Progressive decline in the ability to perform everyday activities is a core feature of AD and related dementias [34
]. Consistent with its characterization as a prodrome for AD and other dementias [7
], older adults with MCI exhibit detectable restrictions in daily function which, in turn, predict progression to AD [17
]. Yet, little is known about the AD-associated biomarkers that predispose patients with MCI to functional impairment. In this study, we examined two conceptually plausible candidates—whole brain atrophy and APOE ε4 genotype.
Our overarching aim was to delineate the functional significance of increased whole brain atrophy and APOE ε4 positivity in MCI. This is the first study to undertake such an investigation. Our analysis of differences in rates of decline in FAQ as a function of diagnostic status revealed that, relative to controls, patients with MCI experienced an additional 1.22-point biannual deterioration in function whereas patients with mild AD experienced an additional 2-point worsening. The observed trajectories are in accord with reports from prior investigations [4
]. Furthermore, MCI patients who progressed to mild AD exhibited a faster rate of decline in FAQ compared to those who remained stable. Although the FAQ is a relatively coarse measure of daily function, the finding that it reliably distinguishes functional trajectories across the AD spectrum suggests that it is a valid measure of functional abilities in MCI, and is sensitive to change over time.
When examined independently, increased VBR and APOE ε4 positivity were each associated with a faster rate of decline in FAQ. When examined jointly, we found an interactive effect—the impact of VBR on rate of change in FAQ was three times (0.41/0.13) as large among APOE ε4 positive patients as among APOE ε4 negative patients. Overall, patients who had elevated VBR and were APOE ε4 positive experienced the most precipitous decline in function. This indicates that brain atrophy and APOE ε4 positivity have a synergistic relationship with regards to everyday functioning in MCI.
In addition to the characteristic accumulation of amyloid plaques and neurofibrillary tangles, disease progression in AD also results in the systematic and widespread loss of neurons and synapses [36
]. Neuroimaging permits the surrogate visualization and quantification of this cell loss via indices of brain atrophy, such as VBR [37
]. Since extent of brain atrophy arguably reflects severity of AD pathology, it is not surprising that those MCI patients who had increased VBR also experienced more rapid functional degradation. This conclusion is buttressed by our exploratory observation that MCI patients who progressed to mild AD exhibited a more rapid rate of decline in FAQ compared to those patients who remained stable. The finding that greater brain atrophy is associated with faster functional decline parallels prior reports of decline in cognitive function as a result of increased cerebral atrophy [12
]. Indeed, it is theoretically plausible that the effect of brain atrophy on functional decline is mediated by its effect on cognitive decline [14
Possession of one or more copies of the APOE ε4 allele is a well-established risk factor for cognitive decline even among healthy older adults [38
], and for the development of AD [19
], though it has been shown to be neither necessary nor sufficient for the latter [39
]. The precise mechanisms by which APOE ε4 impacts cognitive function or risk of AD is unclear. Possible explanations include its role in amyloid aggregation, fibrillization, and clearance [40
], in neurofibrillary tangle formation [41
], in regulation of the brain’s vasculature [42
], in the metabolism of lipids [40
], and in the modulation of other risk factors [43
]. The association we found between APOE ε4 and rate of functional decline, as well as APOE ε4’s modulatory effect on the relationship between VBR and functional decline, is likely due to APOE ε4’s involvement in these biological processes. Even so, we note that prior attempts to link APOE ε4 to functional decline in the elderly have yielded inconsistent results, likely due to methodological variations [44
The findings from this study have important clinical implications. With the ongoing race to develop disease-modifying drugs for AD, it has become critical to identify markers for monitoring disease progression and response to treatment [45
]. Several studies have posited cerebral atrophy (whether of specific structures, of the whole brain, or simply as ventricular enlargement) as one such marker [28
]. However, attention has been drawn to a need for these markers to be related not only to the underlying pathological process but, perhaps more importantly, to clinically meaningful outcomes [45
]. The findings from this study suggest that VBR may be considered a bone fide marker because it is associated with functional abilities, in addition to cognitive performance [29
]. Similarly, the ability to prevent, decelerate, or reverse limitations in complex daily activities could be a target outcome for MCI treatment studies [47
This study demonstrated that cerebral atrophy and APOE ε4 status have prognostic value with respect to decline in everyday function; and that such functional degradation is more precipitous among patients who experience increased brain atrophy in the context of being APOE ε4 positive. Because progressive functional decline is a signature of AD, this study’s findings suggest that MCI patients who have increased brain atrophy and are APOE ε4 positive may be at elevated risk of progressing to AD. Indeed, if one considers an FAQ score of 12.8 (the mean FAQ score of the AD patients at baseline) the threshold for progression to AD, the trajectories displayed in suggests that APOE ε4 negative patients (whether with normal or elevated VBR) do not attain this threshold within the thirty six-month duration of the study; APOE ε4 positive patients with normal VBR attain this cutoff between 30 and 36 months; and APOE ε4 positive patients with elevated VBR reach this cutoff about one year earlier, between 18 and 24 months. This is potentially useful information for clinicians, patients, and their families with respect to long-term planning. In addition, identifying the class of MCI patients at increased risk for functional change may enable timely implementation of interventions to compress such changes, thereby improving quality of life for both patients and caregivers [4
A possible limitation of this study is the use of an informant-report instrument in assessing everyday functioning. Although easy to obtain, report-based information is susceptible to biases such as those due to erroneous recall, social desirability, and the cognitive/psychological state of the reporter. In addition, FAQ’s relatively gross rating scale and generic nature may be considered limitations. Even so, as noted earlier, its ability to clearly distinguish the functional trajectories of control, MCI, and mild AD participants validates its use in MCI, especially in large-scale studies where the use of more elaborate functional measures may be logistically constrained. A common criticism of cerebral atrophy measures based on ventricular volume is that ventricular expansion could occur as a result of non-degenerative factors such as altered cerebrospinal fluid dynamics, chronic alcohol abuse, cardiovascular diseases, and treatment with diuretics [48
]. These confounds were likely excluded via ADNI’s study entry criteria. Furthermore, though elevated VBR admittedly may be the result of ventricular expansion, brain tissue loss, or both, “in the ex vacuo
state enlarged ventricular space” occurs only in proportion to dissolute brain parenchyma [31
]. Thus, VBR veritably captures the extent of cerebral atrophy. Finally, the MCI patients in this study were of the amnestic variety. As such, it is not known whether our findings would replicate within non-amnestic MCI samples.
In summary, this study is the first to demonstrate that cerebral atrophy and APOE ε4 genotype have prognostic utility with regard to rate of functional decline among patients with MCI, and that their effects may be synergistic. Even so, it is important to highlight that a very recent study [49
] found that the age of onset of sporadic AD was not necessarily a function of possession of ≥ 1 copies of the APOE ε4 allele. Rather, age of onset was influenced by possession of the longer forms of the polymorphic poly-T variant, rs10524523, in the translocase of the mitochondrial membrane 40 (TOMM40) homolog gene that is located in the same region of linkage disequilibrium as APOE. This finding was particularly robust among persons who were APOE ε3/4. It would be of great interest to determine whether TOMM rs10524523 poly-T length influences rate of prospective cognitive and functional decline in MCI to a greater extent than mere APOE carriership, especially among persons who are APOE ε3/4. Another avenue for future research is identifying VBR thresholds that have maximum power for predicting functional decline in MCI. Such thresholds may simultaneously serve in identifying MCI patients who are most likely to progress to AD, for enrollment in clinical trials. Finally, it would be informative to determine whether measures of regional atrophy (e.g., of medial temporal structures) are more sensitive to change in everyday function relative to the VBR and other measures of whole brain atrophy.