Predictive Factors for Disease Progression in Patients With Early-Onset Alzheimer’s Disease

INTRODUCTION

Alzheimer’s disease (AD) usually occurs at older ages, but sometimes at younger ages, which is called early-onset (EO) AD. EOAD patients are defined as patients with AD who are diagnosed before the age of 65 years. Since they have longer life expectancy and are still included in the productive population, we have no choice but to seriously consider disease progression. Furthermore, it has been reported that EOAD has different clinical manifestations, courses, and neuroimaging findings compared to late-onset (LO) AD. Several studies have reported that EOAD patients tend to show not only more severe impairment in attention, language, apraxia, and visuospatial function [1–5] but also relative sparing of memory function in early stages unlike LOAD [6]. In addition, various studies have confirmed that EOAD patients demonstrate greater hypoperfusion, hypometabolism, gray matter atrophy, white matter atrophy, and destruction in functional connectivity [3, 7–11]. Although there have been many studies on characteristics of EOAD, only a few studies have been conducted on disease progression, and the results were conflicting. Some studies have demonstrated faster disease progression in EOAD than in LOAD [12–17], while others have not [18, 19]. Moreover, there are fewer studies focusing on factors for disease progression. In general, apolipoprotein E (APOE) ɛ4 carrier, female gender, and higher education accelerate cognitive decline in AD [20–22]. A previous study on EOAD has indicated that the APOE ɛ4 noncarrier group progresses faster than the carrier group [13], which is opposite to current concepts. To our knowledge, studies focusing on predictive factors only for EOAD are rare. Therefore, the aim of this study was to investigate the rate of disease progression in Korean patients with EOAD and the influence of various factors, such as onset age, gender, education, and APOE genotype on disease progression.

MATERIALS AND METHODS

RESULTS

Table 1 summarizes the clinical and demographic characteristics of the participants. In 288 patients with EOAD, women were predominant (60.8%). The mean (SD) age at onset and the educational level were 61.5 (5.5) (range, 41–64) years and 8.7 (4.7) years, respectively. The mean (SD) of baseline (at diagnosis) scores were: 19.0 (4.9) for K-MMSE, 15.7 (9.4) for SIADL, 0.9 (0.5) for CDR, and 4.9 (2.7) for CDRSOB. The distribution of CDR revealed 128 of CDR 0.5, 132 of CDR 1, and 28 of CDR 2 (Table 1). The gender proportion of CDR was not so different between male and female group: male group- 47(41.6% in male) for CDR 0.5, 52 (46.0% in male) for CDR 1, and 14 (12.4% in male) for CDR 2 versus female group- 81 (46.3% in female) for CDR 0.5, 80 (45.0% in female) for CDR 1, and 14 (8.0% in female) for CDR 2. Compared with two gender groups, the mean baseline scores of K-MMSE and CDRSOB showed no statistical differences (K-MMSE 19.6 ± 5.0 in male versus 18.5 ± 4.9 in female, p = 0.060; CDRSOB 5.1 ± 2.9 in male versus 4.8 ± 2.5 in female, p = 0.420).

Based on the linear mixed models, EOAD patients showed an annual decline in the K-MMSE score (β [S.E], –1.54 [0.2]/year; 95% confidence intervals (CI), –2.03 to –1.04; p <  0.001) and annual increases in the SIADL score (β [S.E], 3.46 (0.4)/year; 95% CI, 2.51 to 4.41; p <  0.001) and CDRSOB in the score (β [S.E], 1.15 (0.1)/year; 95% CI, 0.90 to 1.40; p <  0.001) after adjustment for age at onset, gender, educational level, follow-up period, and APOE genotype. Follow-up period naturally affected all parameters because AD progresses over time. As shown in Table 2, considering factors except for the follow-up period, the educational level was the strongest factor for changes in all clinical parameters (p <  0.001 for K-MMSE, p = 0.044 for SIADL, and p = 0.035 for CDRSOB). When stratified independently by gender (male versus female), linear mixed models demonstrated that disease progression was faster in females as assessed by SIADL (p = 0.002), while it was not significant when assessed by K-MMSE (p = 0.941) or CDRSOB (p = 0.324). However, there were no differences in the annual changes in all the three parameters according to APOE genotype or onset age. Higher educated patients showed faster disease progression as assessed by K-MMSE, SIADL, and CDRSOB, and female patients showed faster disease progression as assessed by SIADL (Table 3 and Fig. 1). After categorized by CDR, the linear mixed models adjusting for age at onset, gender, educational level, follow-up period, and APOE genotype revealed that annual change of K-MMSE was relatively bigger while annual change of SIADL was comparatively smaller in CDR 2 group compared with CDR 0.5 and 1 groups (K-MMSE: –1.83/year [95% CI, –2.34 to –1.32; p <  0.001] in CDR 0.5 versus –1.52/year [95% CI, –2.09 to –0.96; p <  0.001] in CDR 1 versus –2.79/year [95% CI, –5.11 to –0.48; p <  0.001] in CDR 2; SIADL: 4.36/year [95% CI, 3.34 to 5.38; p <  0.001] versus 4.05/year [95% CI, 2.78 to 5.33; p <  0.001] versus 1.87/year [95% CI, 0.62 to 5.16; p <  0.001]). However, the annual changes of CDRSOB were similar among three groups (1.28/year [95% CI, 1.03 to 1.53; p <  0.001] in CDR 0.5 versus 1.57/year [95% CI, 1.18 to 1.96; p <  0.001] in CDR 1 versus 1.46/year [95% CI, 0.30 to 2.62; p <  0.001]).

DISCUSSION

In this study, we confirmed changes in clinical parameters and factors for disease progression in Korean EOAD. The main finding is that higher-educated and female EOAD patients worsen significantly faster in cognitive and functional ability as assessed by MMSE, SIADL, and CDRSOB, while no association was found with age at disease onset and APOE genotype.

In AD patients, decline of MMSE scores vary from 2.7 to 4.5 points per year among studies: 3–3.4 points per year in the pre-acetylcholine esterase inhibitor (AChEI) era [33, 34] and 2.4 points per year in the post-AChEI era [35]. Approximately 1 point difference in the MMSE score could be considered a general effect of antidementia medication. In EOAD patients, changes of MMSE scores were reported as –2.4 points/year by Van der Vlies et al. [13], –0.82 points/year by Grønning et al. [14], and –1.99 points/year by Zhao et al. [17]. In our study, the annual decline in the K-MMSE score was 1.54 points, which is lower than those of previous studies except for a study by Grønning et al. [14]. A prior Korean AD study by Suh et al. [36] reported an average annual decline in MMSE score of 2.3 points in a community-based sample of 107 patients who did not take any AChEI before or during the study period, which was difficult to compare because there are differences in age and the use of antidementia medication and because the difference in the annual decline in MMSE scores between the two studies are similar (<1 point). Our result of an annual increase in the CDRSOB score of 1.15 points/year seems to be lower than that of a previous study that showed an increase of 3.8 points for 2 years in mild AD [37]. The overall rate of disease progression was not faster than expected. The plausible explanations for this may be that this study was conducted on the basis of a hospital-based population and that our participants appeared to take better care and to have better compliance with antidementia drugs.

Education effect is similar between earlier studies and ours [38–42]. The “cognitive reserve model” offers the key to understanding the education effect in this study. This implies the rate of disease progression is faster in individuals with high cognitive reserve than in those with low cognitive reserve. Consequently, the protective effect of education disappears after disease onset because education can no longer overcome progressive pathologic changes to severe AD [43].

Age at disease onset and APOE genotype did not affect the rate of disease progression. In the perspective of APOE genotype, the negative results between APOE genotype and cognitive decline in our study suggest that APOE genotype be a risk factor for AD occurrence rather than a prognostic factor or can be less pronounced in EOAD as in the results of Davidson et al. [44] and Kleiman et al. [45]. The effect of APOE genotype on disease progression in EOAD patients remains to be clarified considering inconsistent result by previous study [13].

In this study, functional decline was remarkably faster in females than in males, which is in disagreement with the result of a previous study that reported that male gender is an independent risk factor for a decline in the IADL score [46]. The poorer functional ability may be partly explained by Korean traditional culture that elderly men work less and more depend on women at home or slower worsening of cognitive function might be hidden by cultural background that men do not complain of clinical symptoms probably due to denial or anosognosia [47]. It could be excluded that the possibility of much proportion of severe status of dementia in female compared with male, based on the gender proportion of CDR. Thus, elderly women may be more sensitive to changes in the IADL score than elderly men in Korea.

Our result revealed similar progression rate in CDR 0.5 and 1 group while it showed a mixed pattern in CDR 2 group. It was slightly different in CDR 2 group from previous result that has already been reported that there is no significant difference in the rate of disease progression according to disease severity in Korean AD [36], however, we should not jump to a conclusion easily because insufficient numbers of patients with moderate stage EOAD in our study can make the result uncertain.

This study has some limitations. First, this is not a population-based study, which raises the possibility of a bias in patient selection or limits the generalization of our results. Further population-based EOAD studies are warranted. Second, this is not a prospective cohort study, resulting in various follow-up intervals, a relatively low follow-up rate, and a short follow-up period than well-designed cohort studies. Third, we did not confirm the antidementia drug types and compliance in detail. Since these two factors may have influenced the progression rate, the careful interpretation should be needed. Despite the above limitations, the important strength of our study is to provide a basis for expanding our understanding of the clinical course of patients with EOAD in real-life practice. Moreover, since we observed changes in SIADL, CDRSOB, and K-MMSE scores, our results could offer clinicians a realistic combination of the predictors for detecting the disease progression. In conclusion, our results suggest that the rate of disease progression in Korean patients with EOAD may not be so fast in real-life hospital-based clinical practice and that education and female gender rather than APOE genotype or age at onset would be important independent predictors of disease progression in patients with EOAD.