Relationship between inflammatory markers and mild cognitive impairment in Chinese patients with type 2 diabetes: a case-control study

A case-control study was conducted at the Metabolic Disease Hospital of Tianjin Medical University from June 2016 to January 2017. The enrollment process was described in Fig. 1. Three hundred thirty-six elderly diabetic patients who were hospitalized for diabetes or its complications were enrolled. All the diabetic patients were met current international diagnostic standards for diabetes [14] and were treated with anti-diabetic medications, including oral hypoglycemic agents (OHAs) and/or insulin injection on the basis of lifestyle interventions. Subjects with hypertension were treated with antihypertensive therapy. Subjects with dyslipidemia were treated with statins, except for those with fasting serum triglyceride were treated with fibrates to prevent acute pancreatitis. Subjects were excluded from the study if they exhibited: (1) type 1 diabetes, (2) acute or severe chronic diabetic complications, such as diabetes ketosis, (3) proliferative diabetic retinopathy, diabetic foot infection or ulcer, (4) severe diabetic nephropathy: estimated glomerular filtration rate (eGFR) < 45 ml/min • 1.73m²; or urinary albumin creatinine ratio > 300 mg/g; or eGFR 45–60 ml/min • 1.73m² and urinary albumin creatinine ratio 30-300 mg/g [15]. The eGFR was estimated by the Chronic Kidney Disease Epidemiology Collaboration creatinine equation [16], (5) history of severe hypoglycemia or hypoglycemia unawareness which the individuals required the assistance of another person to administer treatment [17], (6) acute coronary syndrome or other severe cardiovascular diseases, (7) neurological disorders, such as established dementia, Parkinson’s disease and seizures, (8) any apparent past history of a cerebrovascular events, or presence of any clinically significant abnormality in brain computed tomography or magnetic resonance imaging scan, (9) visual or hearing disabilities; psychiatric disorder, (10) acute or chronic infection, or taking NSAIDS, antibiotics or steroids within the 3 months. Altogether, 39 subjects were excluded according the above exclusion criteria. Totally 298 T2D patients completed Mini-Mental State Examination (MMSE) [18] and a short memory questionnaire [19]. Sixty-three T2D patients with MCI were detected as cases. Accordingly, among the 235 diabetic subjects with normal cognitive function (NCF), 63 T2D ones were selected randomly as controls matched by same gender, similar age (±3 years) and education using an on-line Random Number Generator/Picker according to the visiting orders of subjects. Consequently, the final sample was composed of 63 cases and 63 controls (Fig. 1).

All subjects underwent MMSE [18] and a short memory questionnaire [19] by trained interviewers. The questionnaire aimed to distinguish if the patients complained about decreased subjective memory. The MMSE is a widely used brief cognitive screening test, which assesses space/time orientation, verbal memory, language, attention/calculation, and visuoconstructive abilities. Another questionnaire was used to determine whether the subjects had any impairment in activities of daily living (ADL) [20]. The patients who met the Petersen’s diagnostic criteria [21] were defined as having MCI: subjective memory complaint and the evidence of objective memory impairment; the MMSE scores were above 13, but ≤18 or ≤ 20 or ≤ 24, with zero, or 1 to 6 years, or more than 6 years education, respectively; intact or minimal damage of ADL; but failure to meet the DSM-IIIR criteria of dementia [22]. The participants with NCF were defined as the subjects who had no subjective memory complaint or objective memory damage, and MMSE scores over the above values based on the corresponding education level.

The study protocol was approved by the Ethics Committee of Tianjin Medical University. Written informed consent was received from each patient.

Height was measured without shoes to the nearest 1 cm and weight measured in light clothing to the nearest 0.1 kg on a beam balance scale. Body mass index (BMI) was calculated as weight in kilograms divided by the square of the height in meters (kg/m²). Blood pressure was measured twice on the left arm with subjects in a seated position after 15 min rest. These measures were averaged to get a mean blood pressure. Information about demographic characteristics, medication history, duration of diabetes, anti-diabetic medications (including OHAs and/or insulin) and diabetic complications, as well as lifestyle habits (smoking, drinking and physical activity) were collected by trained interviewers. Venous blood samples were taken from all participants after an overnight fast (12 h at least) and the samples were stored at − 80 °C until assessment assayed performed. The glycated hemoglobin A1c (GHbA1c) was measured using high-performance liquid chromatography method. The levels of lipid profile including total cholesterol (TC), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C) and fasting plasma glucose (FPG) were measured enzymatically on a Modular Analytics P800 analyzer (Roche Diagnostics, Indianapolis, IN).

Mean and standard deviations (SD) or median (inter-quartile range) were used as descriptive statistics for continuous variables and percentages for categorical variables. For continuous variables, the paired-sample t test or Wilcoxon rank-sum test was used for between-group comparisons, and chi-square test for categorical variables. Spearman correlation analysis was conducted to analyze the correlation between serum folate and inflammatory markers. Logistic regression analysis was used to assess the association between the inflammatory markers and risk of MCI. Odds ratio (OR) and corresponding 95% confidence intervals (CI) were calculated. Model 1 was used to calculate the crude OR, and model 2 was adjusted for factors related diabetes including duration of diabetes (per year), HbA1c (< 7.0%/≥7.0%), anti-diabetic medications (only OHAs /only insulin/insulin plus OHAs), diabetic retinopathy (yes/no) and diabetic nephropathy (yes/no), and common cardiovascular diseases risk factors including smoking status (yes/no), body mass index (< 24.0 kg/m²/24.0–28.0 kg/m²/≥28.0 kg/m²), hypertension (yes/no), LDL-C (< 2.6 mmol/L/≥2.6 mmol/L). Model 3 was additional adjusted for serum folate. All statistical operations were performed using SPSS (version 13.0, Chicago, IL, USA). All reported P values were two-sided and P value less than 0.05 was considered to indicate a statistically significant difference.

The socioeconomic data and MMSE scores of the subjects were shown in Table 1. Compared with the subjects in T2D-NCF group, those in T2D-MCI group had lower MMSE scores (20.38 ± 3.68 vs. 26.75 ± 2.84; P < 0.001). The mean level of TC (5.45 ± 1.10 mmol/L) in T2D-MCI group was significantly higher than that in T2D-NCF group (5.00 ± 1.00 mmol/L, P < 0.05). The percentage of diabetic retinopathy (42.9%), diabetic nephropathy (41.3%), LDL-C ≥ 2.60 mmol/L (84.1%) in T2D-MCI group were significantly higher than that in T2D-NCF group (19.0, 19.0 and 66.7%, respectively; P < 0.05). There were no significant differences in BMI, systolic blood pressure (SBP), diastolic blood pressure (DBP), the duration of diabetes, triglycerides, HDL-C, FPG, GHbA1c, the percent of hypertension, ADL scores, smoking status and drinking status between T2D-MCI group and T2D-NCF group (P > 0.05). In addition, there were no significant differences in all the above indices between the low folate subgroup (< 7.0 μg/L) and the high folate subgroup (≥7.0 μg/L) in T2D-MCI group or T2D-NCF group (P > 0.05) (Table 1).

The median value of serum folate in T2D-MCI group was significantly lower than that in T2D-NCF group (T2D-MCI: 6.25(5.30–7.94) μg/L vs. T2D-NCF: 8.07(6.65–10.40) μg/L) (P < 0.01) (Table 2). Compared with the T2D-NCF group, the median values of serum TNF-α, IL-6 and hs-CRP were significantly higher than that in T2D-MCI group (P < 0.05) (Table 2).

Spearman correlation analysis showed that the serum folate was negatively correlated with hs-CRP, TNF-α, and IL-6 (r value: − 0.33, − 0.19, − 0.21, respectively; P < 0.05) (Table 3).

Table 4 showed the odds ratios of MCI across tertiles of inflammatory markers. In univariate analysis, compared with the subjects in the lowest tertile, those in the upper tertile of IL-6 (OR = 3.09, 95%CI: 1.57–6.10; P < 0.05) and hs-CRP (OR = 2.25, 95%CI: 1.14–4.44; P < 0.05) were associated with higher odds of MCI. The TNF-α was not associated with risk of MCI (P > 0.05). In multivariate analysis, the significant associations persisted after adjusted for the confounding variables including smoking, overweigh and obesity, duration of diabetes, anti-diabetic medications, HbA1c, hypertension, LDL-C, diabetic retinopathy and diabetic nephropathy. The significant association between IL-6 and MCI persisted after further adjusting for serum folate, whereas the association between hs-CRP and MCI disappeared. Since the greatest reduction in neural tube defects has been observed at serum concentrations≥7.0 μg/L [23], we further divided all the subjects into two subgroups according to the level of serum folate (< 7.0 μg/L/≥7.0 μg/L) and evaluated the associations between hs-CRP and MCI among the two subgroups (Table 5). The significant association between hs-CRP and MCI only existed in the low folate subgroup (< 7.0 μg/L) after adjusted for the above confounding factors (OR = 3.34, 95%CI: 1.05–10.64), not in the high folate subgroup (≥7.0 μg/L; OR = 2.16, 95% CI: 0.68–6.88).