RESEARCH ARTICLE

The relationship between cognitive performance and insulin resistance in non-diabetic patients with mild cognitive impairment Tae-Eun Kim1, Dong Hyun Lee1, Yoon-Jeong Kim1, Ji Oh Mok2, Chul- Hee Kim2, Jeong-Ho Park1, Tae-Kyeong Lee1, Kwangsun Yoo3, Yong Jeong3, Yunhwan Lee4 and Sun Ah Park1 1

Department of Neurology, Soonchunhyang University Bucheon Hospital, Bucheon, Republic of Korea Department of Endocrinology, Soonchunhyang University Bucheon Hospital, Bucheon, Republic of Korea 3 Laboratory for Cognitive Neuroscience and NeuroImaging, Department of Bio and Brain Engineering, Korea Advanced Institute of Science and Technology, Daejeon, Republic of Korea 4 Department of Preventive Medicine and Public Health, Ajou University School of Medicine, Suwon, Republic of Korea Correspondence to: S. A. Park, MD, PhD, Professor of Neurology, E-mail: [email protected] 2

Objective: Insulin resistance (IR) is a distinct and early feature of type 2 diabetes mellitus and metabolic syndrome. IR is thought to play a vital role in cognitive impairment. We conducted this study to understand the early characteristics of cognitive dysfunctions attributable to IR. Methods: This study included 85 consecutive non-diabetic elderly participants with mild cognitive impairment (MCI). IR was estimated with the homeostasis model assessment of insulin resistance (HOMA-IR). Cognitive performances were analyzed as a function of scores on the HOMA-IR. Results: The group analysis those with and without IR did not show any differences in the cognitive performance although higher HOMA-IR was closely associated with lower performances in immediate recall on the Seoul Verbal Learning Test (SVLT-I) (r = 0.244, p = 0.026) and Controlled Oral Word Association Test (COWAT) (r = 0.270, p = 0.013). In subgroup analysis by APOE status, SVLT-delayed (p = 0.027) and COWAT (p = 0.016) scores were found to be significantly lower in the IR than the non-IR among those with APOE ε4 allele. In multiple regression analysis, impairment on the COWAT remained significantly correlated with scores on HOMA-IR (β = 0.271, t = 2.340, p = 0.022). However, IR status was identified to interact with APOE ε4 carriership toward poor performances in the COWAT (β = 0.335, t = 2.285, p = 0.026). Conclusion: This study found a domain-specific impact of HOMA-IR scores on cognitive performances in non-diabetic patients with MCI. This association was profound only in APOE ε4carriers. Copyright # 2014 John Wiley & Sons, Ltd. Key words: apolipoprotein E; mild cognitive impairment; insulin resistance History: Received 18 February 2014; Accepted 2 July 2014; Published online 25 July 2014 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/gps.4181

Introduction Insulin resistance (IR) is the most distinctive feature of type 2 diabetes mellitus (T2DM) and emerges much earlier than overt T2DM. Increased serum insulin levels accompanied by decreased cellular responses to insulin characterize IR (Kahn, 2003). As demonstrated by an increase in pro-inflammatory cytokines in cerebrospinal fluid, hyperinsulinemia induces central Copyright # 2014 John Wiley & Sons, Ltd.

nervous system inflammation (Fishel et al., 2005). Additionally, the deficiency of available insulin-degrading enzyme for amyloid β protein (Aβ) degradation induces an increase in Aβ concentrations in the brain (Farris et al., 2003). As molecular evidences of IR are frequently observed in the Alzheimer’s disease (AD) brains (Steen et al., 2005; Liu et al., 2011; Talbot et al., 2012), it appears that, starting at the early stage of AD, IR plays a critical role in cognitive impairment Int J Geriatr Psychiatry 2015 30: 551–557

T.-E. Kim et al.

552

(Talbot et al., 2012). In contrast to the many studies that have explored cognitive deficits in T2DM, very few investigations have examined early-stage IR (i.e. before the onset of T2DM). And previous research on non-diabetic IR has focused on subjects with normal cognition (Kenna et al., 2013) or those with AD (Burns et al., 2012; Pavlik et al., 2013). However, as the results pertaining to AD were contradictory, the following studies including mild cognitive impairment (MCI) patients have been needed (Burns et al., 2012; Pavlik et al., 2013). The current study recruited consecutive patients with MCI to understand the characteristics of the early cognitive deficits that are related to IR. Materials and methods Subjects

This study was approved by the Institutional Review Board of our institute, and all subjects provided written informed consent before enrolling. Participants had been recruited from an ongoing registry, to see the relationship between IR and cognitive impairment. Physical and neurological examinations, comprehensive interviews, neuropsychological testing, brain magnetic resonance imaging (MRI), and blood sampling were performed. The diagnosis of MCI was made based on established criteria (Petersen et al., 2010): complaints of cognitive impairment, scores on objective measures of cognitive impairment 1.0 standard deviation (SD) below the appropriate age- and education-adjusted norms (Ahn et al., 2010), ability to perform the activities of daily living, and failure to meet criteria for a diagnosis of dementia according to the Diagnostic and Statistical Manual of Mental Disorders-IV. We excluded subjects with any of the following: (i) hearing or visual impairment that would interfere with the interviews and neuropsychological testing, (ii) neurological disorders, (iii) psychiatric disorders, (iv) history of use of psychoactive substances or cognitive enhancers that might potentially affect neuropsychological test results, (v) significant systemic disorders that potentially interfere with cognitive functions (e.g. cardiac diseases, respiratory problems, uncontrolled hypertension, malignancy, hepatic diseases, and renal diseases), or (vi) overt T2DM. T2DM was defined as (i) the presence of a history of diabetes and use of antidiabetic medicine, (ii) a fasting blood glucose level equal to or higher than 7.0 mmol/L or higher, or (iii) hemoglobin A1C equal to or higher than 6.5%. Copyright # 2014 John Wiley & Sons, Ltd.

Neuropsychological testing

All participants completed standardized neuropsychological tests including the Korean version of the Boston Naming Test, which assesses naming ability (K-BNT) (Kim and Na, 1999); the Rey Complex Figure Test (RCFT), which assesses visual functioning and visual memory; the Seoul Verbal Learning Test (SVLT), which measures verbal memory; the semantic and phonemic Controlled Oral Word Association Test (COWAT), which assesses semantic and executive functioning; and the Digit Span Test, which measures attention. Overall cognitive performance was evaluated using the Korean version of the Mini-Mental State Examination (K-MMSE) (Kang et al., 1997) and the Clinical Dementia Rating (CDR). Norms according to age, sex, and educational level were based on data obtained from 447 normal subjects (Ahn et al., 2010). Accordingly, raw scores were converted to z-scores based on group means. Composite z-scores were then created for naming, visual functioning, visual memory, verbal memory, attention, and executive functioning by averaging the z-transformed data of individual tests. The composite z-scores were used to validate differences between groups. Measurements of IR-related blood parameters

IR was validated with the ratio of the fasting-state levels of insulin and glucose using the homeostasis model assessment of insulin resistance (HOMA-IR). Blood samples were collected from venous blood after overnight fast. Fasting blood glucose levels were measured by the glucokinase method, and insulin and C-peptide levels were measured by radioimmunoassay. HOMA-IR scores were obtained with the following formula: HOMA-IR = fasting insulin (μIU/mL) × fasting plasma glucose (mmol/L) / 22.5 (Matthews et al., 1985). Apolipoprotein- ε4 (APOE ε4) genotyping

Genomic DNA was extracted from blood samples using a DNA extraction kit (Qiagen, Germantown, MD, USA). For APOE genotyping, polymerase chain reactions were performed using APOE genotyping PrimerMix Kit (Bio-Core, Seoul, Korea) following the manufacturer’s recommendations. Each reaction mixture was heated at 95 °C for 12 min, followed by 35 cycles of amplification (94 °C for 30 s, 63 °C for 30 s, and 72 °C for 45 s) and a final extension step for 5 min at 72 °C. The PCR products were evaluated by electrophoresis on a 2% agarose gel. Int J Geriatr Psychiatry 2015 30: 551–557

Cognitive performance and insulin resistance

Brain MRI and volumetric measures

Brain MRI was performed in all subjects using a 3.0 Tesla MRI scanner (Signa HDxt, GE Healthcare). Threedimensional, T1-weighted images were taken with the following imaging parameters: 1-mm coronal slice thickness, TR/TE = 6.7/2.9 ms, 12° flip angle, 1 number of excitations, 20 × 20 cm field of view, 31.25 kHz bandwidth, and 256 × 256 pixels matrix size. Hippocampal volumes were automatically measured using T1 MRI data with FreeSurfer (Athinoula A. Martinos Center for Biomedical Imaging, Charlestown, MA, USA), as previously described (Fischl et al., 2001; Fischl et al., 2002; Fischl et al., 2004). The T1 image was processed based on the default setting of the reconstruction process provided in FreeSurfer. Briefly, motion correction, removal of nonbrain tissue, Talairach transformation, segmentation of subcortical white matter and deep gray matter volumetric structures, intensity normalization (Sled et al., 1998), tessellation of the gray matter/white matter boundary, and automated topology correction (Fischl et al., 2001; Segonne et al., 2007) were processed. Statistical analyses

Statistical analyses were conducted using SPSS Ver. 17.0 for Windows (Chicago, IL, USA). Correlations involving data from all subjects were analyzed using Pearson’s correlation coefficient. Non-IR (HOMA-IR scores ≤2.61) and IR (HOMA-IR scores >2.61) were defined according to the HOMA-IR cut-off values for dichotomized groups (Ascaso et al., 2003). Comparisons were performed for the two groups using independent t-tests for continuous variables and chi-square tests for categorical variables. Subgroup analyses were performed after incorporating APOE ε4 carrier status. Due to the small number of subjects, mean comparisons of APOE ε4 carriers were conducted using the Mann–Whitney test and Fisher’s exact test. Associations between cognitive performance and HOMA-IR scores were examined with multiple regression after age, sex, educational level, hypertension, and APOE ε4carrier status were included as covariates. The interaction between APOE ε4 and IR was tested by including a product term in the regression model. A p-value 2.61) (Ascaso et al., 2003). The two groups did not differ in terms of age, sex, educational level, or presence of hypertension. Further, the differences in Table 2 The correlations of HOMA-IR with neuropsychological and neuroimaging data

NP tests K-MMSE K-BNT RCF T-copy SVLT-I SVLT-D RCFT-I RCFT-D COWAT Digit span Neuroimaging Lt H vol (mL) Rt H vol (mL)

Mean ± SD

Correlations

23.5 ± 3.9 0.9 ± 1.7 1.0 ± 2.9 1.1 ± 1.0 1.2 ± 1.3 1.1 ± 1.0 1.1 ± 1.0 0.6 ± 0.9 0.1 ± 1.1

r = 0.053, p = 0.630 r = 0.102, p = 0.354 r = 0.147, p = 0.181 r = 0.244, p = 0.026* r = 0.152, p = 0.165 r = 0.106, p = 0.338 r = 0.150, p = 0.174 r = 0.270, p = 0.013* r = 0.080, p = 0.479

3624 ± 1036 3565 ± 1404

r = 0.097, p = 0.564 r = 0.119, p = 0.476

*p < 0.05. r, Pearson’s correlation coefficient; K-MMSE, Korean version of mini-mental state examination; K-BNT, Korean-Boston Naming Test; RCFT-Copy, Rey Complex Figure Test- copy; SVLT-I, Seoul Verbal Learning Test- immediate recall; SVLT-D, Seoul Verbal Learning Test- delayed recall; COWAT, Controlled Oral Word Association Test; H, hippocampus; vol, volume. p < 0.05 is considered statistically significant.

Int J Geriatr Psychiatry 2015 30: 551–557

T.-E. Kim et al.

554

z-scores of individual neuropsychological tests did not reach statistical significance (Supplementary table 1). Previous reports have demonstrated the augmenting effect of APOE ε4 on cognitive decline and the development of AD in patients with T2DM (Peila et al., 2002; Irie et al., 2008). Thus, the same analysis after stratifying the sample according to APOE ε4 carrier status was conducted. Because our sample contained few APOE ε4 carriers, non-parametric analysis was performed in this group. APOE ε4 carriers in the IR group obtained lower scores on the delayed recall of verbal items (SVLT-D) and COWAT than did carriers in the non-IR group (Table 3). However, the comparable groups of non-carriers did not differ significantly in cognitive performance. Subsequent multiple regression analysis demonstrated that COWAT scores remained significantly correlated with HOMA-IR scores (β = 0.271, p = 0.022) when age, sex, hypertension, educational level, and APOE ε4 status were considered as covariates (Table 4).

Because the domain-specific deteriorations depending on IR were evident in APOE ε4 carriers, the interaction between APOE ε4 carrier status and IR on cognitive domains was assessed. In COWAT scores (β = 0.335, p = 0.026), a significant interaction between APOE ε4 carrier status and IR was identified. Discussion In this study, we evaluated the contribution of IR to cognitive decline in non-diabetic MCI patients. Unlike most of the literatures exploring the role of IR in cognitive impairment (Kuusisto et al., 1997; den Heijer et al., 2003; Kanaya et al., 2004; Geroldi et al., 2005), our study excluded various confounding factors related to T2DM by including only non-diabetic patients. Specifically, the cerebrovascular disorders that frequently accompany T2DM, systemic diabetic complications, and anti-diabetic medication-related effects can affect cognitive functions. Additionally, by focusing on MCI, we

Table 3 The comparisons of neuropsychological and neuroimaging data depending on insulin resistance incorporating APOEε4 carriership APOE ε4-

APOE ε4+

HOMA-IR ≤2.61 (n = 28)

HOMA-IR >2.61 (n = 30)

p-value

HOMA-IR ≤2.61 (n = 10)

HOMA-IR >2.61 (n = 8)

p-value

Laboratory Glucose Insulin C-peptide HOMA-IR

92.8 ± 10.4 6.6 ± 2.8 1.7 ± 1.1 1.5 ± 0.7

103.5 ± 14.2 26.1 ± 20.1 4.9 ± 3.9 7.0 ± 6.2

0.002

The relationship between cognitive performance and insulin resistance in non-diabetic patients with mild cognitive impairment.

Insulin resistance (IR) is a distinct and early feature of type 2 diabetes mellitus and metabolic syndrome. IR is thought to play a vital role in cogn...
141KB Sizes 0 Downloads 7 Views