Abstract
Keywords
INTRODUCTION
Dementia, which is characterized by a progressive decline in cognitive function, is a major concern in aging societies. Although a number of treatments have been approved, an effective treatment for the disease has yet to appear. At this stage, early detection of cognitive decline is essential for disease prevention. The condition of mild cognitive impairment (MCI) was first described in 1999 as a common disorder of aging that increases the risk of dementia, including Alzheimer’s disease (AD), and patients with MCI progress to dementia at an annual rate of 15.0% [1]. Shankle et al. [2] demonstrated that it is feasible to apply the most sensitive MCI screening tests in community healthcare settings.
The pathogenesis of AD is thought to involve the accumulation of advanced glycation endproducts (AGEs) [3]. Accumulation of AGEs in cells and tissues is a normal feature of aging but is accelerated in AD [4], suggesting a contribution of AGEs to the pathogenesis of dementia. Recent research has shown that diabetic mellitus (DM) patients have increased levels of AGEs in association with inflammation and oxidative stress [5], leading to dementia [6] and MCI [7]. AGEs in tissues can be estimated using the relatively simple noninvasive measurement of skin autofluorescence (SAF), a method based on the fluorescent properties of some AGEs [8]. A validation study indicated that SAF levels are determined by the accumulation of AGEs and demonstrated a strong correlation between SAF and the skin content of AGEs, including pentosidine, N(ɛ)-(carboxymethyl)lysine (CML), and N(ɛ)- (carboxyethyl) lysine (CEL) [8]. Nevertheless, the possible involvement of AGEs in the pathophysiology of MCI in healthy populations has not been fully investigated.
Here, to better understand the association of MCI with SAF, we examined the involvement of tissue AGE accumulation in MCI in healthy community-dwelling subjects.
Our three hypotheses were as follows: (1) SAF significantly correlates with brain atrophy evaluated by THA; (2) SAF is significantly higher in MCI patients than in normal subjects; and (3) SAF is an independent determinate of MCI in apparently healthy subjects.
METHODS
Subjects and methods
The subjects attended a voluntary medical check-up program, specifically designed to evaluate aging-related disorders, including atherosclerosis, cardiovascular disease, physical function, and cognitive impairment [9–11]. All clinical data were obtained through the checkup process. Inclusion in this study was limited to participants aged 40 years and older who had not received regular medications for hypertension or diabetes mellitus for at least 6 months. To minimize the impact of disease, participants with a history of cardiovascular diseases, including stroke and ischemic heart disease, were excluded. In this cross-sectional observational study, several atherosclerosis-related clinical parameters were measured, including those on brain MRI. A total of 226 subjects participated in the study. Enrollment was between April 2013 and April 2015. This clinical study was conducted with consideration of the protection of subjects as outlined in the Declaration of Helsinki, and was approved by the ethics committee of the Ehime University Graduate School of Medicine. Written informed consent form was obtained from all participants beforeexamination.
Measurement of SAF
The AGE-Reader (DiagnoOptics Technologies BV, Groningen, The Netherlands) is a desktop device that uses the characteristic fluorescent properties of certain AGEs to estimate the accumulation level of AGEs in the skin. All measurements by the AGE-Reader were performed at the volar side of the forearm between 10 and 15 cm below the elbow, as described previously [8]. Measurements were calculated by dividing the mean value of emitted light intensity per nm between 420 and 600 nm by the mean value of excitation light intensity per nm between 300 and 420 nm, expressed in arbitrary units (AU). An earlier validation study showed an intra-individual Altman error percentage of 5.0% per day and 5.9% due to seasonal change [8].
Assessment of MCI
MCI was assessed by the Japanese version of the MCI screening test [12], a 10-minute, computationally scored, staff-administered test. The validity and specificity of this test in differentiating normal aging from MCI have been described elsewhere. Cross-validation has been confirmed using the Clinical Dementia Rating score as a reference. The overall accuracy in discriminating both amnestic and mixed cognitive domain types of MCI from normal aging is 97% [13].
Laboratory tests
Blood samples were collected between 9.00 and 10.00 am from the cubital vein following an overnight fast. Low-density lipoprotein cholesterol (LDL-C) levels were calculated using the Friedewald formula [14]. Estimated glomerular filtration rate (eGFR) was calculated using the Cockcroft–Gault formula [15].
Assessment of baPWV
Brachial ankle pulse wave velocity (baPWV) was measured with a volume plethysmographic apparatus (FORM/ABI; Omron Colin Co. Ltd., Komaki, Japan) as previously described [16]. BaPWV was measured after at least 5 minutes of rest. Measurements were reported as the average value for bilateral measurements.
Assessment of brain atrophy
Atrophy of the brain was assessed by examining the temporal horn area (THA), as previously described [10]. In brief, bilateral THA was measured as an index of medial temporal lobe atrophy at the suprasellar cistern (pentagon) level, using T2-weighted images, with the mean value of the two measurements used in the analyses. MRI assessment of medial temporal lobe atrophy has been shown to be a powerful and independent predictor of progression to dementia in MCI patients [17].
Statistical analysis
All continuous variables were expressed as the mean±standard deviation (SD), unless otherwise indicated. Normal distribution (Kolmogroc-Smirnov test) and homoscedasticity (Levene test) of data were verified. Comparisons between the two groups were assessed using the unpaired t-test for parametric variables and Mann-Whitney U test for nonparametric variables. The chi-square test was used to assess frequency differences. Correlations between variables were evaluated using Pearson’s correlation coefficient. Factors independently associated with MCI were assessed via logistic regressions analysis. In all comparisons, a two-tailed p < 0.05 was considered statistically significant. Analyses were performed using the SPSS software package for Windows version 17 (SPSS, Chicago, IL).
RESULTS
Risk factors for MCI
MCI was diagnosed in 18 (7.9%) of 226 subjects. Clinical characteristics of subjects with and without MCI are summarized in Table 1. The MCI group had a significantly higher age compared with the normal group. Although other known risk factors for dementia, including SBP, did not significant differ between subjects with and without MCI, values for sCr, eGFR, BNP, baPWV, THA, and SAF were significantly higher in those with MCI. In univariate analyses, SAF significantly correlated with age, sCr, eGFR, UACR, BNP, baPWV, and THA (Table 2). There was no significant difference in education level between the two groups.
SAF and MCI
To further investigate whether some parameters could predict the presence of MCI, the prevalence of MCI was analyzed by tertile of the respective index. The prevalence of MCI was significantly different between the tertiles of age, eGFR, BNP, baPWV, THA, and SAF (Fig. 1A–F). To obtain SAF cut-off values, ROC curve analysis for the presence of MCI was performed. The highest sensitivity–(1–specificity) was obtained with SAF at 2.27 AU, and this value was accordingly defined as high SAF associated with the presence of MCI. The odds ratio of high SAF indicating the presence of MCI is summarized in Table 3. Even after adjustment for other possible confounding parameters including age and THA, SAF≥2.27 was significantly related to the presence of MCI (odds, 6.402; 95% CI, 1.590–25.773, p = 0.009).
DISCUSSION
In this study, we analyzed the possible association of SAF and MCI in community-dwelling, middle aged to elderly healthy subjects. Our results supported our three hypotheses, namely that: (1) SAF significantly correlates with brain atrophy evaluated by THA; (2) SAF is significantly higher in MCI patients than in normal subjects; and (3) SAF is an independent determinate of MCI in apparently healthy subjects. These findings that SAF appears to be an independent MCI screening marker for healthy people and that increased SAF might predict the occurrence of dementia suggest that SAF examination should be included in routine medical check-ups.
In particular, our results highlight that SAF may be associated with cognitive decline resulting from non-diabetic AGE production. Although the underlying mechanism by which increased SAF contributes to MCI in healthy subjects is unclear, two clinical studies [18, 19] have reported significant associations between SAF and brain atrophy and cognitive decline in DM patients. Moran et al. [18] demonstrated that increased SAF is associated with brain atrophy independent of age and sex in type 2 DM patients. Spauwen et al. [19] also showed that higher SAF was significantly associated with cognitive decline after adjustment for diabetes, which is consistent with the result of our study. However, after further adjustment for systolic blood pressure, cardiovascular disease and eGFR, the associations were attenuated.
In our study, these associations were still positive after adjustment for several confounding factors, indicating a discrepancy between studies. One possible explanation is the study population; while their study examined patients with type 2 DM and 20% had cardiovascular disease, our study examined non-diabetic subjects. More recently, van Waateringe et al. demonstrated that SAF levels are associated with several clinical and lifestyle factors including age, BMI, and renal function in the non-diabetic condition in accordance with our results [20]. In addition, AGEs are formed not only in the presence of hyperglycemia, but also in diseases associated with high levels of oxidative stress, such as chronic kidney disease (CKD). In this setting, higher circulating AGE levels may cause neurotoxicity via oxidativestress [21].
To our knowledge, this is the first study to demonstrate a highly significant association between MCI and SAF in a non-diabetic healthy population.
CVD risk factors and MCI
It is well known that uncontrolled cardiovascular disease (CVD) risk factors increase the risk of cognitive impairment including MCI. Cerebral hypoperfusion can contribute to structural and functional changes in the brain, which ultimately leads to cognitive impairment. Okamoto et al. showed strong evidence for the relationship between cerebral amyloid angiopathy and microinfarcts with cerebral hypoperfusion as the mediating factor. [22]. Given this perspective, hypertension, DM, dyslipidemia may be potential risk factors for MCI, by affecting cerebral hypoperfusion. However, there are still no definitive conclusions about the relation between CVD risk factors and MCI based on publisheddata [23].
CKD is also a significant risk factor for cognitive impairment. Recently, Torres et al. demonstrated that CKD patients who had an eGFR < 60 mL/min/1.73 m2 were at higher risk of cognitive impairment than those with eGFR≥60 mL/min/1.73 m2 [24]. As for BNP, a 14-year prospective study showed that BNP is an independent risk marker for dementia. Baseline BNP levels were significantly associated with the risk of dementia in the entire study population, adjusted for multiple cardiovascular risk factors [25]. It is considered that arterial stiffness, including baPWV, is a cause of cognitive impairment. Pooled analyses of cross-sectional studies showed that baPWV is a contributor to microvascular brain disease [26]. In our study subjects, although MCI prevalence markedly differed among tertiles of eGFR, BNP, and baPWV, these markers did not have a strong correlation with MCI. On multiple regression analyses suggested that SAF might be the most powerful predictor for occurrence of MCI.
Prevalence of MCI
The estimated prevalence of MCI in adults aged 65 years and older is 10% to 20% [27]. In our study, 7.9% of participants were diagnosed with MCI, which is consistent with another Asian study [28] but lower than that in Australian (17.7%) [29] and US (29%) studies [30].
Limitations
This study has a number of limitations that warrant attention. First, the sample size was small. For this reason, this study was underpowered to examine the breadth of relatively rare (<10% prevalence rate) prevalence of MCI. Second, as we did not measure serum levels of AGEs, some non-fluorescent AGEs might not have been detected by SAF measurement in this study. Third, study findings might not relate to non-Japanese populations and it is uncertain if the data can be extrapolated to other countries. However, it has been reported that SAF strongly correlates with the total AGE content, including non-fluorescent AGEs, measured in skin biopsies of subjects [8].
SAF is a useful variable for assessing the accumulation of AGEs. Our study confirms that SAF is elevated in subjects with MCI, which mainly manifests as brain atrophy. A better understanding of the complexity of the influence of SAF and MCI on the development of cognitive decline will require a larger longitudinal study.
Conclusions
These preliminary results suggest that SAF might be an independent MCI screening marker for healthy people, and that increased SAF might predict the occurrence of dementia. We therefore recommend the inclusion of SAF examination in routine medical check-ups.
