Obesity, Brain Microstructure, and Cognition in Ageing

Obesity has been associated with accelerated cognitive decline during ageing and an increased risk for dementia (Fergenbaum et al., 2009; Fitzpatrick et al., 2009; Gustafson et al., 2009; Whitmer et al., 2008). However, recent discussions are questioning this detrimental relationship (Prickett et al., 2015). A protective effect of midlife obesity was found in a large epidemiologic study (Qizilbash et al., 2015), whereas a meta-analysis showed that midlife obesity increases the risk of dementia (Albanese et al., 2017). Methodological difference and reverse causation could be the reasons for such conflicting results (Gustafson, 2015; Kivimäki et al., 2018). For example, some studies used either body mass index (BMI) or waist-to-hip ratio (WHR); their results suggested abdominal or central obesity, which is measured by WHR or waist circumference, is more detrimental for cognitive decline than global obesity measured by BMI (Gustafson et al., 2009; Whitmer et al., 2008). Therefore, in order to assess the effect of body fat on cognitive function in a large group of community-dwelling healthy adults, we used both BMI and WHR as indicators for adiposity (Study 1). To better understand cognitive changes, it is important to investigate obesity-related brain changes. Regarding to brain structure, high body fat has been associated with decreased total brain volume and regional grey and white matter volume (Willette and Kapogiannis, 2015). However, the findings regarding white matter volume are inconsistent; obesity was found to be associated with decreased as well as increased white matter volumes (Bobb et al., 2014; Driscoll et al., 2012; Karlsson et al., 2013; Walther et al., 2010). Because of the inconsistency, diffusion-weighted imaging (DWI), which assesses the microstructural architecture, has been considered to have higher agreements among studies on obesity and decreased directional property of white matter (Willette and Kapogiannis, 2015). However, a recent study reported a positive association between obesity and white matter microstructure (Birdsill et al., 2017). The conflicting results may be explained by the relative small–medium sample size (n = 15– 268) among the studies and limited statistical power for data-driven whole brain voxel- wise analyses. Following the thought of increasing statistical power with bigger samples might be beneficial and more reliable, we investigated the link between body fat and white matter microstructure using a population-based sample of 1255 participants in Study 1. The obesity-related changes in the brain is considered to be a result of neuroinflammation. The neuroinflammation is likely caused by high fat intake- and excess adipose tissue-induced peripheral inflammation, which in turn leads to insulin resistance and hyperglycaemia; and these all together could also affect the hippocampus and result in memory inhibition (O’Brien et al., 2017; Pugazhenthi et al., 2016). A number of studies reported a negative association between obesity and hippocampal volume (Debette et al., 2011; Raji et al., 2010; Taki et al., 2008). However, others found a positive association (Widya et al., 2011) or no association (Bobb et al., 2014; Debette et al., 2011; Driscoll et al., 2012). This could be due to the fact that hippocampus comprises several subfields (cornu ammonis fields [CA1–4], the dentate gyrus [DG], and the subiculum), and these subfields have distinct functional properties (Deng et al., 2010; Strien et al., 2009; Yassa and Stark, 2011). It has also been reported that the hippocampal subfields are affected by ageing differently (Malykhin et al., 2017). Because the hippocampus is highly susceptible to degenerative processes (Pfefferbaum et al., 2013; Raz et al., 2010) and possesses neurogenesis ability in adults (Eriksson et al., 1998), it is often a key target in intervention studies. Several studies indicate that lifestyle interventions can be effective to combat obesity and to restore cognitive functions (Siervo et al., 2011). A double-blind randomised controlled trial of 1260 older individuals (aged 60–77 years) from the FINGER study showed that the intervention group obtained higher scores in a neuropsychological test battery compared to the control group after a 2-year multidomain intervention (diet, exercise, cognitive training, vascular risk monitoring) (Ngandu et al., 2015). Dietary nutrients that have anti-inflammatory or anti-oxidative effects, such as omega-3 fatty acids and polyphenols, have been associated with improved cognitive function and brain structural changes (Gómez-Pinilla, 2008; Huhn et al., 2015). Resveratrol is one type of polyphenols and occurs in variety of plants such as red grapes and blueberries (Baur et al., 2006). It has been reported to be associated with glucose metabolism and insulin sensitivity (Liu et al., 2014) as well as improved cognitive functions (Huhn et al., 2015; Marx et al., 2018). However, some studies reported no improvements in cognition after intervention (Köbe et al., 2017; Wightman et al., 2015). Besides the possibility that resveratrol does not affect cognition, another reason for these results could be that resveratrol is only effective in healthy ageing population (Evans et al., 2017; Witte et al., 2014) but not in young (Wightman et al., 2015) or patient population (Köbe et al., 2017). Therefore, replication studies with healthy older adults could help to evaluate whether there is an effect of resveratrol on cognitive function. Following this idea, we have conducted a double-blind randomised controlled trial (Study 2) with comprehensive neuropsychological test battery to assess the effect of resveratrol using 60 elderly subjects. In Study 1, we applied whole brain voxel-wise analysis to explore the correlations between overall obesity (measured by BMI) or abdominal obesity (WHR) and white matter microstructure. We found a negative correlation between BMI as well as WHR and fractional anisotropy (FA), a measure of microstructural architecture, in multiple white matter tracts independent of confounding factors. We further explored the indirect link of obesity and cognitive dysfunction using mediation analysis. In the mediation analysis, an indirect path through obesity-associated clusters was considered. We found that although obesity had no direct effect on executive functions and processing speed, it affected cognitive performance through lower FA in callosal and associative fibre tracts. We found the correlation between obesity and memory performance was not mediated by FA in the selected white matter tracts. In Study 2, we conducted a randomised trial for the effect of resveratrol on memory performance and hippocampal structure. We found that intake of resveratrol did not show any beneficial effect on either glucose metabolism or cognitive performance. Neither volume nor mean diffusivity (MD), another measure of microstructural architecture, showed changes after the intervention compared to the placebo group. However, subtle ageing- or lifestyle-related changes in the MD of the hippocampus were detected. This demonstrated that MD outperforms volumetric measures for detecting subtle changes of the hippocampus. The reason we could not observe any changes after resveratrol intake might be that this compound does not have an effect or that other lifestyle changes undermined the effect of resveratrol as neuroplasticity can be influenced by many factors. This thesis highlights that body fat is associated with lower FA in the white matter of the brain. This may indicate some widespread damage to the white matter; and mediation analysis indicates abdominal obesity is linked to poorer executive functions and processing speed through lower FA. Further this thesis shows that adding a dietary supplementation of resveratrol for six months does not improve memory or hippocampal structure in the present cohort of healthy adults with a large BMI range. Future studies should investigate longitudinal changes of body fat and brain structure in order to establish the causal relationship among obesity, white matter microstructure, and cognitive function. And more comprehensive lifestyle interventions combining diet, exercise, and cognitive training should be considered instead of one single approach to prevent and hopefully preserve obesity induced changes in cognition and in the brain