risk of incident AD 7±3 years later, in 691 dementia-free participants of the Original FHS cohort [21
]. Participants in the second tertile of IL-1 production levels were at increased risk of developing incident AD compared to those in the lowest tertile (Hazard Ratio=2.84; 95% Confidence Interval, 1.09-7.43) and individuals within the highest tertile of production showed a trend toward an increased risk but with borderline significance (Hazard Ratio=2.61; 95% Confidence Interval, 0.96-7.07). High production of TNF-α was associated with future AD (Hazard Ratio=2.59; 95% Confidence Interval, 1.09-6.12 comparing highest to lower two tertiles). In contrast, levels of circulating inflammatory cytokines (CRP, IL-6) were not associated with the risk of AD. This study demonstrates the possible importance of spontaneous production of cytokines by PBMC in reflecting intracerebral inflammation as opposed to serum cytokine levels which reflect systemic inflammation, but due to the blood brain barrier may not adequately reflect intracerebral inflammation. PBMCs cross the blood brain barrier and are detected as brain macrophages that impact neuronal and glial function.