Dorien Bangma

72 | CHAPTER 4 no difficulties with ‘expressing a choice’ when making financial decisions (Lui et al., 2013), but need more time than healthy controls to complete a FDM task (Lassen-Greene et al., 2017; Okonkwo et al., 2006). In a computer-based simulation task, people living with MCI were found to have more difficulties than healthy controls with the use of an ATM (Czaja et al., 2017). As expected, and in accordance with the findings mentioned above, an overall medium to large pooled mean effect size was found in the meta-analysis when comparing the performances of people living with MCI and healthy controls on tests of FDM ( g = 0.95 [0.78; 1.11], SE = 0.08, p < .001 based on 16 studies; Figure 4.2). Significant heterogeneity was, however, found ( Q (15) = 55.2, p < .001, I 2 = 72.8%). More than half of the studies on MCI used the FCI (n = 9) and for these studies a significantly larger pooled mean effect size was found ( g = 1.09 [0.92; 1.25], SE = 0.09, p < .001; Q(1) = 8.80, p = .003) compared to the seven studies using FDM tests other than the FCI ( g = 0.70 [0.52; 0.89], SE = 0.10, p < .001). The funnel plot showed a significant asymmetry (p = .005; Figure 4.3). One study (Bangen et al., 2010) compared people living with amnestic MCI, people living with non-amnestic MCI and healthy controls and only found significant differences between people living with amnestic MCI and healthy controls, while other group comparisons were not significant. These findings are consistent with another study comparing people living with different subtypes of MCI (Pereira, Yassuda, et al., 2010). Interestingly, based on binary logistic regression, the ‘money management’ subscale of the ILS was a significant contributor in the prediction of the subtype of MCI (i.e., amnestic or non-amnestic MCI; Bangen et al., 2010). Longitudinal studies. Over time, people living with MCI showed a significant deterioration on multiple domains of FDM (Clark et al., 2014; Gerstenecker et al., 2016; Lassen-Greene et al., 2017; Martin et al., 2019; Niccolai et al., 2017; Pereira, Oliveira, et al., 2010; Triebel et al., 2009). In one study (Triebel et al., 2009), after a one-year follow-up, 18% of participants converted from MCI to AD and showed a stronger deterioration than MCI non- converters on overall FDM (i.e., total score) and, more specifically, on ‘checkbook management’ of the FCI. Furthermore, deterioration was found in people living with MCI after two-years for ‘checkbook management’, ‘bank statement management’, ‘bill payment’ and ‘investment decision making’ of the FCI (Gerstenecker et al., 2016). Two other studies also reported significant deterioration after two-years in FDM performances in people living with MCI (Clark et al., 2014; Niccolai et al., 2017). Clark et al. (2014) demonstrated that MCI converters showed more often a significant decline on the FCI (i.e., a decline of more than 10 points) after one as well as after two-year(s) follow-up compared to MCI non-converters. However, statistics for group comparisons were not reported. MCI converters may already have more difficulties with some aspects of FDM at baseline compared to MCI non-converters (Triebel et al., 2009), however, two other studies did not support this finding (Clark et al., 2014; Pereira, Oliveira, et al., 2010). Furthermore, baseline performances of people living with MCI could not predict the deterioration after two years (Niccolai et al., 2017). Lassen-Greene et al. (2017) explored the performances on the FCI in more detail and reported that in people living with MCI only the accuracy of performances, not the speed of processing, deteriorates over a three-years period. Finally, Martin et al. (2019) performed a three and six-years follow-up using