Relationship between glycemic control, microalbuminuria and cognitive functions in elderly type 2 diabetic patients

Abstract

Aim: The prevalence of diabetes is increasing in elderly populations, and is thought to be an important risk factor for cognitive dysfunction in this age group. Methods: The study included 104 patients aged over 60 years who were followed-up for type 2 diabetes for at least 6 months, in addition to 44 controls. Glycemic parameters, microangiopathic complications, microalbumin elimination, and the Standardized Mini Mental State Examination (SMMSE) scores were used as indicators of cognitive function. Results: The SMMSE scores of diabetic patients were significantly lower than the control group (p < 0.05). The average SMMSE score for normoalbuminuric diabetic patients was 22.36 ± 4.66, compared with 22.61 ± 4.90 for the microalbuminuria patients (p = 0.84). A positive correlation was found between SMMSE scores and patients’ hemoglobin values and education levels, whereas a negative correlation was noted between SMMSE scores and systolic and diastolic blood pressures and hemoglobin A1c levels (p < 0.05). Patients with diabetic neuropathy, a microvascular complication of diabetes, were found to have significantly lower SMMSE scores (p = 0.011). Conclusion: Elderly diabetic patients showed decreased cognitive function compared to volunteers. No relationship was established between microalbuminuria and cognitive functions, although diabetic neuropathy was found to be related to decreased cognitive function.

• Albuminuria
• cognitive functions
• dementia
• diabetic microvascular complications
• Standardized Mini Mental State Examination
• Type 2 diabetes mellitus

Introduction

Type 2 diabetes mellitus (DM) is becoming a serious concern due to its increasing incidence in societies, as well as potential to cause several organ dysfunctions resulting from systemic complications. The microangiopathic complications of type 2 DM include nephropathy, neuropathy, and retinopathy, and macroangiopathic complications include stroke, ischemic heart disease, and peripheral vascular ischemia. Dementia is a cognitive disorder that is primarily accompanied by memory dysfunction and various and changing psychiatric and behavioral symptoms, including personality changes.1 Numerous studies have reported reduced cognitive functions in diabetic patients, especially among the elderly; however, the exact mechanism for this condition is not clear.2,3 The rise in the number of diabetic elderly patients mandates close evaluation of dementia and cognitive functional disorders in this particular population. SMMSE evaluates orientation, memory, attention, and language use, and is a simple and valid means to evaluate cognitive functions.4

The occlusive type cerebrovascular diseases, especially those affecting the small vessels, seen in the elderly, cause reduction in cognitive functions.5,6 There is no established relationship between microvascular diseases in elderly individuals with type 2 diabetes and cognitive dysfunction. Endothelial dysfunction is observed in type 2 DM. Microalbuminuria, as a sign of diabetic nephropathy, is also known to be related to endothelial dysfunction. Therefore, development of atherosclerosis, which is seen in diabetic patients, and diabetic glomerulosclerosis, are thought to follow a similar path and both are considered as significant risk factors for stroke and coronary heart diseases in this patient group.7 When considering the relationship of microvascular diseases to cognitive dysfunction in diabetic patients, microalbuminuria could be an early sign of cognitive dysfunction. We investigated the relationship between albumin clearance and cognitive functions in type 2 diabetic patients older than 60 years of age.

Materials and methods

Patients

All of the participants were informed about the study, consent was obtained prior to enrollment in the study, and the study was approved by the Local Ethics Board. The study included 104 type 2 diabetic patients older than 60 years, who were followed up at the Endocrinology and Metabolism Department, Uludag University Medical School for at least 6 months; and 44 volunteers of the same age group. Comparison of the diabetic patients and the control group, whose diabetes are ruled out by measuring fasting plasma glucose twice, is made.8 Type 1 diabetic patients, type 2 diabetic patients who had experienced macrovascular complications within the previous 6 months, and those with autoimmune diseases, malign diseases, and late-stage liver or renal failure were excluded from the study. The patients with diabetes used oral antidiabetic drugs or insulin were included in the study. Users of the drugs which influence the glucose metabolism such as steroids, chemotherapeutics, antiepileptics, oral contraceptives were excluded. Furthermore, patients with active neurological disease or depression were also excluded. Macrovascular complications of the patients are anamnestic questioned and verified by examining the patients’ hospital records.

Study design

The medical histories of the patients were obtained, including duration of diabetes, tobacco-alcohol consumption, and presence of diabetic complications. Blood pressure and pulse readings following a 15-minute rest were recorded after a minimum of two measurements. Other anthropometric measurements included height, weight, body mass index (BMI), waist circumference, hip circumference, and waist circumference/hip circumference ratios were measured.

Patients’ glycemic control (glycosylated hemoglobin A1c (HbA1c), fasting plasma glucose (FPG), post-prandial glucose (PPG), and fasting insulin levels (FPI)) were evaluated between 8 to 9 am, following a minimum of 10 h fasting. BMI was calculated by the investigators as weight in kilograms divided by the square of height in meters. Spot urine microalbumin and creatinine measurements were collected and urinary microalbumin/creatinine ratios (MCr) were calculated. Microalbuminuria was defined as MCr >30 µg/mg in the spot urine. Patients were then grouped as Type 2 diabetics without microalbuminuria (Group 1), Type 2 diabetics with microalbuminuria (Group 2), and the control group (Group 3).

Evaluation of cognitive functions

SMMSE was completed on all participants enrolled in the study. SMMSE includes questions on 11 subjects, grouped under five categories: orientation, immediate recall, attention, delayed verbal recall, and language. The SMMSE score ranges between 0 and 30, with the lower score indicating worsening level of cognitive impairment, and the test takes approximately 10 min to administer.9 It provides a baseline score of cognitive function, and assists in early identification of deficits in those functions. The Turkish version of the SMMSE has been shown to be reliable and valid in Turkish population.1

Statistical analysis

Statistical analysis of data was performed using the Statistical Package for Social Sciences software version 17.0 (IBM Acquires SPSS Inc., Somers, NY), which is used in the Uludag University network. Data are presented as mean ± standard deviation (SD). Chi-square test was used for clinical data variables. Mann–Whitney U-test analysis and Student t-test was used to assess the differences among the study groups. Pearson’s and Spearman’s correlation coefficients were used where appropriate. In all statistical analyses, results were considered statistically significant if two-tailed p was less than 0.05.

Results

A total of 148 patients were enrolled in the study (99 females, 49 males) with average age of 65.72 ± 5.12 years. The anthropometric and laboratory findings of the 104 diabetic patients and 44 volunteers are shown in Table 1. The average age for diabetic patients was 65.86 ± 5.19 years compared to 65.40 ± 5.00 years in the control group (p = 0.55). When diabetic patients were grouped based on albumin creatinine clearance, 60 patients were found to be normoalbuminuric (Group 1, MCr < 30 µg/mg) and 44 were microalbuminuric (Group 2, 30 µg/mg ≤ MCr < 300 µg/mg).

Table 1. The anthropometric characteristics and laboratory parameters of type 2 diabetic patients and control group.

	Diabeticpatients (n = 104)	Control group (n = 44)	p
Age (year)	65.86 ± 5.19	65.40 ± 5.00	NS
Gender (female/male)	62/42	27/17	NS
Smokers (n, %)	18 (17.3%)	5 (17.2%)	NS
Alcohol consumption (n, %)	3 (2.9%)	1 (2.27%)	NS
Body weight (kg)	79.75 ± 15.09	73.75 ± 14.57	0.019
BMI (kg/m^2)	30.28 ± 5.66	28.64 ± 5.19	NS
Waist circumference (cm)	102.17 ± 12.70	96.27 ± 13.18	0.013
Hip circumference (cm)	113.02 ± 11.84	111.04 ± 11.73	NS
WHR	0.90 ± 0.07	0.86 ± 0.06	0.002
FPG (mg/dL)	155.93 ± 57.12	96.95 ± 19.02	<0.01
A1c (%)	7.55 ± 1.57	−	−
TC (mg/dL)	194.40 ± 36.07	193.84 ± 35.83	NS
HDL-C (mg/dL)	46.22 ± 13.37	56.81 ± 12.70	<0.01
LDL-C (mg/dL)	117.44 ± 33.57	116.61 ± 34.82	NS
TG (mg/dL)	158.50 ± 77.40	131.30 ± 59.27	0.034
MCr (µg/mg)	45.15 ± 59.55	−	−
SMMSE	22.47 ± 4.74	24.22 ± 3.56	0.025

Note: NS: not significant, BMI: body mass index, WHR: waist–hip ratio, FPG: fasting plasma glucose, A1c: hemoglobin A1c, TC: total cholesterol, HDL-C: high-density lipoprotein cholesterol, LDL-C: low-density lipoprotein cholesterol, TG: triglyceride, MCr: urinary microalbumin/creatinine ratios, SMMSE: standardized mini mental state examination.

Weight, waist circumference, and waist circumference/hip circumference ratios of diabetic patients were found to be significantly higher than the control group (p > 0.05). Glycemic parameters were also found to be higher in diabetic patients (Table 1). There was no statistically significant difference in anthropometric and glycemic parameters between group 1 and group 2 patients (Table 2). The mean duration of diabetes was 133.01 ± 81.90 months in group 1 and 145.68 ± 107.43 months in group 2 (p = 0.89).

Table 2. The anthropometric characteristics and laboratory parameters of type 2 diabetic patients with normoalbuminuria and microalbuminuria.

	Patients with normoalbuminuria (group 1, n = 60)	Patients with microalbuminuria (group 2, n = 44)	p
Age (year)	65.30 ± 5.14	66.63 ± 5.22	NS
Duration of diabetes (month)	133.01 ± 81.90	66.63 ± 5.22	NS
Body weight (kg)	79.11 ± 11.73	80.63 ± 18.85	NS
BMI (kg/m^2)	30.47 ± 4.83	30.03 ± 6.67	NS
Waist circumference (cm)	102.51 ± 10.67	101.70 ± 15.16	NS
Hip circumference (cm)	114.10 ± 9.99	111.56 ± 13.96	NS
WHR	0.89 ± 0.06	0.91 ± 0.08	NS
SBP (mmHg)	132.25 ± 12.83	132.95 ± 11.17	NS
DBP (mmHg)	81.96 ± 10.31	80.45 ± 10.66	NS
Hemoglobin (g/L)	12.87 ± 1.68	12.77 ± 1.88	NS
FPG (mg/dL)	161.63 ± 62.50	148.15 ± 48.44	NS
PPG (mg/dL)	216.06 ± 89.06	214.25 ± 69.76	NS
A1c (%)	7.54 ± 1.49	7.57 ± 1.68	NS
TC (mg/dL)	192.51 ± 36.37	196.97 ± 35.92	NS
HDL-C (mg/dL)	46.00 ± 10.88	46.52 ± 16.30	NS
LDL-C (mg/dL)	115.48 ± 35.03	120.21 ± 31.60	NS
TG (mg/dL)	157.55 ± 82.04	159.83 ± 71.31	NS
MCr (µg/mg)	12.11 ± 6.41	90.20 ± 69.55	< 0.01
SMMSE	22.36 ± 4.66	22.61 ± 4.90	NS

Note: NS: not significant, BMI: body mass index, WHR: waist–hip ratio, SBP: systolic blood pressure, DBP: diastolic blood pressure, FPG: fasting plasma glucose, PPG: postprandial plasma glucose, A1c: hemoglobin A1c, TC: total cholesterol, HDL-C: high-density lipoprotein cholesterol, LDL-C: low-density lipoprotein cholesterol, TG: triglyceride, MCr: urinary microalbumin/creatinine ratios, SMMSE: standardized mini mental state examination.

The average SMMSE scores were 22.36 ± 4.66 in group 1, 22.61 ± 4.90 in group 2, and 24.22 ± 3.56 in group 3. There was no statistically significant difference in average SMMSE scores between groups 1 and 2 (p = 0.84). Compared with the control group, diabetic patients showed significantly lower SMMSE scores (p < 0.05). Correlation analysis was performed to all patients. Factors associated with diabetes mellitus like blood pressure, HbA1c, etc., and also social factors evaluated with correlations analysis. Then correlating factors were analyzed with linear regression model. Pearson’s correlation analyses showed SMMSE was positively correlated with hemoglobin and education levels, and was negatively correlated with systolic and diastolic blood pressures and HbA1c levels (p < 0.05) (Table 3). Education level was found most related factors with MMS scores. There was no relationship noted between SMMSE and MCr.

Table 3. The factors related with SMMSE scores.

	Correlations		Linear regression
	r	p*	β	CI (5–95%)		p**
Grup (diabetes/control)	0.185	0.024	0.01	−1.54	1.80	0.87
Gender (female/male)	−0.306	<0.01	−0.18	−3.35	0.44	0.056
Economic level (bad versus good)	0.271	0.01	0.03	−1.11	1.56	0.74
Education level (primary versus higher)	0.538	<0.01	0.41	1.01	2.56	<0.001
Hemoglobin (g/L)	0.166	0.044	−0.01	−0.44	0.38	0.88
Systolic blood pressure (mmHg)	−0.176	0.032	0.05	−0.06	0.09	0.60
Diastolic blood pressure (mmHg)	−0.286	<0.01	−0.25	−0.20	0.02	0.017
Hemoglobin A1c (%)	−0.221	0.01	−0.09	−0.66	0.19	0.28

Note: r: correlation coefficient, β: regression coefficient; *p < 0.05 significant, **p < 0.01 significant.

CI: 95.0% confidence interval for β.

Fifteen out of 104 diabetic patients presented with diabetic retinopathy. The SMMSE scores in patients with retinopathy averaged 22.26 ± 4.71, compared with 22.59 ± 4.84 in patients without diabetic retinopathy (p = 0.76). Twenty-eight of the diabetic patients were noted as having neuropathy and their average SMMSE score was 20.57 ± 4.22, whereas it was 23.28 ± 4.69 in patients without neuropathy. The average SMMSE score was significantly lower in patients with neuropathy compared to those without (p = 0.011). When coronary and peripheral vascular disease data were examined, 30 patients were noted to have macrovascular disease. Those patients had an average SMMSE score of 22.30 ± 4.90, whereas those without any vascular disease averaged 22.64 ± 4.65. No statistically significant difference was noted between the two groups (p = 0.76).

Discussion

Our study results showed that SMMSE scores, as an indicator of cognitive function, were significantly lower in diabetic patient groups compared to the controls. Cognitive functions were positively correlated with hemoglobin and education levels, and negatively correlated with systolic and diastolic blood pressures as well as HbA1c levels. We did not detect a significant relationship between cognitive functions and microalbuminuria.

Various studies have reported that diabetes in elderly individuals is associated with reduced cognitive functions.10,11 Cukierman et al.10 showed that patients with diabetes were 1.2 to 1.7 times more likely to show reduced cognitive functions compared to those without diabetes. In another study, Mogi et al.12 found that SMMSE scores in elderly diabetic patients were significantly lower compared to controls, and that cognitive functions were negatively correlated with HbA1c levels and insulin use. We found that the SMMSE scores of elderly diabetic patients were significantly lower compared to the control group. When our diabetic patients were grouped as insulin users (n = 42) and non-users (n = 62), we were not able to show any significant relationship between the average SMMSE scores of those groups (p = 0.256).

Elderly patients are more likely to present with reduced cognitive functions and have increased risk of dementia.13 Hemoglobin levels among elderly patients are also known to be related to reduced cognitive functions.13 Reduction in the oxygen carrying capacity of the blood due to anemia may result in reduced cognitive functions and increased risk of silent cerebro-vascular strokes. Another study reported a positive correlation between hemoglobin levels and SMMSE scores of the participants enrolled.14 Similarly, we found that hemoglobin levels and SMMSE scores were positively correlated in our study. Education level is also known to affect cognitive performance during later stages of adult life. While some studies found no such relationship,4 we identified a positive correlation between the level of education and SMMSE scores in our study group.

Hypertension is a significant risk factor for vascular dementia in elderly patients. However, there are different arguments presented for the causes of Alzheimer type dementia.15 In several cross-sectional studies, a negative correlation was found between hypertension and Alzheimer disease in later stages of life.16 It is thought that the reduction in blood pressure may cause cerebral hypoperfusion and neurodegeneration, both conditions potentially contributing to the development of Alzheimer’s.15,16 Triantafyllidi et al.17 showed that impaired cognitive function is associated with an increased large artery stiffness and microalbumin excretion in newly diagnosed, untreated hypertensive patients. Their findings supported the hypothesis that cognitive impairment induced by impaired microcirculation is linked to large artery stiffness and microvascular damage. Similar to earlier studies, we found a negative correlation between the systolic and diastolic blood pressures and cognitive functions. HbA1c, as the average blood glucose level over the previous 3 months, is used as a control parameter in controlling diabetes. The findings of the present study are consistent with previous studies, which showed a negative correlation between HbA1c levels and cognitive function.18,19

The causative relationship between diabetes and cognitive function has not been shown exclusively. It is thought that microvascular diseases seen in diabetic patients may be related to cognitive dysfunction.14 Some studies have shown negative correlation between microalbuminuria and cognitive functions.14,20,21 Afşar et al.14 reported a significant relationship between SMMSE scores of macroalbuminuric and normoalbuminuric patients, but did not find a significant difference between microalbuminuric and normoalbuminuric sub-groups. In another study, Joosten et al.22 found a significant relationship between albuminuria and cognitive dysfunction in young patients, but failed to show a similar relationship in elderly patients. Barzilay et al.23 estimated that 2.6 to 5.9% of cases of cognitive decline in the general population, and 6.8 to 8.7% in those with DM, are associated with microalbuminuria or macroalbuminuria. They suggest that, among people with macrovascular disease, microvascular disease in one organ system may reflect microvascular disease in another. Bruce et al.24 showed that urinary albumin excretion is a potentially modifiable risk factor for cognitive function in older diabetic individuals. Abbatecola et al.25 suggested that microalbuminuria in older patients with impaired glucose tolerants, there was a significant decline in cognitive functioning compared to that in those with normoalbuminuria. Their findings showed that urinary albumin excretion is a predictor of vascular disease in older patients with plasma glucose fluctuations. In our study, we were not able to show a significant relationship in SMMSE scores between microalbuminuric and normoalbuminuric patients, and there was no statistically significant correlation between microalbuminuria and SMMSE scores.

Other microvascular complications of diabetes, diabetic retinopathy, and diabetic neuropathy are thought to be related to reduced cognitive functions. In a study of individuals with no known cerebrovascular diseases, retinopathy was found to be related to reduced cognitive functions, and it was concluded that an underlying disease caused retinopathy and reduced cognitive functions.26 Vascular dysfunctions in elderly patients result in reduced blood flow to the nerves and reduced endoneurial oxygen pressure. On the other hand, high blood glucose levels have direct toxic effects on nerves, resulting in development of diabetic neuropathy and reduced cognitive function.27 Another study found that, apart from vascular dysfunctions, polyol pathway and advanced glycation end-product accumulation may also result in development of neuropathy and reduced cognitive function.28 We were not able to detect any significant relationship between retinopathy and cognitive functions; however, we found that patients with diabetic neuropathies have lower SMMSE scores.

In conclusion, elderly diabetic patients present with reduced cognitive functions compared to controls. No relationship was found between microalbuminuria and cognitive functions. Diabetic neuropathy, a form of diabetic complications, was related to reduced cognitive dysfunction. Prospective studies with larger sample groups are needed in order to determine the relationship between diabetic microvascular complications and cognitive functions in elderly diabetic patients.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.