Association of plasma endothelial lipase levels on cognitive impairment

We recruited 97 participants including cognitively normal individuals (N = 39), amnestic MCI (N = 38), and AD-type dementia (N = 20) from the memory clinic in Samsung Medical Center, from February 2017 to August 2017. Additionally, we recruited 29 cognitively normal participants from the local community. Five participants were excluded due to the withdrawal of consent, and a participant was excluded due to the failure of obtaining an amyloid PET scan. Of the remaining 120 participants, 11 subjects with a coefficient of variation of 30% or more in the EL measurements were excluded. A total of 109 elderly participants (mean age; 75.3 years) over 65 years of age were recruited from a clinic and local community (Additional file 1: Figure S1). The subjects consisted of 51 cognitively normal controls (NC), 38 patients with amnestic mild cognitive impairment (aMCI), and 20 patients with AD. All participants underwent neurological examination, neuropsychological (NP) testing, and routine blood tests. Blood tests for all participants included a complete blood count, blood chemistry tests, vitamin B12/folate, syphilis serology, thyroid function tests and APOE genotyping. Participants with current or past neurological or psychiatric illnesses such as major depressive disorders, epilepsy, brain tumors, encephalitis or severe head trauma that would affect cognitive function were excluded. On MRI, patients with structural lesions such as tumors, traumatic brain injuries, or hydrocephalus were also excluded. The institutional review boards approved this study. Written, informed consent was obtained from patients and caregivers.

The NC was volunteered community based elderly who had no history of neurological or psychiatric illness, and had no abnormalities upon neurological examination. The NC exhibited normal cognition on the detailed neuropsychological tests. Patients met the following criteria for aMCI proposed by Petersen et al. [28]: (1) subjective memory complaints by the patient or an informant; (2) relatively normal performance in other cognitive domains; (3) normal activities of daily living (ADL), as judged clinically; (4) objective memory decline below − 1.0 SD on either verbal or visual memory tests; and (5) not demented. AD was diagnosed based on National Institute on Aging-Alzheimer’s Association (NIA-AA) research criteria for probable AD [29].

All participants underwent a standardized neuropsychological battery called the Seoul Neuropsychological Screening Battery, which consisted of tests of attention, language, visuospatial, memory, and frontal/executive functions [30]. Tests that were scored included the following: The Korean version of the Boston Naming Test (K-BNT), Rey-Osterrieth Complex Figure Test (RCFT: copying, immediate and 20-min delayed recall, and recognition), Seoul Verbal Learning Test (SVLT: immediate, 20-min delayed recall and recognition), phonemic and semantic Controlled Oral Word Association Test (COWAT) and the Stroop Test (word and color reading). SVLT and RCFT were performed to assess verbal and nonverbal learning and memory. The K-BNT and RCFT copy were done to evaluate language and visuospatial function, respectively. Phonemic and semantic COWAT and the Stroop Test were performed to evaluate frontal/executive function. Scores were considered abnormal when they were lower than − 1.0 SD below age- and education-adjusted norms. All participants also were administered the Clinical Dementia Rating Scale Sum of Boxes (CDR-SOB). Clinical Dementia Rating (CDR) on 5-point scale was scored using the CDR-SOB. The Korean version of the Mini-Mental Status Examination (K-MMSE, range 0 to 30) was also administered. Tests were administered by experienced staff and supervised by board-certified clinical neuropsychologists.

Patients underwent florbetaben (FBB) PET or flutemetamol (FMM) PET at Samsung Medical Center using a Discovery STe PET/CT scanner (GE Medical Systems, Milwaukee, WI, USA) in three-dimensional scanning mode. ¹⁸F-florbetaben PET was classified as positive when visual assessment was scored as 2 or 3 on brain amyloid-plaque load (BAPL) scoring system [31, 32]. Visual interpretation of ¹⁸F-flutemetamol PET images relied upon a systematic review of five brain regions (frontal, parietal, posterior cingulate and precuneus, striatum and lateral temporal lobes) [33]. More detailed procedures are described in supplementary methods section (Additional file 1: Methods).

The plasma concentrations of EL were measured using the ELISA kit (Cloud-Clone Corp., Houston, TX; SEA469Hu) according to the manufacture’s protocol. Plasma sample was separated with EDTA as anticoagulant. EL release is induced by administration of heparin, but EL levels in pre heparin condition are significantly correlated with levels of post heparin condition [34]. To assess the correlation between plasma EL mass and cognitive impairment, therefore, EDTA-plasma was obtained without administration of heparin. Each level is mean values of the duplicate measurement.

Group differences were calculated using either t-tests or an analysis of variance (ANOVA) followed by post hoc tests for continuous variables and chi-square tests for nominal variables. Linear trend was estimated using the ANOVA results and with a polynomial contrast. Pearson correlation coefficients were used to quantify associations between variables. Simple and multiple logistic regression analyses were conducted to calculate odds ratios (ORs) of cognitive impairment (MMSE score ≤ 25) per 1 standard deviation (SD) increment of plasma EL concentration. The MMSE cut-off chosen for the general cognition function varied from 23 to 26 [35]. We determined MMSE score ≤ 25 as cognitive impairment. All calculations, graphs and statistical analyses were performed using SPSS ver. 19.0 (IBM Corporation, Armonk, NY) and GraphPad Prism 5 (GraphPad Inc., La Jolla, CA).

Demographic and clinical data are shown in Table 1. The mean age of patients with AD was 79.1 years old, which was significantly older than the cognitively normal control (NC) group (73.8 years old, p <  0.001) and MCI group (75.3 years old, p <  0.05). There was no between-group difference in the number of years of education, body mass index (BMI) score or lipid parameters, such as TC, HDL-C, low-density lipoprotein cholesterol (LDL-C) and triglyceride (TG). But between-group difference in sex ratios, total protein or albumin was observed. ApoE4 allele carriers in the AD group comprised 55%, which was significantly more frequent than those in the NC (22%) and MCI (37%) groups. Positive Amyloid PET brain scans were found in 18% of the control group, 47% of the MCI group and 65% of the AD group. The average MMSE score of the MCI group (25.9 ± 2.7) was lower than control group (27.6 ± 2.1; p <  0.05), and the average MMSE score of the AD group (16.6 ± 6.4) was lower than both other groups (both p <  0.001). Clinical dementia rating (CDR) scores were higher in AD group compared to both other groups (both, p <  0.001). However, the difference in CDR score between NC and MCI groups was not observed.

Plasma EL showed no significant between-diagnostic group differences (Table 1). EL was inversely correlated with HDL-C (r = − 0.24, p = 0.015) in the blood lipid profile, and was not correlated with TC, LDL-C, or triglyceride (Additional file 1: Figure S2). Additionally, plasma EL was positively correlated with white blood cell (WBC) and platelet counts. (Additional file 1: Table S1).

The analysis by classification of dementia severity according to CDR scores, showed that there were significant group differences between CDR groups (Fig. 1; one-way ANOVA; p = 0.03). The EL levels were significantly higher in the CDR 1 group (30.5 ± 5.1) than in the CDR 0 (19.4 ± 1.9, p = 0.012), CDR 0.5 (21.4 ± 1.1, p = 0.01) or CDR 2 (15.5 ± 3.9, p = 0.003) groups (Fig. 1). There was no trend of EL concentration in the whole CDR severity range. However, as the severity of dementia increased from CDR = 0 to CDR = 1, the EL concentration tended to increase (p for trend = 0.013; Fig. 1) and elevated EL levels were significantly correlated with MMSE score (r = − 0.29, p = 0.003; Additional file 1: Figure S3). However, the elevated levels of plasma EL were not correlated with the presence of ApoE4 (p = 0.265) or amyloid deposit in the brain (t-test p = 0.199) (Additional file 1: Figure S4).

To determine the association between EL and cognitive impairment (MMSE score ≤ 25), we investigated ORs in subjects from cognitively normal to mild dementia (Table 2 and Table 3). EL concentration was categorized into two ranges according to the mean + 1 standard deviation (31.6 ng/ml). Participants that fell into the upper range of EL had a higher unadjusted OR of cognitive impairment (OR = 5.4, 95% confidence interval [CI], 1.7–17.5) than those that fell into lower range. When adjusted for age, sex, education, BMI, ApoE e4 allele and Amyloid-PET imaging, and additionally adjusted for disease history including diabetes hyperlipidemia, heart disease, hypertension, stoke, traumatic brain injury, the ORs of cognitive impairment in the upper range group compared with the lower range group were similar to those of the unadjusted model (OR = 4.8, 95% CI 1.3–17.9; OR = 5.6, 95% CI 1.4–22.9, respectively)..