Using clinical-pathological data from more than 1,000 individuals, Boyle et al. show that age-related neuropathologies are differentially related to late-life cognitive trajectories. Whereas some contribute to progressive cognitive deterioration, others impair cognition but exert relatively stable effects over time. The findings have implications for preventing and treating late-life cognitive decline.
Keywords: cognitive ageing, neuropathology, Alzheimer’s, small vessel disease, TDP-43
Abstract
The objective of this study was to examine whether the effects of age-related neuropathologies on cognition change over time. Participants were 1096 deceased persons from two clinical-pathologic studies. All were without dementia at baseline, completed a detailed battery of cognitive tests annually over up to 21 years, died, and underwent detailed neuropathologic examinations to identify Alzheimer’s disease pathology, vascular pathologies (i.e. macro- and microscopic infarcts, atherosclerosis, arteriolar sclerosis, and cerebral amyloid angiopathy), Lewy bodies, transactive response DNA-binding protein 43 (TDP-43) pathology, and hippocampal sclerosis. A time-varying effects model was used to flexibly characterize the trajectory of global cognition and assess whether the effects of demographics and each neuropathologic index on cognition changed over time. Results indicated that the mean trajectory of global cognition was characterized by gradual cognitive decline beginning ∼15 years before death and accelerated decline in the last few years. With the exception of microinfarcts and arteriolar sclerosis, all neuropathologies were associated with the cognitive trajectory. However, the nature of their associations varied. Alzheimer’s disease pathology, macroscopic infarcts, Lewy bodies, TDP-43 pathology, and hippocampal sclerosis were associated with progressive cognitive decline, with their deleterious effects increasing over time. By contrast, atherosclerosis and cerebral amyloid angiopathy pathology were associated with a lower level of cognition but their effects were relatively stable over time. These results suggest that age-related neuropathologies are differentially related to late life cognitive trajectories. Whereas some contribute to progressive cognitive deterioration, others lower the level of cognition but exert relatively stable effects over time.
Introduction
Loss of cognition in old age is primarily driven by the accumulation of age-related neuropathologies. It is now well established that neuropathologic indices of the three most common causes of dementia, Alzheimer’s disease, cerebrovascular disease, and Lewy bodies, are frequently observed in the brains of older persons with and without dementia and are important determinants of cognitive impairment and decline (Driscoll et al., 2006; Sonnen et al., 2007; Markesbery, 2010; Bennett et al., 2012c; Boyle et al., 2013). More recent data, however, suggest that less well-studied neuropathologies also are common and related to cognitive impairment and decline; these include transactive response DNA-binding protein 43 (TDP-43) pathology, hippocampal sclerosis, and additional vascular pathologies including cerebral atherosclerosis, arteriolar sclerosis, and cerebral amyloid angiopathy (Nelson et al., 2010, 2011, 2013; Wilson et al., 2013; Boyle et al., 2015; Nag et al., 2015; Arvanitakis et al., 2016). Further, despite interrelationships between some of these neuropathologies, their effects on late life cognitive decline are relatively independent.
Although the deleterious effects of age-related neuropathologies on cognition are increasingly recognized, little is known about how the effects of neuropathology on cognition manifest over the course of ageing. Age-related neuropathologies have different patterns of accumulation, frequently co-occur, and many cannot be quantified in vivo, making it difficult to determine how they impact cognition as individuals age. Clinical-pathologic studies suggest that age-related neuropathologies may have differential effects on late life cognitive trajectories, with some (e.g. Alzheimer’s disease pathology, infarcts) associated with cognitive decline, and others (e.g. microinfarcts) associated with a lower level of cognition but not rate of decline (Boyle et al., 2013; Skrobot et al., 2016). Notably, however, most studies used linear modelling approaches, despite evidence that cognitive decline in old age is non-linear (Hall et al., 2001; Dodge et al., 2011). Some studies have used changepoint models to address non-linearity, and this approach reveals more complex associations; for example, whereas Alzheimer’s disease pathology and gross infarcts are more strongly related to early (i.e. preterminal) cognitive decline, Lewy bodies are more strongly associated with later (i.e. terminal) decline (Wilson et al., 2010, 2015; Boyle et al., 2013). We recently expanded on the changepoint models by using a novel sigmoidal modelling approach to examine the relation of neuropathology with four components of the cognitive trajectory (Wilson et al., 2016). Results again revealed differential associations of specific neuropathologies with components of the cognitive trajectory. Although changepoint and sigmoidal models accommodate non-linearity and provide evidence of complex neuropathology-cognition associations, they are limited by a priori assumptions about the functional form of the cognitive trajectory. Further, they do not allow for the effects of neuropathology to change flexibly over time. Thus, how the effects of neuropathology on cognition unravel over the course of ageing remains largely unknown.
In this study, we expand on prior work by employing a time-varying effects model (TVEM), a unique and flexible modelling approach, to examine whether the effects of multiple age-related neuropathologies on cognition change over time (Yang, 2012). TVEM differs from prior approaches in two important ways: first, it does not assume a specific functional form (e.g. linear, sigmoidal) for outcomes such as longitudinal cognitive decline. Second, TVEM allows the effect of predictors (i.e. neuropathology) to change over time, again without assuming a specific functional form. The approach therefore can flexibly capture the trajectory of late life cognition and reveal dynamic effects of age-related neuropathologies on cognition over time. Participants in this study were 1096 deceased older persons from two longitudinal clinical-pathologic cohort studies of ageing. All were without dementia at baseline, completed a minimum of two cognitive evaluations (maximum = 21), died and underwent uniform neuropathologic examinations for identification of age-related neuropathologies.
Materials and methods
Participants
This study included deceased participants from the Religious Orders Study and the Rush Memory and Aging Project (Bennett et al., 2012a, b). The Religious Orders Study involves older Catholic priests, nuns, and monks throughout the USA, and the Rush Memory and Aging Project involves older lay persons recruited from the Chicago area. Both studies require detailed annual clinical evaluations and brain autopsy. Participants provide informed consent and an anatomical gift act, and the studies were approved by the Institutional Review Board of Rush University Medical Center.
Eligibility for these analyses required the absence of dementia at the study baseline, at least two cognitive evaluations, and death with a brain autopsy and complete neuropathologic examination. At the time of these analyses, 3124 individuals had enrolled in the parent studies and completed the baseline evaluation; 194 persons with dementia at baseline were excluded. Of the remaining 2930 individuals, 241 subjects only completed the baseline assessment (100 died and 141 dropped out prior to the first follow-up evaluation). Of the remaining 2689 subjects, 1278 died, and a brain autopsy was done on 1121 persons (88%); of those, neuropathological examination was complete for 1096 individuals (97.8%).
Clinical diagnosis and assessment of global cognitive function
Details of the diagnostic procedures have been described previously (Bennett et al., 2012a, b). After review of all test results and information about education, occupation, effort on testing, and potential sensory or motor problems, a neuropsychologist provided a rating of the presence or absence of cognitive impairment. Subsequently, an experienced clinician diagnosed dementia following standard criteria that require a history of cognitive decline and impairment in two or more cognitive domains, one of which must be memory for the diagnosis of clinical Alzheimer’s disease (McKhann et al., 1984).
Each annual clinical evaluation included a battery of 19 cognitive tests. The Mini-Mental State Examination (MMSE) was used to describe the cohort, and a subset of items from the MMSE and a short form of Complex Ideational Material were used only for clinical classification (Folstein et al., 1975; Bennett et al., 2012a, b). Performance on the remaining 17 tests was used to create a composite measure of global cognition in order to minimize floor and ceiling artefacts in longitudinal analyses. For computation of the composite measure of global cognition, scores on individual tests were converted to z scores using the baseline mean and standard deviation of all participants in both studies, and the z-scores were averaged together. Further information on this measure and the individual tests is published elsewhere (Wilson et al., 2002, 2005, 2015).
Neuropathologic evaluation
Brain autopsy procedures have been described in detail and were done by examiners who were unaware of all clinical information (Schneider et al., 2007; Boyle et al., 2013). Briefly, brains were removed, weighed, and hemispheres were cut coronally using a Plexiglas® jig into 1 cm slabs. One hemisphere was fixed in 4% paraformaldehyde. After gross examination of both hemispheres, nine brain regions of interest (i.e. midfrontal, midtemporal, inferior parietal, anterior cingulate, entorhinal and hippocampal cortices, basal ganglia, thalamus and midbrain) were dissected from the 1 cm slabs of fixed tissue and processed and embedded in paraffin. Sections (6 µm) from the paraffin blocks were stained for assessment of pathology.
Alzheimer’s disease pathology (i.e. neuritic plaques, diffuse plaques and neurofibrillary tangles) was visualized in five cortical regions using a modified Bielschowsky silver stain. Counts of silver-stained neuritic plaques, diffuse plaques, and neurofibrillary tangles were used to create a continuous global measure of Alzheimer’s disease pathology. Gross (chronic) infarcts were identified visually by examining slabs and pictures from both hemispheres, confirmed histologically, and treated as present or absent in analyses. Lacunar infarcts that were ≤1 cm in the subcortical grey or white matter were considered macroscopic infarcts. Microinfarcts (chronic) were those that were not visible to the naked eye but were identified under microscopy using haematoxylin and eosin stain in a minimum of nine regions, including six cortical regions, two subcortical regions, and midbrain; microinfarcts were treated as present or absent in analyses. Lewy body pathology was assessed using antibodies to α-synuclein in six regions including substantia nigra, limbic and neocortices and treated as present or absent in analyses. Hippocampal sclerosis was identified as severe neuronal loss and gliosis in hippocampus or subiculum using haematoxylin and eosin stain and treated as present or absent in analyses (Nag et al., 2015). TDP-43 cytoplasmic inclusions were assessed in six regions using antibodies to phosphorylated TDP-43 (pS409/410; 1:100); inclusions in each region were rated on a six-point scale and the mean of the regional scores was used in analyses (Wilson et al., 2013). Arteriolar sclerosis was detected histologically by use of haematoxylin and eosin-stained sections of the anterior basal ganglia and rated as absent, mild, moderate or severe; for these analyses, arteriolar sclerosis was rated absent or present to facilitate interpretation of the results, with present including mild, moderate and severe ratings (Arvanitakis et al., 2016). Atherosclerosis was assessed on gross examination by visual inspection of vessels in the circle of Willis and rated as absent, mild, moderate or severe; for these analyses, atherosclerosis was rated absent or present, with present including mild, moderate and severe ratings (Arvanitakis et al., 2016). Cerebral amyloid angiopathy was assessed using amyloid-β immunostaining in four regions (i.e. midfrontal, inferior temporal, angular, and calcarine); in each region, meningeal and parenchymal vessels were assessed for amyloid deposition and scored from 0–4, where 0 = no deposition; 1 = scattered segmental but no circumferential deposition; 2 = circumferential deposition up to 10 vessels; 3 = circumferential deposition in > 10 vessels and up to 75% of the vessels; and 4 = circumferential deposition in over 75% of the vessels. The score for each region was the maximum of the meningeal and parenchymal scores and a continuous summary score was created by averaging scores across regions (Boyle et al., 2015).
Statistical analyses
We used TVEM, a unique and flexible statistical method for exploring dynamic associations between predictors (e.g. neuropathology) and longitudinal outcomes (i.e. cognition) (Shiyko et al., 2012; Yang, 2012). TVEM differs from traditional regression analysis in that it does not assume a specific functional form (e.g. linear, quadratic) for the longitudinal outcome, and it allows the effect of predictors (i.e. their coefficients) to vary flexibly over time, again without assuming a particular functional form. As a result, the model captures complex cognitive trajectories and provides estimated coefficients and corresponding confidence bands for the association of each neuropathologic index with cognition at each cognitive assessment. By allowing the effect of neuropathology to vary flexibly over time, this method can directly inform on the dynamic effects of different neuropathologies on cognitive trajectories over time; for example, it is well appreciated that a clinical stroke can be associated with rapid cognitive decline followed by stabilization or even improvement in cognitive function, and this method can capture such changing effects over time. Our model included age at death, sex, education and all nine of the age-related neuropathologies, with all variables treated as time-varying. Further, to summarize the time-varying effects of each neuropathologic index, we calculated the mean estimated coefficient and range of estimates for each index over time.
Results
Characteristics of study participants
Participants (n = 1096) in these analyses had a mean age at baseline of 80.2 years [standard deviation (SD) = 6.8] and a mean of 16.4 years of education (SD = 3.7). They died at a mean age of 88.9 years (SD = 6.6) after completing a mean of 7.7 years of follow-up (SD = 4.5); additional descriptive data are provided in Table 1. Notably, there was substantial decline in cognition during the follow-up period, as evidenced by the difference between the global cognitive scores at baseline and proximate to death; this is consistent with the observation that, among eligible persons who started without dementia, by the time of death, 427 (39.2%) persons had dementia. There also was a considerable burden of neuropathology among decedents. Table 2 shows the correlations of pathology variables with demographics.
Table 1.
Descriptive characteristics of the cohort
| Variable | Mean, SD (%) |
|---|---|
| Age at death, years | 88.94 ± 6.59 |
| Education, years | 16.36 ± 3.67 |
| Gender, female, n (%) | 733 (66.9) |
| Ethnicity, white, non-Hispanic, n (%) | 1064 (97.2) |
| Baseline MMSE | 27.94 ± 1.99 |
| Proximate to death MMSE | 21.73 ± 8.65 |
| Baseline global cognition | −0.02 ± 0.51 |
| Global cognition proximate to death | −0.82 ± 1.10 |
| Global Alzheimer’s disease pathology | 0.70 ± 0.59 |
| Braak stage (% 0, I/II, III/IV, V/VI) | 1, 17.3, 58.2, 23.5 |
| TDP-43 (% positive) | 0.60 ± 0.94 (50) |
| Hippocampal sclerosis, n (% present) | 78 (7.2) |
| Gross infarcts, n (% present) | 378 (34.5) |
| Cerebral amyloid angiopathy (% present) | 1.11 ± 1.06 (77.7) |
| Severity: % rated mild, moderate, severe | (43.3, 23, 11.4) |
| Atherosclerosis, n (% present) | 891 (81.7) |
| Severity: % rated mild, moderate and severe | 47.4, 27.3, 7.0 |
| Arteriolar sclerosis, n (% present) | 745 (68.5) |
| Severity: % rated mild, moderate and severe | 36.6, 23.3, 8.6 |
| Lewy bodies, n (% present) | 241 (22.0) |
| Microscopic infarcts, n (% present) | 321 (29.3) |
MMSE = Mini-Mental State Examination.
Table 2.
Correlations of pathology with demographics
| Pathology variable | Age at death | Male sex | Education |
|---|---|---|---|
| Global Alzheimer’s disease pathology | 0.14 (<0.0001) | −0.14 (<0.0001) | −0.03 (0.307) |
| TDP-43 | 0.21 (<0.0001) | −0.07 (0.023) | −0.04 (0.201) |
| Hippocampal sclerosis | 0.14 (<0.0001) | −0.04 (0.244) | −0.03 (0.356) |
| Gross infarcts | 0.15 (<0.0001) | 0.03 (0.359) | −0.04 (0.168) |
| Cerebral amyloid angiopathy | 0.12 (<0.0001) | −0.06 (0.056) | 0.03 (0.269) |
| Atherosclerosis | 0.09 (0.003) | 0.02 (0.451) | 0 (0.750) |
| Arteriolar sclerosis | 0.09 (0.002) | −0.09 (0.002) | −0.06 (0.056) |
| Lewy bodies | 0.06 (0.050) | 0.05 (0.118) | −0.01 (0.797) |
| Microscopic infarcts | 0.06 (0.033) | 0 (0.923) | −0.07 (0.031) |
Data are presented as correlation coefficient (P-value).
Estimating the trajectory of global cognition and the effect of demographics
We used TVEM to estimate the mean trajectory of global cognition over up to 21 years of observation (mean = 7.7 years, SD = 4.5 years); this analysis required complete data on all variables and included age at death, sex, education, and the nine age-related neuropathologies (i.e. Alzheimer’s disease pathology, macro and microscopic infarcts, Lewy bodies, TDP-43 pathology, hippocampal sclerosis, cerebral amyloid angiopathy pathology, atherosclerosis, and arteriolar sclerosis), with all variables treated as time-varying (i.e. their effects could vary flexibly over time). As shown in Fig. 1, the mean trajectory of global cognition was characterized by gradual cognitive decline beginning ∼15 years prior to death, with an accelerated rate of decline in the few years just prior to death. As shown in Fig. 2A, the association of age with global cognition was evident beginning >15 years before death, when every additional decade of age was associated with a 0.18-unit decrease in global cognition [standard error (SE) = −0.03, P = 1.12 × 10−8]. Notably, this association remained relatively stable over time (mean estimate = −0.012, range of estimates: −0.015 to −0.009; SD = 0.002), as indicated by the nearly flat curve. As shown in Fig. 2B, the association of education with cognition was evident >15 years before death, when every additional year of education was associated with a 0.024 unit increase in global cognition (SE = 0.005, P = 3.53 × 10−7); this association also remained relatively stable (mean estimate = 0.027, range of estimates: 0.023 to 0.031, SD = 0.003). Finally, the association of male sex with cognition was evident > 15 years prior to death (Fig. 2C), with males having lower global cognition than females (estimate = −0.121, SE = 0.032; P = 1.78 × 10−4); however, this effect levelled off and was not evident in the last decade of life (range of estimates: −0.183 to −0.004, SD = 0.053).
Figure 1.
The estimated mean trajectory of global cognitive function (dotted red line) and the corresponding 95% confidence band (shaded area) derived from the fully-adjusted TVEM model. Dark blue indicates the area of the trajectory with the most dense cognitive data, and light blue indicates tails where cognitive data are more sparse.
Figure 2.
The estimated effects of age (A, dotted red line), education (B, dotted red line), and male sex (C, dotted red line) on the trajectory of global cognition and the corresponding 95% confidence bands (shaded areas) derived from the fully-adjusted TVEM model. Dark blue indicates the area of the trajectory with the most dense cognitive data, and light blue indicates tails where cognitive data are more sparse.
Effect of age-related neuropathologies on the trajectory of global cognition
Simultaneous examination of the relation of age-related neuropathologies revealed that, with the exception of microscopic infarcts and arteriolar sclerosis, all other age-related neuropathologies (i.e. Alzheimer’s disease pathology, macroscopic infarcts, Lewy bodies, TDP-43 pathology, hippocampal sclerosis, cerebral amyloid angiopathy pathology and atherosclerosis) were associated with the trajectory of global cognition. Interestingly, however, the nature of the associations of these neuropathologies with cognition varied. Specifically, Alzheimer’s disease pathology, macroscopic infarcts, Lewy bodies, TDP-43 pathology, and hippocampal sclerosis were associated with progressive cognitive deterioration, with their effects increasing over time (Figs 3 and 4). As shown in Fig. 3A, the association of Alzheimer’s disease pathology on global cognition was evident ∼17 years before death, when every one additional unit of Alzheimer’s disease pathology was associated with a −0.072 unit decrease in global cognition (SE = 0.035; P = 0.048); this effect doubled ∼13 years and tripled ∼10 years before death (range of estimates: −0.580 to −0.054, SD = 0.135). The association of hippocampal sclerosis and cognition was evident ∼13 years before death (estimate = −0.112, SE = 0.055; P = 0.042) and doubled and tripled ∼10 and ∼3 years before death, respectively (range of estimates: −0.350 to 0.010, SD = 0.116); (Fig. 3B). The association of macroscopic infarcts and cognition was evident ∼13 years before death (estimate = −0.056, SE = 0.028; P = 0.044) and doubled and tripled ∼7 and ∼4 years, respectively, before death (range of estimates: −0.272 to 0.026, SD = 0.072; Fig. 3C). The association of Lewy bodies was evident ∼10 years before death (estimate = −0.046, SE = 0.022; P = 0.043) and doubled and tripled ∼8 and ∼5 years before death, respectively (range of estimates: −0.281 to 0.096, SD = 0.102; Fig. 4A). Finally, the association of TDP-43 became evident ∼7 years before death (estimate = −0.029, SE = 0.015; P = 0.049) and doubled and tripled ∼5 and 4 years before death, respectively (range of estimates: −0.155 to 0.023, SD = 0.052; Fig. 4B).
Figure 3.
The estimated effects of Alzheimer’s disease pathology (A, dotted red line), hippocampal sclerosis (B, dotted red line), and macroscopic infarcts (C, dotted red line) on the trajectory of global cognition and the corresponding 95% confidence bands (shaded areas) derived from the fully-adjusted TVEM model. Dark blue indicates the area of the trajectory with the most dense cognitive data, and light blue indicates tails where cognitive data are more sparse.
Figure 4.
The estimated effects of Lewy body pathology (A, dotted red line) and TDP-43 pathology (B, dotted red line) on the trajectory of global cognition and the corresponding 95% confidence bands (shaded areas) derived from the fully-adjusted TVEM model. Dark blue indicates the area of the trajectory with the most dense cognitive data, and light blue indicates tails where cognitive data are more sparse.
In contrast to the neuropathologies with increasingly deleterious effects on cognition, the associations of two vascular pathologies, atherosclerosis and cerebral amyloid angiopathy pathology, with cognition were relatively stable over time. That is, the association of atherosclerosis was evident >15 years prior to death and reduced global cognition by an average of –0.178 unit, but this association did not change substantially over time (range of estimates −0.208 to −0.146, SD = 0.023) (Fig. 5A). The association of cerebral amyloid angiopathy pathology with cognition was evident >11 years before death but was small and also remained stable over time (average = −0.023, range of estimates: −0.044 to −0.007, SD = 0.012) (Fig. 5B).
Figure 5.
The estimated effects of atherosclerosis (A, dotted red line) and cerebral amyloid angiopathy pathology (B, dotted red line) on the trajectory of global cognition and the corresponding 95% confidence bands (shaded areas) derived from the fully-adjusted TVEM model. Dark blue indicates the area of the trajectory with the most dense cognitive data, and light blue indicates tails where cognitive data are more sparse.
Discussion
In this study, we examined the time-varying relations of several age-related neuropathologies with the trajectory of late life cognitive decline using data from >1000 persons in whom cognitive function was assessed annually for up to 21 years. Findings showed that the deleterious effects of many age-related neuropathologies emerged more than a decade prior to death. However, the nature of their associations with the trajectory of cognition varied. Whereas the effects of some neuropathologies were progressive, with their deleterious effects increasing over time, the effects of others were relatively stable over time. These findings suggest that age-related neuropathologies are differentially related to late life cognitive trajectories, with some contributing to progressive cognitive deterioration and others lowering the level of cognitive function at some point without leading to subsequent progressive cognitive decline.
The present study builds on prior work in three important ways. First, the application of TVEM to longitudinal cognitive data allows for a flexible (i.e. without assuming a particular functional form) examination of the trajectory of late life cognitive decline. Notably, the vast majority of studies of cognitive decline in old age have used linear modelling approaches; this works well when cognitive assessments are conducted many years apart and/or the number of evaluations is limited. However, late life cognitive decline is non-linear when measured at relatively short (e.g. annual) intervals over many years (Sliwinski, 2006; Wilson et al., 2010, 2012; Dodge et al., 2011; Boyle et al., 2013). Further, although changepoint and other models that characterize components of the cognitive trajectory have been used and represent an improvement over linear models, they are limited by a prespecified functional form and other assumptions, such as the optimal number of changepoints to use; these models therefore may not accurately represent the natural history of cognitive ageing and may be suboptimal for persons whose trajectories are inconsistent with the specified functional form. Flexible modelling approaches such as that used here are needed to accommodate nonlinearity and characterize more nuanced patterns of cognitive decline over time. The present findings suggest that, on average, cognitive ageing is characterized by gradual decline beginning more than a decade before death, with accelerated decline in the few years just before death. This is consistent with prior reports of terminal decline and suggests that terminal decline may in fact account for a considerable proportion of the overall cognitive decline observed in studies of older persons.
Second, this study advances our understanding of the temporal relations of multiple age-related neuropathologies with cognitive decline. By incorporating time-varying effects, we were able to show that the deleterious effects of many age-related neuropathologies are evident more than a decade before death. The effects of Alzheimer’s disease pathology, atherosclerosis, and hippocampal sclerosis emerged relatively earliest, followed by macroscopic infarcts, cerebral amyloid angiopathy and Lewy bodies, respectively. This is consistent with in vivo data demonstrating that Alzheimer’s disease pathology and atherosclerosis manifest years prior to the onset of overt dementia (Villemagne et al., 2011; Rowe et al., 2013). By contrast, the effects of TDP-43 pathology emerged relatively later. These findings are generally consistent with those from recent work in which we examined the temporal relations of neuropathology with four different components of late life cognitive trajectories (i.e. starting level, midpoint of decline, slope of decline, and final level) (Wilson et al., 2016). In that study, most of the age-related neuropathologies were related to the middle or later stages of decline. One exception was that very high levels of TDP-43 pathology were associated with early changes in cognitive function, but moderate and lower levels of TDP-43 pathology were associated with relatively later changes, as observed in the present analyses. Although future work is needed to more securely establish the temporal relations of age-related neuropathologies with cognition, the finding that age-related neuropathologies vary in terms of the timing of their effects on cognition has implications for the administration of interventions to prevent or treat late life cognitive decline.
Third, this study greatly extends our understanding of the nature of the association of age-related neuropathologies with cognitive decline over the course of ageing. In our analyses, many of the age-related neuropathologies (i.e. Alzheimer’s disease pathology, macroscopic infarcts, Lewy bodies, TDP-43 pathology, and hippocampal sclerosis) were associated with progressive cognitive deterioration, with their deleterious effects increasing over time. By contrast, two of the vascular pathologies, atherosclerosis and cerebral amyloid angiopathy, lowered the level of cognition but their effects were relatively stable over time. Two other vascular pathologies (i.e. microscopic infarcts and arteriolar sclerosis) were not associated with the trajectory of global cognition. These findings extend prior work showing that, whereas Alzheimer’s disease and other degenerative pathologies are strongly related to cognitive decline, select vascular pathologies have relatively weaker associations with the cognitive trajectory or are related primarily to level of cognitive function (Boyle et al., 2013; Arvanitakis et al., 2016). Together, these data increase our understanding of the longitudinal footprint associated with various pathologic processes and suggest that some age-related neuropathologies are more potent than others over the course of ageing. Thus, a focus on treatment or prevention of the most aggressive forms of neurodegeneration in old age is likely to yield the greatest benefit in terms of reducing the burden of cognitive decline.
This study has strengths and limitations. Strengths include high rates of follow-up participation and autopsy, reducing bias from selective attrition, and the assessment of cognition using a validated and robust measure for many years prior to death; the availability of dense longitudinal data allows for the precise estimation of the cognitive trajectory. Further, the use of TVEM allowed us to flexibly model the cognitive trajectory and simultaneously consider the dynamic effects of multiple age-related neuropathologies on cognition. A limitation is that the TVEM approach is not based on person-specific trajectories, thereby limiting some inferences. Additionally, as neuropathology data are collected post-mortem, we cannot establish the evolution or presence of these neuropathologies during life. Further, the neuropathologies included in this study were not all assessed with the same degree of precision (e.g. some were measured continuously and others were rated as present or absent; some were measured in multiple brain regions and others were measured in only a single region), and this may have obscured our ability to quantify the impact of some age-related neuropathologies. For example, atherosclerosis and arteriolar sclerosis were each evaluated in one brain region and semiquantitative ratings were dichotomized as present or absent for the analysis, and the inclusion of persons with mild disease may have diluted the effects of those neuropathologies. Notably, however, in supporting analyses, we required moderate or severe ratings of atherosclerosis and arteriolar sclerosis and the results were essentially unchanged (data not shown), thereby lending confidence in our primary results. Finally, some neuropathologies such as white matter disease were not assessed, although we are now collecting data on white matter pathology and will be able to examine its impact when sufficient data accrue. Despite these limitations, our analysis examined multiple neuropathologies simultaneously and revealed that the effects of age-related neuropathologies change over the course of ageing. This is in keeping with the idea that many age-related neuropathologies accumulate slowly over years or decades, but these findings have critical implications for our understanding of the varied impacts of these neuropathologies on late life cognitive trajectories. Thus, although future work is needed to further clarify the temporal and dynamic relations of neuropathology with cognition, the present results begin to address the complexity of when and how the effects of age-related neuropathologies on cognition unfold. Such knowledge ultimately will be needed to guide prevention and treatment efforts to reduce the large and increasing public health burden posed by late life cognitive decline and dementia.
Acknowledgements
The authors are grateful to the participants of the Religious Orders Study and the Rush Memory and Aging Project, and the faculty and staff of the Rush Alzheimer’s Disease Center.
Funding
This research was supported by the National Institute on Aging (R01AG34374, R01AG42210, R01AG17917, P30AG10161, R01AG15819) and the Illinois Department of Public Health.
Glossary
Abbreviation
- TVEM
time-varying effects model
References
- Arvanitakis Z, Capuano AW, Leurgans SE, Bennett DA, Schneider JA. Relation of cerebral vessel disease to Alzheimer’s disease dementia and cognitive function in elderly people: a cross-sectional study. Lancet Neurol 2016; 15: 934–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bennett DA, Schneider JA, Arvanitakis Z, Wilson RS. Overview and findings from the religious orders study. Curr Alzheimer Res 2012a; 9: 628–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bennett DA, Schneider JA, Buchman AS, Barnes LL, Boyle PA, Wilson RS. Overview and findings from the rush Memory and Aging Project. Curr Alzheimer Res 2012b; 9: 646–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bennett DA, Wilson RS, Boyle PA, Buchman AS, Schneider JA. Relation of neuropathology to cognition in persons without cognitive impairment. Ann Neurol 2012c; 72: 599–609. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Boyle PA, Wilson RS, Yu L, Barr AM, Honer WG, Schneider JA. et al. Much of late life cognitive decline is not due to common neurodegenerative pathologies. Ann Neurol 2013; 74: 478–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Boyle PA, Yu L, Nag S, Leurgans S, Wilson RS, Bennett DA. et al. Cerebral amyloid angiopathy and cognitive outcomes in community-based older persons. Neurology 2015; 85: 1930–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dodge HH, Wang CN, Chang CC, Ganguli M. Terminal decline and practice effects in older adults without dementia: the MoVIES project. Neurology 2011; 77: 722–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Driscoll I, Resnick SM, Troncoso JC, An Y, O’Brien R, Zonderman AB. Impact of Alzheimer’s pathology on cognitive trajectories in nondemented elderly. Ann Neurol 2006; 60: 688–95. [DOI] [PubMed] [Google Scholar]
- Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatric Res 1975; 12: 189–98. [DOI] [PubMed] [Google Scholar]
- Hall CB, Ying J, Kuo L, Sliwinski M, Buschke H, Katz M. et al. Estimation of bivariate measurements having different change points, with application to cognitive ageing. Stat Med 2001; 20: 3695–714. [DOI] [PubMed] [Google Scholar]
- Markesbery WR. Neuropathologic alterations in mild cognitive impairment: a review. J Alzheimer’s Dis 2010; 19: 221–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology 1984; 34: 939–44. [DOI] [PubMed] [Google Scholar]
- Nag S, Yu L, Capuano AW, Wilson RS, Leurgans SE, Bennett DA. et al. Hippocampal sclerosis and TDP-43 pathology in aging and Alzheimer disease. Ann Neurol 2015; 77: 942–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nelson PT, Abner EL, Schmitt FA, Kryscio RJ, Jicha GA, Smith CD. et al. Modeling the association between 43 different clinical and pathological variables and the severity of cognitive impairment in a large autopsy cohort of elderly persons. Brain Pathol 2010; 20: 66–79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nelson PT, Schmitt FA, Lin Y, Abner EL, Jicha GA, Patel E. et al. Hippocampal sclerosis in advanced age: clinical and pathological features. Brain 2011; 134 (Pt 5): 1506–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nelson PT, Smith CD, Abner EL, Wilfred BJ, Wang WX, Neltner JH. et al. Hippocampal sclerosis of aging, a prevalent and high-morbidity brain disease. Acta Neuropathol 2013; 126: 161–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rowe CC, Bourgeat P, Ellis KA, Brown B, Lim YY, Mulligan R. et al. Predicting Alzheimer disease with beta-amyloid imaging: results from the Australian imaging, biomarkers, and lifestyle study of ageing. Ann Neurol 2013; 74: 905–13. [DOI] [PubMed] [Google Scholar]
- Schneider JA, Arvanitakis Z, Bang W, Bennett DA. Mixed brain pathologies account for most dementia cases in community-dwelling older persons. Neurology 2007; 69: 2197–204. [DOI] [PubMed] [Google Scholar]
- Shiyko MP, Lanza ST, Tan X, Li R, Shiffman S. Using the time-varying effect model (TVEM) to examine dynamic associations between negative affect and self confidence on smoking urges: differences between successful quitters and relapsers. Prev Sci 2012; 13: 288–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Skrobot OA, Attems J, Esiri M, Hortobagyi T, Ironside JW, Kalaria RN. et al. Vascular cognitive impairment neuropathology guidelines (VCING): the contribution of cerebrovascular pathology to cognitive impairment. Brain 2016, in press. [Epub ahead of print]. [DOI] [PubMed] [Google Scholar]
- Sliwinski MJ SR, Hall RB, Katz M, Verghese J, Lipton RB. On the importance of distinguishing pre-terminal and terminal cognitive decline. Euro Psychol 2006; 11: 172–81. [Google Scholar]
- Sonnen JA, Larson EB, Crane PK, Haneuse S, Li G, Schellenberg GD. et al. Pathological correlates of dementia in a longitudinal, population-based sample of aging. Ann Neurol 2007; 62: 406–13. [DOI] [PubMed] [Google Scholar]
- Villemagne VL, Pike KE, Chetelat G, Ellis KA, Mulligan RS, Bourgeat P. et al. Longitudinal assessment of Abeta and cognition in aging and Alzheimer disease. Ann Neurol 2011; 69: 181–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wilson RS, Barnes LL, Krueger KR, Hoganson G, Bienias JL, Bennett DA. Early and late life cognitive activity and cognitive systems in old age. J Int Neuropsychol Soc 2005; 11: 400–7. [PubMed] [Google Scholar]
- Wilson RS, Beckett LA, Barnes LL, Schneider JA, Bach J, Evans DA. et al. Individual differences in rates of change in cognitive abilities of older persons. Psychol Aging 2002; 17: 179–93. [PubMed] [Google Scholar]
- Wilson RS, Boyle PA, Yu L, Segawa E, Sytsma J, Bennett DA. Conscientiousness, dementia related pathology, and trajectories of cognitive aging. Psychol Aging 2015; 30: 74–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wilson RS, Capuano AW, Bennett DA, Schneider JA, Boyle PA. Temporal course of neurodegenerative effects on cognition in old age. Neuropsychology 2016; 30: 591–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wilson RS, Leurgans SE, Boyle PA, Schneider JA, Bennett DA. Neurodegenerative basis of age-related cognitive decline. Neurology 2010; 75: 1070–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wilson RS, Segawa E, Hizel LP, Boyle PA, Bennett DA. Terminal dedifferentiation of cognitive abilities. Neurology 2012; 78: 1116–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wilson RS, Yu L, Trojanowski JQ, Chen EY, Boyle PA, Bennett DA. et al. TDP-43 pathology, cognitive decline, and dementia in old age. JAMA Neurol 2013; 70: 1418–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yang J, Tan X, Li R, Wagner A. TVEM: time-varying effect model. SAS macro suite users’ guide (version 200). University Park, PA: The Methodology Center, Penn State; 2012. [Google Scholar]