Elevated Plasma Free Fatty Acid Susceptible to Early Cognitive Impairment in Type 2 Diabetes Mellitus

Abstract

Background:

Elevated free fatty acid (FFA) induces lipotoxicity, attributed to diabetes and cognitive decline. Sterol regulatory element-binding protein-1c (SREBP-1c) regulates lipid metabolism.

Objective:

We investigated the roles of FFA in mild cognitive impairment (MCI) of type 2 diabetes mellitus (T2DM) patients and determine its association with rs11868035 polymorphism.

Methods:

We recruited 191 Chinese T2DM patients into two groups through Montreal Cognitive Assessment. Demographic and clinical data were collected, multiple domain cognitive functions were tested, plasma FFA levels were measured through ELISA, and SREBP-1c rs11868035 genotype was detected using the Seqnome method.

Results:

In comparison with the healthy-cognition group (n = 128), the MCI group (n = 63) displayed lower glucose control (p = 0.012) and higher plasma FFA level (p = 0.021), which were independent risk factors of MCI in T2DM patients in multivariate regression analysis (OR = 1.270, p = 0.003; OR = 1.005, p = 0.036). Additionally, the plasma FFA levels of MCI patients were positively correlated with Stroop color word test-C time scores (r = 0.303, p = 0.021) and negatively related to apolipoprotein A1 levels (r = –0.311, p = 0.017), which are associated positively with verbal fluency test scores (r = 0.281, p = 0.033). Both scores reflected attention ability and executive function. Moreover, the G allele carriers of rs11868035 showed higher digit span test scores than non-carriers in T2DM patients (p = 0.019) but without correlation with plasma FFA levels.

Conclusion:

In T2DM, elevated plasma level of FFA, when combined with lower apolipoprotein A1 level portends abnormal cholesterol transport, were susceptible to early cognitive impairment, especially for attention and execution deficits. The G allele of SREBP-1c rs11868035 may be a protective factor for memory.

Keywords

Free fatty acid lipid profile mild cognitive impairment sterol regulatory element-binding protein-1c type 2 diabetes mellitus.

INTRODUCTION

The global prevalence of diabetes has been rising over recent decades [1], with 693 million people expected to suffer diabetes by 2045 [2]. Furthermore, relevant epidemiological studies have manifested that type 2 diabetes mellitus (T2DM) is the most common cause of mild cognitive impairment (MCI) [3], with an approximately 56%increased risk of developing Alzheimer’s disease (AD) [4], thus seriously affecting the quality of life. MCI, the transitional stage between normal cognition and AD dementia, plays a key role in the early warning of AD occurrence [5, 6]. Hence, finding related risk factors of MCI in patients with diabetes can effectively prevent the progression of diabetes-related AD in the early stage. However, the exact mechanism of cognitive decline caused by T2DM has not been completely understood and requires further studies.

Lipid metabolism disorder plays significant parts in the development of diabetic cognitive impairment [7 –9]. Elevated levels of triglycerides (TG), total cholesterol (TC), low-density lipoprotein cholesterol (LDL-c), and apolipoprotein B (ApoB), and decreased levels of high-density lipoprotein cholesterol (HDL-c) and apolipoprotein A1 (ApoA1) can contribute to abnormal lipid metabolism, which is a risk factor for T2DM and AD [10 –18]. However, the normalization of the above blood lipid indicators does not mean that it can effectively protect the cognitive function of patients with diabetes. In patients with type 2 diabetes, elevated free fatty acid (FFA) contributes to lipotoxicity [19], involved in various complications of diabetes, including cognitive decline. In vitro studies have confirmed that FFA stimulates the formation of Aβ and hyperphosphorylation of tau protein, which were the early pathological characteristics in AD [20 –24]. Similar results from many clinical trials have also indicated that FFA levels may affect the development of AD [25]. Additionally, FFA, as a non-esterified fatty acid, can enter the brain by passive diffusion or by protein-mediated transport with a protein such as CD36 and hyperglycemia increased blood–brain barrier permeability by affecting vascular endothelial function [26 –28]. Elevated plasma FFA levels in patients with diabetes may affect the levels of FFA in the brain, increase the risk for AD and may be an important risk for predicting cognitive impairment in diabetes. Together, the chronic elevated FFA concentration may explain the increased cognitive impairment in patients with T2DM.

Sterol regulatory element-binding protein (SREBP), the lipid synthesis transcription factor of cholesterol and fatty acid synthesis, has three isoforms, namely, SREBP-1a, -1c, and -2. Among these isoforms, SREBP-1c is located in the proximal short arm of chromosome 17 (17p11.2) and mainly regulates fatty acid metabolism [29 –31]. In a study of single nucleotide polymorphisms (SNP) rs11868035 of SREBP-1c gene, Liu et al. found that the insulin resistance index of the rare homozygotes GG of rs11868035 was significantly lower than that of AA in the T2DM group [32]. Additionally, rs11868035 polymorphisms exhibited significant association with plasma TC levels and insulin resistance (IR) and played an important role in the mechanism of diabetes-related cognitive dysfunction [33 –35]. In patients with T2DM, SREBP-1c SNP (rs11868035) is associated with IR and regulates the synthesis of FFA, which directly affects Aβ deposition and tau protein phosphorylation and is attributed to IR. Hence, a correlation might be present between SREBP-1c SNP (rs11868035) and FFA in the development of MCI in T2DM.

The present study aimed to detect the plasma FFA concentration and SREBP-1c SNP (rs11868035) in patients with T2DM with or without MCI to investigate the roles of plasma FFA levels and SREBP-1c SNP (rs11868035) in cognitive impairment and further explore the correlation among FFA, lipid profile, and rs11868035 polymorphisms. The results aim to elucidate the lipid metabolism combined with FFA susceptible to cause cognitive impairment and determine whether rs11868035 is involved in regulation. Our findings may provide evidence that FFA causes cognitive impairment in T2DM.

MATERIALS AND METHODS

Participation criteria

From 2015 to 2018, the study was conducted at the Department of Endocrinology, Affiliated ZhongDa Hospital of Southeast University involving 191 Chinese Han hospitalized patients aged 40–80 years, who not only met the standards of T2DM established by the World Health Organization in 1999 [36], but also had disease duration of T2DM more than 3 years. In addition, participants with any of the following conditions were excluded: 1) acute complication of diabetes such as diabetic ketoacidosis in the past 3 months, 2) cerebrovascular accident confirmed by cranial imaging scan in the past year, 3) clear diagnosis of degenerative diseases of the nervous system, such as AD, 4) visual and auditory resolution defects that are difficult to complete neuropsychological testing, 5) application of psychiatric drugs in the past 3 months, 6) systemic diseases such as malignant tumors, anemia, and serious infections, and 7) thyroid disease. All patients with MCI met the following guidelines proposed by the National Institute on Aging Alzheimer’s Association: 1) declined cognitive function reported by self, informant, or clinician, 2) one or more cognition-domain impairment diagnosed with objective evidence such as Montreal Cognitive Assessment (MoCA), 3) normal daily living as measured by questionnaire of activities of daily living, and 4) absence of dementia [37]. In addition, all individuals were Chinese Han, their participation was approved by the Research Ethics Committee of the Affiliated ZhongDa Hospital of Southeast University, and informed consent of the participants were obtained. The trial registration is entitled “Advanced Glycation End Products Induced Cognitive Impairment in Diabetes: BDNF Signal Meditated”.

Demographic and clinical characteristics collection

Sociodemographic characteristics, including age, gender, education level, history of smoking and drinking, diabetes duration, and hypertension duration, were obtained by standard interview. Information about medical history, such insulin usage and metformin usage, were self-reported or confirmed during hospitalization. Physical attributes, including blood pressure, height, and weight, were measured using standard methods. The body mass index (BMI) was calculated by dividing the body weight by the square of height (kg/m²), and hypertension was considered under the following criteria: 1) For patients who do not use high-blood-pressure medications, systolic blood pressure (SBP)≥140 mmHg, and/or diastolic blood pressure (DBP)≥90 mmHg and 2) history of hypertension and receiving hypertensive drugs. Clinical characteristics, including postprandial blood glucose (PBG), fasting blood glucose (FBG), glycosylated hemoglobin (HbA1c,%), TC, TG, LDL-C, HDL-C, ApoA1, and ApoB were measured in the laboratory center of the Affiliated Zhongda Hospital of Southeast University. Standardized quality control procedures were implemented as directed by the Chinese Laboratory Quality Control.

Neuropsychological test data

For each enrolled volunteer, multiple cognitive domain tests, including MoCA, Mini-Mental State Examination (MMSE), Trail Making Test-A and B (TMT-A and B, respectively), Verbal Fluency Test (VFT), Digit Span Test (DST), Clock Drawing Test (CDT), Auditory Verbal Learning Test (AVLT), and Stroop Color Word Test (SCWT), were performed to comprehensively evaluate their executive function, semantic memory, verbal and visual functions, attention, visuospatial skills, and psychomotor speed. These tests were conducted by an experienced neuropsychiatrist, who had no knowledge of the study design.

Measurement of plasma FFA

Patients with T2DM were recruited after hospitalization the next morning and fasting for 12 h. Venous blood (2 ml) was extracted into tubes containing EDTA, and then centrifuged for 15 min at 2,000 rpm. The upper plasma fraction and lower protein components were collected respectively, and frozen at –80°C immediately. The FFA concentration was detected by plasma fraction by using ELISA kits (Jin Yibai Biological Technology, Nanjing, China) according to the manufacturer’s instructions.

Genotyping of SREBP-1c gene rs11868035 polymorphism

Genomic DNA was extracted from the stored protein components by using a DNA purification kit (Puregene, Gentra Systems, Minneapolis, MN, USA), according to the manufacturer’s instructions. Then, genomic DNA was sent to CNKINGBIO (Beijing, China) to detect the genotypes of SREBP-1c rs11868035 through the Sequenom method by using the Agena Bioscience’s MassARRAY platform.

Statistical analysis

Statistical analyses were conducted using SPSS version 23.0 (SPSS Inc., Chicago, IL, USA). Data and parameters were reported as mean±standard deviations, medians (interquartile range) as appropriate, or percentages. Variables with normal or homoscedastic distribution were compared using Student’s t test and analysis of variance, whereas asymmetrical or heteroscedastic variables were analyzed using non-parametric Mann–Whitney U and Kruskal–Wallis tests. The Chi-squared (χ²) test was used to compare the distributions of categorical variables or test for the allelic and genotypic distributions in Hardy–Weinberg equilibrium. Pearson or Spearman rank correlation analysis was conducted to determine the relationship between plasma FFA levels and multiple cognitive domains and lipid profiles. Logistic regression analysis was performed to explore the relationship between demographic characteristics and the plasma FFA and to assess the risk of cognitive impairment in patients with T2DM. Statistical significance was considered at p < 0.05 (two-side).

RESULTS

Demographic, clinical, and cognitive characteristics of the participants

Participant characteristics, demographic, clinical, and neuropsychological test data are summarized in Table 1. The 191 participants were divided into two groups, namely, the MCI group (n = 63) and healthy-cognition group (n = 128), based on MoCA, which is a highly sensitive method (The MoCA score of healthy cognition is≥26, education years less than 12 years plus 1 point). First, our results show that the incidence of MCI in T2DM was approximately 32.98%. Then, compared with the healthy-cognition group, patients with MCI showed substantially increased age, percentage of female, hypertension duration, diabetes duration, HbA1c, TC, and ApoA1 but significantly reduced education level (all p < 0.05). The MCI group showed significantly higher levels of plasma FFA than the healthy-cognition group (430.75±98.66 versus 401.20±74.20μmol/L, p = 0.021). The neuropsychological test scores of multiple cognitive domains were significantly lower in the MCI group than in the healthy-cognition group (all p < 0.05). No significant differences were identified in terms of smoking, drinking, BMI, PBG, FBG, SBP, DBP, TG, LDL-c, HDL-c, ApoB, coronary heart disease, diabetic nephropathy, diabetic retinopathy, diabetic peripheral neuropathy, the percentage of insulin, or lipid-lowering drugs use (all p ≥ 0.05).

Table 1

Demographic, clinical and cognitive characteristics

	MCI group (n = 63)	Control group (n = 128)	p
Age (y)	62.05±8.69	59.38±8.69	0.047¹^*
Female, n (%)	37 (58.73%)	43 (33.60%)	0.001³^*
Education Level (y)	9.29 (8.00–12.00)	10.99 (9.00–12.00)	0.001²^*
Smoking, n (%)	15 (23.81%)	42 (32.80%)	0.201³
Drinking, n (%)	7 (11.11%)	27 (21.10%)	0.090³
BMI (kg/m²)	25.02 (23.01–26.91)	24.64 (22.61–26.67)	0.636²
Hypertension, n (%)	42 (66.67%)	65 (50.80%)	0.038³^*
SBP (mmHg)	136.76 (124.00–148.00)	138.61 (122.25–150.00)	0.729²
DBP (mmHg)	82.21 (78.00–89.00)	82.34 (74.00–90.00)	0.803²
Hypertension duration (y)	9.71 (0.00–15.00)	6.26 (0–10.00)	0.020²^*
Diabetes duration (y)	12.56 (8.00–15.00)	10.61 (6.00–13.00)	0.029²^*
Use of insulin, n (%)	31 (50.79%)	82 (64.10%)	0.050³
Use of metformin, n (%)	38 (60.32%)	88 (68.75%)	0.248³
Use of statins, n (%)	39 (20.4%)	61 (31.9%)	0.064³
Use of statins and fibrates, n (%)	39 (20.4%)	65 (34%)	0.147³
CHD	6 (3.1%)	13 (6.8%)	0.891³
DN	7 (3.7%)	5 (2.6%)	0.054³
DR	4 (2.1%)	2 (1.0%)	0.075³
DPN	8 (4.2%)	11 (5.8%)	0.372³
HbA1c (%)	9.52 (8.00–10.80)	8.67 (7.33–9.48)	0.012²^*
FBG (mmol/L)	8.42 (6.63–10.09)	8.27 (6.20–9.70)	0.594²
PBG (mmol/L)	15.40 (12.50–18.00)	14.57 (11.80–16.90)	0.384²
TG (mmol/L)	1.83 (0.89–2.19)	1.70 (0.95–2.13)	0.597²
TC (mmol/L)	4.80±1.09	4.46±1.08	0.042¹^*
LDL-c (mmol/L)	2.92±1.00	2.75±0.82	0.214¹
HDL-c (mmol/L)	1.23 (1.01–1.40)	1.15 (0.94–1.33)	0.060²
TC/LDL-c	1.76 (1.49–1.83)	1.66 (1.51–1.79)	0.950²
HDL-c/TC	0.27 (0.20–0.31)	0.27 (0.22–0.30)	0.638²
HDL-c/LDL-c	0.48 (0.33–0.62)	0.45 (0.33–0.52)	0.960²
ApoA1 (g/L)	1.23±0.28	1.14±0.29	0.037¹^*
ApoB (g/L)	0.89 (0.67–1.06)	0.83 (0.67–0.94)	0.320²
FFA (μmol/L)	430.75±98.66	401.20±74.20	0.021¹^*
Cognition test scores
MoCA	22.84 (22.00–25.00)	27.77 (27–29)	<0.001²^*
MMSE	26.65 (25.00–28.00)	29.12 (29–30)	<0.001²^*
DST	10.79 (10.00–12.00)	12.39 (11–13)	<0.001²^*
VFT	14.75 (13.00–17.00)	18.26 (15–21)	<0.001²^*
CDT	3.11 (2.00–4.00)	3.59 (3.00–4.00)	<0.001²^*
TMTA	74.41 (56.00–84.00)	59.57 (45.00–62.75)	<0.001²^*
TMTB	199.38 (158.00–236.00)	145.76 (98.50–167.50)	<0.001²^*
AVLT IR	15.60 (13–17)	19.44 (16–23)	<0.001²^*
AVLT DR	4.60 (3–6)	6.41 (5–8)	<0.001²^*
SCWT A (time)	37.89 (27–50)	30.99 (24–36)	0.006²^*
SCWT A (number)	49.78 (50–50)	49.62 (50–50)	0.012²^*
SCWT B (time)	61.40 (44.00–74.00)	48.11 (37.00–55.25)	<0.001²^*
SCWT B (number)	47.78 (47–50)	49.08 (49–50)	<0.001²^*
SCWT C (time)	113.46 (82.00–136.00)	88.32 (69.25–106.00)	<0.001²^*
SCWT C (number)	44.37 (43–48)	47.36 (46–50)	<0.001²^*

The data was presented as mean±SD, n (%), or the median (interquartile range) unless otherwise specified. BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; CHD, coronary heart disease; DN, diabetic nephropathy; DR, diabetic retinopathy; DPN, diabetic peripheral neuropathy; HbA1c, glycosylated hemoglobin; FBG, fasting blood-glucose; PBG, postprandial blood glucose; TG, triglyceride; TC, total cholesterol; LDL-c, low density lipoprotein cholesterol; HDL-c, high density lipoprotein cholesterol; ApoA1, Apolipoprotein A1; ApoB, Apolipoprotein B; FCP, fasting C-peptide; HOMA-IR, homeostatic model assessment of insulin resistance; FFA, free fatty acid; MoCA, Montreal Cognitive Assessment; MMSE, Mini-Mental State Examination; TMT-A and TMT-B, Trail Making Test-A and Trail Making Test-B, respectively; AVLT-IR and AVLT-DR, Auditory Verbal Learning Test-Immediately Recall and Auditory Verbal Learning Test-Delayed Recall, respectively; CDT, Clock Drawing Test; DST, Digit Span Test; VFT, Verbal Fluency Test; SCWT, Stroop Color Word Test; MCI, participants with mild cognitive impairment; Non-MCI, participations without MCI. ¹Student’s t test was employed for normally distributed variables. ²The Mann-Whitney U test was employed for asymmetrically distributed variables. ³The Chi-square test was employed for categorical variables. ^*Significance, p < 0.05.

Partial correlation analysis between plasma FFA levels and cognitive performances and lipid profiles

The partial correlation analysis of plasma FFA concentrations with multiple neuropsychological tests and lipid metabolism profile (TG, TC, HDL-c, LDL-c, ApoA1, and ApoB) were presented in Table 2. The demographic variables were controlled for age, gender, education level, duration of diabetes, and hypertension. The results showed a positive correlation between the plasma FFA level and SCWT C time in the MCI group (r = 0.303, p = 0.021), and this condition primarily reflected patients’ execution and selective attention function, neuronal plasticity, and dorsolateral prefrontal cortical activity. However, the plasma FFA levels were negatively correlated with HDL-c and ApoA1 levels. Additionally, ApoA1 levels were positively related to VFT scores, which mainly reflected executive function and attention ability. However, no association was observed between HDL-c and cognitive function (all p > 0.05).

Table 2

Partial correlations of plasma FFA concentrations and lipid metabolism with neuropsychological tests in T2DM patients with MCI

	FFA		ApoA1		HDL-c
	r	p	r	p	r	p
MoCA	0.112	0.402	0.132	0.325	0.005	0.969
MMSE	–0.082	0.539	0.079	0.557	0.073	0.587
DST	0.013	0.922	0.179	0.180	0.195	0.142
VFT	0.046	0.731	0.281	0.033^*	0.225	0.090
CDT	–0.015	0.913	–0.168	0.207	–0.148	0.268
TMTA	0.012	0.931	0.123	0.356	0.118	0.379
TMTB	0.054	0.688	0.072	0.590	0.106	0.430
AVLT IR	–0.046	0.731	0.101	0.451	0.068	0.614
AVLT DR	0.116	0.387	–0.012	0.927	–0.061	0.648
SCWT A time	–0.130	0.331	0.004	0.978	–0.144	0.280
SCWT A number	0.061	0.647	0.159	0.234	–0.047	0.727
SCWT B time	–0.003	0.987	–0.063	0.637	–0.110	0.409
SCWT B number	–0.005	0.969	0.156	0.242	–0.027	0.842
SCWT C time	0.303	0.021^*	–0.177	0.183	–0.192	0.149
SCWT C number	–0.121	0.366	–0.056	0.677	–0.255	0.053
TG	–0.067	0.616	–0.151	0.258	0.098	0.465
TC	0.017	0.899	0.144	0.392	0.363	0.005^*
LDL-c	–0.072	0.589	–0.086	0.522	0.083	0.536
HDL-c	–0.273	0.038^*	0.705	<0.001^*	-	-
ApoA1	–0.311	0.017^*	-	-	0.706	<0.001^*
ApoB	0.087	0.516	–0.072	0.591	0.155	0.245

Partial correlation was used to control variables, the age, gender, education level, duration of diabetes, and duration of hypertension. MoCA, Montreal Cognitive Assessment; MMSE, Mini-Mental State Examination; TMT-A and TMT-B, Trail Making Test-A and Trail Making Test-B, respectively; AVLT-IR and AVLT-DR, Auditory Verbal Learning Test-Immediately Recall and Auditory Verbal Learning Test-Delayed Recall, respectively; CDT, Clock Drawing Test; DST, Digit Span Test; VFT, Verbal Fluency Test; SCWT, Stroop Color Word Test; HOMA-IR, homeostatic model assessment of insulin resistance; FFA, free fatty acid; HDL-c, high density lipoprotein cholesterol; ApoA1, Apolipoprotein A1; MCI, mild cognitive impairment. ^*Significance, p < 0.05.

Linear regression analysis

Multi-variable linear regression analysis explored the optimal model of effecting SCWT C (time) scores and the extent in which FFA affected attention in patients with T2DM having MCI. All the variables in Table 1, including age, education level, gender, smoking, drinking, BMI, FBG, PBG, SBP, DBP, HbA1c, TG, TC, LDL-c, HDL-c, ApoB, ApoA1, hypertension duration, diabetes duration, plasma FFA, and the use of insulin and metformin, were considered independent variables, while SCWT C (time) was considered as the dependent variable, and the stepwise approach was adopted. In the optimal model, the elevated plasma FFA levels, non-use of metformin, and high BMI were the main risk factors of attention (total R² = 0.212, p = 0.003). FFA could explain 9.4%of the factors affecting attention, the non-use of metformin could explain 5.9%of attention deficit, and the proportion of BMI affecting attention reached 6.0%(Table 3).

Table 3

Assessment results of rick of SCWT C (time) scores in a multivariable linear regression models in T2DM patients with MCI

	Model 1		Model 2		Model 3
	B	95%CI	B	95%CI	B	95%CI
FFA (μmol/L)	0.307^*	0.025–0.213	0.43^*	0.029–0.213	0.344^*	0.043–0.223
Use of melformin (%)			–0.242^*	–36.782– –0.299	–0.350^*	–46.288– –7.496
BMI (kg/m²)					0.270^*	0.199–6.490
R²	0.094		0.153		0.213
p	0.014^*		0.007^*		0.003^*

BMI, body mass index; FFA, free fatty acid. ^*Significance, p < 0.05.

In the multi-variable linear regression analysis, the optimal model of lipid metabolism effecting plasma FFA levels was explored by setting ApoA1 as the risk factor of FFA (r² = 0.079, p = 0.015, Table 4).

Table 4

Assessment results of rick of abnormal plasma FFA levels in a multivariable linear regression models in T2DM patients with MCI

	B	95%CI
ApoA 1	–0.306^*	–190.756– –21.497
R²		0.079
p		0.015^*

^*Significance, p < 0.05.

Distributions of rs11868035 genotypes and allele frequencies in MCI and healthy-cognition groups

The distributions of rs11868035 genotype and allele frequencies in the two groups are shown in Table 5. However, no remarkable differences were observed between the two groups even after adjustment for gender, age, and educational level (all p < 0.05).

Table 5

Distributions of rs11868035 genotypes and allele frequencies between groups

rs 11868035 genotype	MCI n (%)	Non-MCI n (%)	Crude OR (95%CI)	p	Adjusted OR (95%CI)^b	p
Overall	63	128
A	96 (76.19)	202 (78.91)	1.000
G	30 (24.39)	54 (21.09)	1.169 (0.703–1.943)	0.547
AA	38 (60.32)	84 (65.63)	1.000
AG	20 (31.75)	34 (26.56)	1.300 (0.664–2.547)	0.443	1.538 (0.745–3.176)	0.244
GG	5(7.94)	10 (7.81)	1.105(0.354–3.455)	1.000	0.913(0.262–3.177)	0.866
AG + GG	25 (39.68)	44 (34.38)	1.256(0.674–2.341)	0.473	1.389(0.709–2.723)	0.338

The data was presented as n (%).MCI, mild cognitive impairment. The genotype and allele frequencies were compared between the groups with Pearson’s χ² tests except for the cases labeled with a, in which the frequency was determined by continuity-corrected Pearson’s χ² tests. ^bAdjusted for age, educational attainment, gender. ^*Significance, p < 0.05.

Comparison of plasma FFA, lipid metabolism, and cognitive performances between different rs11868035 genotypes

In the MCI group, the increased tendency of LMT-DR scores, which was used to comprehensively assess the memory by measuring verbal learning as delayed free recall, were displayed in order of AA, AG, and GG genotypes (p = 0.038). In addition, DST scores, which assess both attention and short-term memory, were displayed in ascending order according to AA, AG, and GG genotypes (p = 0.016) in all T2DM subjects. The G allele was a protective factor for memory and attention. Although FFA had an ascending trend according to AA, AG, and GG genotypes in all T2DM and MCI subjects, no significant differences were found between the rs11868035 gene polymorphisms and plasma FFA level or other indices, including BMI, HbA1c, TG, TC, HDL-c, LDL-c, HDL-c/TC, HDL-c/LDL-c, LDL-c/TC, ApoA1, and ApoB in the MCI group or in all subjects (p > 0.05, Table 6).

Table 6

Comparison of cognitive test scores in all subjects and the MCI group according to the rs11868035 genotype

	All groups				MCI group
	AA (n = 122)	AG (n = 54)	GG (n = 15)	p	AA (n = 39)	AG (n = 20)	GG (n = 5)	p
MoCA	26.15 (25–28)	25.96 (24–28)	26.47 (25–28)	0.787	22.62 (21–25)	22.90 (22–24)	24.20 (23–25)	0.339
MMSE	28.46 (28–30)	27.80 (26.75–29)	28.60 (28–30)	0.081	26.85 (26–28)	25.95 (24.25–27.75)	27.40 (25.50–29.00)	0.143
DST	11.60 (10.75–13)	12.28 (11–14)	12.53 (11–14)	0.019^*	10.62 (9–11)	11.10 (10–12.75)	10.60 (8.50–12.00)	0.503
VFT	16.85 (14–19)	17.13 (14–20.25)	18.67 (15–22)	0.452	14.41 (13–17)	15.10 (13–17)	15.20 (13.50–17.00)	0.911
CDT	3.36 (3–4)	3.61 (3–4)	3.40 (3–4)	0.162	3.08 (2–4)	3.20 (2.25–4.00)	3.00 (2.50–3.50)	0.780
TMTA	64.07 (48–72)	66.17 (48–77)	62.07 (54–71)	0.911	77.41 (56–86)	68.85 (55–79.25)	72.60 (56.00–90.50)	0.792
TMTB	159.24 (107–2202	174.06 (104.25–196.25)	163.07 (115–215)	0.806	205.62 (158–255)	190.70 (162–229.50)	188.20 (129.50–246.50)	0.912
AVLT IR	18.48 (15–22)	17.24 (14–21.25)	18.73 (16–230	0.411	15.13 (13–17)	15.50 (13.25–19.50)	19.40 (16.50–22.00)	0.073
AVLT DR	5.97 (4–7)	5.39 (3.75–7.00)	6 (4–7)	0.431	4.32 (3–6)	4.65 (2.25–6.00)	6.60 (6.00–7.00)	0.038^*
SCWT A time	34.01 (25–39)	32.19 (23.50–36.25)	31.20 (23–32)	0.312	37.69 (25–50)	37.30 (27–47.75)	41.20 (27.50–47.50)	0.855
SCWT A number	49.67 (50–50)	49.59 (50–50)	50 (50–50)	0.519	49.74 (50–50)	49.80 (50–50)	50.00 (50–50)	0.637
SCWT B time	54.59 (40–67)	48.69 (36.75–54.50)	49.40 (35–60)	0.080	61.59 (44–74)	59.95 (44.25–72.50)	65 (39–89)	0.897
SCWT B number	48.60 (48–50)	48.65 (48–50)	49.00 (48–50)	0.558	47.49 (46–50)	48.15 (47.25–50.00)	48.60 (47.50–49.50)	0.371
SCWT C time	98.95 (75–120)	88.70 (69.75–103.00)	106.73 (64–129)	0.135	114.13 (85–136)	105.25 (74.25–126.50)	140.20 (101–189.50)	0.267
SCWT C number	46.11 (45–49)	46.94 (45.75–49.00)	46.60 (46–49)	0.480	43.85 (43–48)	45.65 (44–48)	44 (38.50–48.50)	0.166
FFA	401.936±76.101	424.068±80.903	436.999±134.079	0.245	414.873±82.396	414.873±82.396	548.779±189.008	0.143
HbA1c	9.08 (7.58–10.53)	8.82 (7.70–9.83)	8.37 (7.00–8.80)	0.604	9.68 (7.98–10.80)	9.38 (8.05–10.43)	8.84 (7.00–10.80)	0.879
TG	1.73 (0.91–2.14)	1,79 (0.94–2.33)	1.69 (1.07–1.97)	0.743	1.82 (0.87–2.22)	1.94 (0.88–2.34)	1.43 (1.08–1.88)	0.799
TC	4.56±1.12	4.59±0.91	4.62±1.49	0.933	4.76±1.20	4.87±0.83	4.81±1.33	0.821
HDL-c	1.20 (0.95–1.40)	1.13 (1.00–1.27)	1.15 (0.87–1.56)	0.861	1.28 (1.01–1.50)	1.18 (1.07–1.30)	1.11 (0.84–1.44)	0.473
LDL-c	2.79±0.84	2.89±0.87	2.67±1.24	0.727	2.87±0.97	3.11±0.97	2.59±1.34	0.690
ApoA1	1.16±0.31	1.15±0.26	1.21±0.27	0.759	1.21±0.29	1.26±0.28	1.23±0.28	0.918
ApoB	0.87 (0.67–0.96)	0.83 (0.68–0.94)	0.80 (0.53–0.98)	0.824	0.90 (0.67–1.04)	0.90 (0.68–1.12)	0.80 (0.56–1.06)	0.810

The data was presented as the median (interquartile range) or mean±SD. Kruskal-Wallis test for comparison of asymmetrically distributed quantitative variables between genotypic subgroups in MCI group and all groups. MoCA, Montreal Cognitive Assessment; MMSE, Mini-Mental State Examination; TMT-A and TMT-B, Trail Making Test-A and Trail Making Test-B, respectively; AVLT-IR and AVLT-DR, Auditory Verbal Learning Test-Immediately Recall and Auditory Verbal Learning Test-Delayed Recall, respectively; CDT, Clock Drawing Test; DST, Digit Span Test; VFT, Verbal Fluency Test; SCWT, Stroop Color Word Test; MCI, participants with mild cognitive impairment; BMI, body mass index; HbA1c, glycosylated hemoglobin; TG, triglyceride; TC, total cholesterol; LDL-c, low density lipoprotein cholesterol; HDL-c, high density lipoprotein cholesterol; ApoA1, Apolipoprotein A1; HOMA-IR, homeostatic model assessment of insulin resistance; FFA, free fatty acid. ^*Significance, p < 0.05.

Logistic regression models

We used simple logistic regression to select possible risk factors of diabetic cognitive impairment into the multivariable regression model. First, in the simple logistic regression model, we included age, gender, education level, BMI, diabetes duration, hypertension duration, SBP, DBP, use of insulin and metformin, HbA1c, FFA, smoking, drinking, HDL-c, LDL-c, TC, TG, ApoA1, and ApoB into the covariate list. The results revealed that female, low education levels, long diabetes and hypertension duration, use of insulin, and high HbA1c and FFA may be associated with MCI in T2DM (Table 7). These significant variables were selected for multivariable regression analysis. Results showed that female, low education level, long diabetes and hypertension duration, use of insulin, high HbA1c, and FFA level were associated with MCI in T2DM.

Table 7

Assessment results of risks of variables of having MCI in logistic regression models in T2DM patients

Variables	Simple logistic regression				Multivariable logistic regression
	β	p	OR	95%CI	β	p	OR	95%CI
Gender (female/male)	–1.053	0.034^*	0.350	0.133–0.925	–1.089	0.003^*	0.337	0.163–0.697
Age (y)	–0.003	0.905	0.997	0.948–1.049
Education Level (y)	–0.252	<0.001^*	0.777	0.680–0.887	–0.219	<0.001^*	0.803	0.713–0.905
Diabetes duration (y)	0.102	0.005^*	1.106	1.031–1.190	0.078	0.014^*	1.081	1.016–1.151
Hypertension duration (y)	0.044	0.047^*	1.045	1.001–1.091	0.050	0.007^*	1.051	1.013–1.090
SBP	–0.023	0.095	0.977	0.951–1.004
DBP	0.031	0.160	1.032	0.988–1.077
BMI	0.109	0.104	1.116	0.976–1.275
HbA1c (%)	0.203	0.022^*	1.225	1.029–1.459	0.239	0.003^*	1.270	1.085–1.485
FFA (μmol/L)	0.007	0.015^*	1.007	1.001–1.012	0.005	0.036^*	1.005	1.000–1.010
Smoking, n (%)	–0.645	0.273	0.525	0.165–1.664
Drinking, n (%)	0.850	0.235	2.340	0.575–9.520
Use of insulin, n (%)	1.348	0.002^*	3.851	1.621–9.147	1.295	0.001^*	3.652	1.683–7.927
Use of metformin, n (%)	0.277	0.544	1.319	0.539–3.227
TG (mmol/L)	–0.023	0.916	0.978	0.644–1.484
TC (mmol/L)	0.213	0.740	1.238	0.351–4.360
LDL-c (mmol/L)	–0.076	0.914	0.927	0.233–3.693
HDL-c (mmol/L)	0.416	0.738	1.516	0.133–17.298
ApoA1 (g/L)	0.552	0.614	1.737	0.203–14.891
ApoB (g/L)	0.180	0.809	1.197	0.279–5.137

The simple logistic regression and multivariable logistic regression were used. BMI, body mass index; HbA1c, glycosylated hemoglobin; FBG, fasting blood-glucose; PBG, postprandial blood glucose; TG, triglyceride; TC, total cholesterol; LDL-c, low density lipoprotein cholesterol; HDL-c, high density lipoprotein cholesterol; FFA, free fatty acid; HOMA-IR, homeostatic model assessment of insulin resistance; MCI, participants with mild cognitive impairment. ^*Significance, p < 0.05.

DISCUSSION

The present study suggested that the plasma FFA concentration in patients with T2DM having MCI was higher than that in the healthy-cognition group. Further logistic regression analysis showed that elevated plasma FFA levels were independent risk factors of diabetic-MCI. Additionally, correlation analysis showed that higher plasma FFA levels were correlated with attention and execution deficit, particularly with low ApoA1 levels, and were also correlated with attention and executive function deficits. Moreover, elevated BMI and non-use of metformin might be correlated with attention deficits. Multivariable logistic regression analysis also showed that high HbA1c levels, long duration of diabetes and hypertension, female, use of insulin, and low education level were risk factors for MCI in patients with T2DM. For SREBP-1c gene rs11868035, although the G allele of rs11868035 appeared to be a protective factor for memory in patients with T2DM having MCI, no remarkable difference was observed in the distributions of genotypes and alleles of rs11868025 between the two groups, and no association was found between rs11868035 polymorphism and plasma FFA levels.

The correlation between FFA levels and cognitive function has been investigated, but the results are inconsistent [25 , 39]. In our study, plasma FFA levels were higher in MCI than in healthy-cognition diabetic patients, and further multivariable regression analysis showed that elevated FFA concentration was an independent risk factor for MCI in patients with T2DM. These findings were consistent with previous studies, in which abnormal FFA levels could influence the development of AD [25, 38]. Conversely, some types of FFA are reduced in patients with AD, and these types include free polyunsaturated and monounsaturated fatty acids, which are the main components of the membrane and are important for brain structure and cognitive function [40]. The total FFA levels, which are mainly characterized by free saturated fatty acid (SAFA), remained higher in AD than in the control group [25]. However, a 23-year follow-up study among 10,059 Israeli men found that long-term high-SAFA diet is negatively correlated with AD, suggesting that SAFA may be beneficial for cognitive function [39]. A possible explanation for these contradictions is that different fatty acid types have different effects on cognition, and the negative effects of one fatty acid on cognition may be offset by the positive effects of another fatty acid. However, the interaction of different FFA remains unknown.

In all T2DM or in the MCI group, the plasma FFA levels were positively correlated with SCWT c (time) scores, thus reflecting patients’ selective attention deficit, neuronal plasticity, and dorsolateral prefrontal cortical activity. Attention deficit is an early symptom of AD, an important feature of cognitive deterioration in AD, and a possible primary cognitive disorder that may significantly affect other cognition performance, such as language, memory, and executive function [41 –43]. Therefore, the possible mechanisms in which FFA causes attention deficit and neuronal plasticity should be explored for the early prevention and treatment of MCI. The etiology of attention deficits in AD are the damage of the frontal and parietal-related areas, the disconnection between anterior and posterior attentional networks, and the reduction of cholinergic function [41 , 43–47].

The above results showed that elevated plasma levels of FFA in T2DM were associated with high risk for MCI, especially for attention deficit. Based on the correlation analysis between FFA and lipid metabolism profiles, FFA was negatively correlated with HDL-c and ApoA1, while no correlation with TG, TC, LDL-c, and ApoB was observed. Further correlation analysis between ApoA1 and cognitive performances showed that ApoA1 was positively correlated with VFT scores, thus reflecting attention and execution function. However, in the present study, ApoA1 was not the independent positive factor for MCI in patients with T2DM according to multi-variable regression analysis. The results were not completely consistent with previous studies, suggesting that ApoA1 was associated with Aβ plaques in AD brains, lower serum ApoA1 levels were associated with cognitive decline, and elevated plasma ApoA 1 levels may effectively prevent learning and memory deficits in patients with AD [14 , 48–50]. The possible causes may be the small sample size, effects of other confounding factors, and the measurement error of cognitive domains. Additionally, although HDL-c was the risk factor of MCI in patients with T2DM, no correction was found between HDL-c and multiple cognitive function, and HDL-c was not the risk factor of MCI in patients with T2DM [15, 16]. This finding was associated with the differences in patient selection and the fact that our study was a cross-sectional study. Moreover, many inconsistent interference factors affected the results between the two groups. ApoA1 was a main protein moiety of HDL-c and account for approximately 70%[51]. Moreover, ApoA1 played a key role in the reverse cholesterol transport-transporting cholesterol from peripheral tissues to the liver metabolism, and it protects the blood–brain barrier in atherosclerosis [52]. Hence, elevated plasma FFA levels, when combined with decreased apolipoprotein A1, may portend abnormal cholesterol transport susceptible to early cognitive impairment, especially for attention and execution deficit.

Based on the comparison of multiple domains of cognitive performances of different rs11868035 genotypes, we found that the carriers of the G allele had high DST scores, which mainly reflect the patients’ attention and short-term memory function, implying that the G allele is a protective factor for attention and short-term memory in patients with T2DM. Moreover, the AVLT DR score of MCI group, which was positively correlated with delayed memory, increased in the order of AA, AG, and GG genotype. These results support that G allele carriers had lower MMSE scores and was a protective factor for cognition in Parkinson’s disease patients [53]. In another study, A allele carriers have higher IR, which is an independent risk factor for AD, the G allele in rs11868035 appears as a protected factor for AD in T2DM [32]. However, the underlying mechanisms have not been clarified, and it was only found that the possible protection of the G allele in memory, and no association was found in the overall cognitive function.

In schizophrenia, rs11868056 GG homozygous subjects showed a higher prevalence of metabolic syndrome and higher TG level and worse processing speed performance, compared with A allele carriers in schizophrenia [33]. However, in non-alcoholic fatty liver disease and control subjects, no association between the rs11868035 genotype and BMI, TC, HDL-c, and LDL-c both in non-alcoholic fatty liver disease and control subjects was found, except that rs11868035 G allele showed significant associations with a lower TG level in control subjects [54]. These are inconsistent. Regrettably, in this present study, no significant relationship was found between rs11868035 genotypes and plasma FFA levels or other lipid metabolism profiles. These inconsistent findings were obtained possibly because different races and the low gene frequency in the Chinese and the small sample size recruited both decreased the testing capabilities. Moreover, T2DM associated cognitive protection was polygenic hereditary effect, with multiple genes and gene-environment interactions, and rs11868035 may have low effect on the overall cognitive protection. In addition, this condition may be caused by a survivorship bias effect. Hence, a few people can defend against risk factors. Therefore, a larger sample size should be set, and more targeted experiments should be conducted to clarify the exact mechanism.

Our study investigated the relationships among plasma FFA levels, SREBP-1c SNP rs11868035, cognitive dysfunction in T2DM patients, comprehensively evaluated the cognitive functions of the patients, such as, memory, attention, executive ability, and visual space, and explored the correction between FFA and lipid metabolism profiles. However, our study had some limitations. First, only total plasma FFA levels were measured. Hence, the exhaustive effects of each type of FFA should be determined. Moreover, although the important role of the APOE genotypes was considered in the pathogenesis of AD, and its possible interaction with the SREBP-1c rs11868035 genotypes may affect the development of AD, the associations of the two genes was hardly determined based on the small sample size in this study [55]. We intended to expand the sample size to determine the possible pathogenic role of rs11868035 genotypes with different APOE genotypes. Next, the MCI and control groups did not match well in the cross-sectional study, which were affected by confounding factors. In addition, drugs such as statins, which were used for patients with diabetes, may influence the metabolism of FFA and lipid. Moreover, we did not evaluate the depression status of the patients, because this condition has an effect on MCI assessment. Finally, the power of test would decrease because of the small sample size, the interpretation degree of some results was limited, and some other possible differences may be ignored.

CONCLUSION

In summary, the plasma FFA concentration in patients with T2DM having MCI was higher than that in the non-MCI group. Moreover, elevated plasma FFA levels were independent risk factors for MCI in patients with T2DM. Additionally, plasma FFA levels were negatively related to ApoA1 levels, and both factors were positively correlated with attention and executive function. Together, these results suggest that the elevated plasma FFA levels combined with a decrease in apolipoprotein A1 may predict abnormal cholesterol transport that is predisposing to early cognitive impairment, especially for attention and execution deficit. Further studies need to be performed to verify the detailed metabolism. Although the G allele appeared to be a protective factor for memory in patients with T2DM having MCI, no remarkable difference was observed in the genotype and allele distribution of rs11868025 between the two groups, and no correlation was observed between rs11868035 polymorphism and plasma FFA levels. The role of SREBP-1c rs11868035 in the early cognition dysfunction of T2DM patients needs to be further investigated using larger sample sizes.

Footnotes

ACKNOWLEDGMENTS

The authors would like to thank all participants of this study.

This work was supported by the National Natural Science Foundation of China (No. 81570732, Shaohua Wang and No. 81870568, Shaohua Wang).

Authors’ disclosures available online ().

References

NCD Risk Factor Collaboration (NCD-RisC) (2016) Worldwide trends in diabetes since 1980: A pooled analysis of 751 population-based studies with 4.4 million participants. Lancet 387, 1513–1530.

Cho

, Shaw

, Karuranga

, Huang

, da Rocha Fernandes

, Ohlrogge

, Malanda

(2018) IDF Diabetes Atlas: Global estimates of diabetes prevalence for 2017 and projections for 2045. Diabetes Res Clin Pract 138, 271–281.

, Wang

, Xiao

(2016) Type 2 diabetes mellitus might be a risk factor for mild cognitive impairment progressing to Alzheimer’s disease. Neuropsychiatr Dis Treat 12, 2489–2495.

Gudala

, Bansal

, Schifano

, Bhansali

(2013) Diabetes mellitus and risk of dementia: A meta-analysis of prospective observational studies. J Diabetes Investig 4, 640–650.

Petersen

(2000) Mild cognitive impairment: Transition between aging and Alzheimer’s disease. Neurologia 15, 93–101.

Petersen

, Doody

, Kurz

, Mohs

, Morris

, Rabins

, Ritchie

, Rossor

, Thal

, Winblad

(2001) Current concepts in mild cognitive impairment. Arch Neurol 58, 1985–1992.

Sallam

, Tumurbaatar

, Zhang

, Tuvdendorj

, Chandalia

, Tempia

, Laezza

, Taglialatela

, Abate

(2015) Peripheral adipose tissue insulin resistance alters lipid composition and function of hippocampal synapses. J Neurochem 133, 125–133.

Yuan

, Wang

(2017) Mild cognitive impairment in type 2 diabetes mellitus and related risk factors: A review. Rev Neurosci 28, 715–723.

Kaiser

, Selzner

, Weber

, Schallner

(2020) Carbon monoxide controls microglial erythrophagocytosis by regulating CD36 surface expression to reduce the severity of hemorrhagic injury. Glia 68, 2427–2445.

10.

Abd Elaaty

, Ismail

, Sheshtawy

, Sultan

, Ebrahim

(2019) Assessment of comorbid mild cognitive impairment and depression in patients with type 2 diabetes mellitus. Diabetes Metab Syndr 13, 1759–1764.

11.

, Li

, Zhao

, Wu

, Ji

, Huang

, Ma

(2016) Relationship between plasma lipids and mild cognitive impairment in the elderly Chinese: A case-control study. Lipids Health Dis 15, 146.

12.

Gonzalez-Escamilla

, Atienza

, Garcia-Solis

, Cantero

(2016) Cerebral and blood correlates of reduced functional connectivity in mild cognitive impairment. Brain Struct Funct 221, 631–645.

13.

Zou

, Zhu

, Deng

, Duan

, Pan

, Tu

, Dai

, Zhang

, Chu

, Lü

(2014) Vascular risk factors and mild cognitive impairment in the elderly population in Southwest China. Am J Alzheimers Dis Other Demen 29, 242–247.

14.

Lewis

, Cao

, Lu

, Mans

, Su

, Jungbauer

, Linton

, Fazio

, LaDu

, Li

(2010) Overexpression of human apolipoprotein A-I preserves cognitive function and attenuates neuroinflammation and cerebral amyloid angiopathy in a mouse model of Alzheimer disease. J Biol Chem 285, 36958–36968.

15.

Cai

, Han

, Sun

, Huang

, Tian

, Shen

, Dong

, Xia

, Wang

(2016) Plasma clusterin and the CLU gene rs11136000 variant are associated with mild cognitive impairment in type 2 diabetic patients. Front Aging Neurosci 8, 179.

16.

Cai

, Han

, Sun

, Huang

, Tian

, Shen

, Wang

(2017) Effects of ABCA1 R219K polymorphism and serum lipid profiles on mild cognitive impairment in type 2 diabetes mellitus. Front Aging Neurosci 9, 257.

17.

Merched

, Xia

, Visvikis

, Serot

, Siest

(2000) Decreased high-density lipoprotein cholesterol and serum apolipoprotein AI concentrations are highly correlated with the severity of Alzheimer’s disease. Neurobiol Aging 21, 27–30.

18.

, Zhao

, Zhuang

, Sun

, Zhang

, Ni

(2018) Low glucagon-like peptide-1 (GLP-1) concentration in serum is indicative of mild cognitive impairment in type 2 diabetes patients. Clin Neurol Neurosurg 174, 203–206.

19.

Sekizkardes

, Chung

, Chacko

, Haymond

, Startzell

, Walter

, Lightbourne

, Brown

(2020) Free fatty acid processing diverges in human pathologic insulin resistance conditions. J Clin Invest 130, 3592–3602.

20.

Wilson

, Binder

(1997) Free fatty acids stimulate the polymerization of tau and amyloid beta peptides. In vitro evidence for a common effector of pathogenesis in Alzheimer’s disease. Am J Pathol 150, 2181–2195.

21.

Patil

, Chan

(2005) Palmitic and stearic fatty acids induce Alzheimer-like hyperphosphorylation of tau in primary rat cortical neurons. Neurosci Lett 384, 288–293.

22.

Marwarha

, Claycombe-Larson

, Lund

, Schommer

, Ghribi

(2019) A diet enriched in palmitate and deficient in linoleate exacerbates oxidative stress and amyloid-beta burden in the hippocampus of 3xTg-AD mouse model of Alzheimer’s disease. J Alzheimers Dis 68, 219–237.

23.

Talbot

, Young

, Jolly-Tornetta

, Lee

, Trojanowski

, Wolf

(2000) A frontal variant of Alzheimer’s disease exhibits decreased calcium-independent phospholipase A2 activity in the prefrontal cortex. Neurochem Int 37, 17–31.

24.

Bloom

(2014) Amyloid-β and tau: The trigger and bullet in Alzheimer disease pathogenesis. JAMA Neurol 71, 505–508.

25.

Fonteh

, Cipolla

, Chiang

, Arakaki

, Harrington

(2014) Human cerebrospinal fluid fatty acid levels differ between supernatant fluid and brain-derived nanoparticle fractions, and are altered in Alzheimer’s disease. PLoS One 9, e100519.

26.

Dhopeshwarkar

, Mead

(1973) Uptake and transport of fatty acids into the brain and the role of the blood-brain barrier system. Adv Lipid Res 11, 109–142.

27.

Chen

, Green

, Orr

, Bazinet

(2008) Regulation of brain polyunsaturated fatty acid uptake and turnover. Prostaglandins Leukot Essent Fatty Acids 79, 85–91.

28.

Yang

, Cao

, Wang

, Li

, Hu

, Wan

, Li

(2021) Inhibition of nitric oxide synthase aggravates brain injury in diabetic rats with traumatic brain injury. Neural Regen Res 16, 1574–1581.

29.

Shimano

(2002) Sterol regulatory element-binding protein family as global regulators of lipid synthetic genes in energy metabolism. Vitam Horm 65, 167–194.

30.

Yahagi

, Shimano

, Hasty

, Amemiya-Kudo

, Okazaki

, Tamura

, Iizuka

, Shionoiri

, Ohashi

, Osuga

, Harada

, Gotoda

, Nagai

, Ishibashi

, Yamada

(1999) A crucial role of sterol regulatory element-binding protein-1 in the regulation of lipogenic gene expression by polyunsaturated fatty acids. J Biol Chem 274, 35840–35844.

31.

Hua

, Wu

, Goldstein

, Brown

, Hobbs

(1995) Structure of the human gene encoding sterol regulatory element binding protein-1 (SREBF1) and localization of SREBF1 and SREBF2 to chromosomes 17p11.2 and 22q13. Genomics 25, 667–673.

32.

Liu

, Liu

, Li

, Xie

, Guo

, Tian

, Ma

, Zhang

, Gao

(2008) Association of sterol regulatory element-binding protein-1c gene polymorphism with type 2 diabetes mellitus, insulin resistance and blood lipid levels in Chinese population. Diabetes Res Clin Pract 82, 42–47.

33.

Vargas-Alarcon

, Gonzalez-Pacheco

, Perez-Mendez

, Posadas-Sanchez

, Cardoso-Saldana

, Ramirez-Bello

, Escobedo

, Nieto-Lima

, Fragoso

(2019) SREBF1c and SREBF2 gene polymorphisms are associated with acute coronary syndrome and blood lipid levels in Mexican population. PLoS One 14, e0222017.

34.

Grarup

, Stender-Petersen

, Andersson

, Jorgensen

, Borch-Johnsen

, Sandbaek

, Lauritzen

, Schmitz

, Hansen

, Pedersen

(2008) Association of variants in the sterol regulatory element-binding factor 1 (SREBF1) gene with type 2 diabetes, glycemia, and insulin resistance: A study of 15,734 Danish subjects. Diabetes 57, 1136–1142.

35.

Huang

, Wang

, Tian

, Huang

, Guo

, Zhang

, Shi

, Wang

(2019) Higher plasma level of Nampt presaging memory dysfunction in Chinese type 2 diabetes patients with mild cognitive impairment. J Alzheimers Dis 70, 303–314.

36.

Alberti

, Zimmet

(1998) Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: Diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabet Med 15, 539–553.

37.

Albert

, DeKosky

, Dickson

, Dubois

, Feldman

, Fox

, Gamst

, Holtzman

, Jagust

, Petersen

, Snyder

, Carrillo

, Thies

, Phelps

(2011) The diagnosis of mild cognitive impairment due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 270–279.

38.

Cui

, Chen

, Liu

, Xie

, Wu

, Wang

(2015) Gas chromatography-mass spectrometry analysis of the free fatty acids in serum obtained from patients with Alzheimer’s disease. Biomed Mater Eng 26, S2165–S2177.

39.

Goldbourt

, Yaari

, Medalie

(1993) Factors predictive of long-term coronary heart disease mortality among 10,059 male Israeli civil servants and municipal employees. A 23-year mortality follow-up in the Israeli Ischemic Heart Disease Study. Cardiology 82, 100–121.

40.

Innis

(2007) Fatty acids and early human development. Early Hum Dev 83, 761–766.

41.

Perry

, Hodges

(1999) Attention and executive deficits in Alzheimer’s disease. A critical review. Brain 122 (Pt 3), 383–404.

42.

Bracco

, Bessi

, Padiglioni

, Marini

, Pepeu

(2014) Do cholinesterase inhibitors act primarily on attention deficit? A naturalistic study in Alzheimer’s disease patients. J Alzheimers Dis 40, 737–742.

43.

Perry

, Watson

, Hodges

(2000) The nature and staging of attention dysfunction in early (minimal and mild) Alzheimer’s disease: Relationship to episodic and semantic memory impairment. Neuropsychologia 38, 252–271.

44.

Klinkenberg

, Sambeth

, Blokland

(2011) Acetylcholine and attention. Behav Brain Res 221, 430–442.

45.

Sarter

, Gehring

, Kozak

(2006) More attention must be paid: The neurobiology of attentional effort. Brain Res Rev 51, 145–160.

46.

Parasuraman

, Martin

(1994) Cognition in Alzheimer’s disease: Disorders of attention and semantic knowledge. Curr Opin Neurobiol 4, 237–244.

47.

Foldi

, Lobosco

, Schaefer

(2002) The effect of attentional dysfunction in Alzheimer’s disease: Theoretical and practical implications. Semin Speech Lang 23, 139–150.

48.

Choi

, Seo

, Yi

, Sohn

, Choe

, Byun

, Lee

, Woo

, Lee

(2016) Amyloid-independent amnestic mild cognitive impairment and serum apolipoprotein A1 levels. Am J Geriatr Psychiatry 24, 144–153.

49.

Harr

, Uint

, Hollister

, Hyman

, Mendez

(1996) Brain expression of apolipoproteins E, J, and A-I in Alzheimer’s disease. J Neurochem 66, 2429–2435.

50.

Song

, Poljak

, Crawford

, Kochan

, Wen

, Cameron

, Lux

, Brodaty

, Mather

, Smythe

, Sachdev

(2012) Plasma apolipoprotein levels are associated with cognitive status and decline in a community cohort of older individuals. PLoS One 7, e34078.

51.

Chistiakov

, Orekhov

, Bobryshev

(2016) ApoA1 and ApoA1-specific self-antibodies in cardiovascular disease. Lab Invest 96, 708–718.

52.

Tall

(1998) An overview of reverse cholesterol transport.A. Eur Heart J 19(Suppl A), 31–35.

53.

Lou

, Li

, Liu

, Li

, Ren

, Luo

(2019) The polymorphism of SREBF1 gene rs11868035 G/A is associated with susceptibility to Parkinson’s disease in a Chinese population. Int J Neurosci 129, 660–665.

54.

Peng

, Chen

, Liu

, Hu

, Lin

(2016) Lack of association between SREBF-1c gene polymorphisms and risk of non-alcoholic fatty liver disease in a Chinese Han population. Sci Rep 6, 32110.

55.

Tang

, Li

, Zhang

, Chen

, Sun

(2019) ɛ3/4 genotype of the apolipoprotein E is associated with higher risk of Alzheimer’s disease in patients with type 2 diabetes mellitus. Gene 703, 65–70.