Total serum homocysteine levels do not identify cognitive dysfunction in multimorbid elderly patients

Abstract

Objectives

Total blood homocysteine (Hcys) and folate levels have been investigated in association with cognitive dysfunction in healthy but not in multimorbid elderly patients. We hypothesized that total serum Hcys is an adequate marker to identify multimorbid elderly patients with cognitive dysfunction assessed by the Short Cognitive Performance Test (SKT) and Mini-Mental State Examination (MMSE).

Design

Cross-sectional study.

Setting

The study center was an acute geriatric hospital.

Participants

A total of 189 multimorbid elderly patients were recruited.

Methods

Cognitive dysfunction was determined according to the SKT and MMSE. Biochemical parameters (Hcys, folate, vitamin B12, hemoglobin), nutritional status (BMI, Mini Nutritional Assessment, nutritional intake), and activities of daily living were assessed.

Results

According to the SKT, 25.4% of patients showed no cerebral cognitive dysfunction, 21.2% had suspected incipient cognitive dysfunction, 12.7% showed mild cognitive dysfunction, 9.0% had moderate cognitive dysfunction, and 31.7% of patients were demented. The median plasma Hcys value was elevated by ∼20% in multimorbid elderly patients, independent of cognitive dysfunction. Serum folate and vitamin B₁₂ concentrations were within normal ranges. We did not find significant differences in nutritional status, activities of daily living, numbers of diseases or medications, or selected biochemical parameters between the SKT groups.

Conclusion

Elevated serum Hcys levels with normal plasma folate and vitamin B12 concentrations were observed in multimorbid elderly patients. The plasma Hcys level did not appear to be an important biological risk factor for cognitive dysfunction in multimorbid geriatric patients.

Introduction

Cognitive impairment is a common age-related disorder that results from a variety of metabolic and other diseases (1). To ensure normal cognitive function the water-soluble vitamins B6, B12, and folate, as well as limiting factors in the metabolism of homocysteine (Hcys), are essential (1, 2). Numerous clinical trials have confirmed that Hcys not only is a risk factor for vascular disease (3, 4) but also is related to cognitive dysfunction or dementia in the elderly (2, 5, 6). Hyperhomocysteinemia is considered to be a direct marker of early cognitive decline, and elevated Hcys levels result from neurotoxic and vasotoxic effects in dementia and Alzheimer’s disease. Because there are limited therapeutic options for Alzheimer’s disease, treatment with high-dose folic acid and vitamin B12 to decrease Hcys levels has attracted considerable interest (7).

However, investigations of the relationship between cognitive dysfunction and vitamin status have been inconclusive, with results ranging from no effect of low vitamin B₁₂ (8, 9) or folate levels (11) to negative effects of low vitamin B₁₂ (10) or folate levels (10, 11, 12) on cognitive dysfunction. Furthermore, increased total Hcys levels have been associated with decreased plasma levels of vitamin B₆, vitamin B₁₂, and folate (13) but also with normal plasma levels of these vitamins (14, 15).

Garcia and Zanibbi reviewed in 2004 (16) the relationship between Hcys and cognitive function scores in healthy elderly individuals. Results varied from significant (10, 17, 18, 19) to nonsignificant associations (20, 21). In unhealthy elderly people, the interpretation of hyperhomocysteinemia is complicated by different causes of multimorbidity. Multimorbidity, a characteristic feature of geriatric patients, is defined by the concurrent appearance of several, mostly chronic diseases that exert mutual influences and lead to functional losses. Unfortunately, until now no clinical studies on hyperhomocysteinemia in this patient group exist. Hence, we examined the usefulness of Hcys determination with regard to special geriatric assessments in multimorbid elderly patients. We hypothesized that total serum Hcys is a possible marker to identify multimorbid elderly patients with cognitive dysfunction assessed by the Short Cognitive Performance Test (SKT).

Methods

Study population

Between October 2005 and December 2006, 189 (131 female, 58 male) multimorbid geriatric patients with acute medical conditions were recruited. The study center was the “Evangelisches Geriatriezentrum Berlin,” Germany, an acute geriatric institution with 132 stationary beds. An acute medical condition was defined as a condition of rapid onset, severe symptoms, and brief duration and included conditions resulting from chronic illnesses that can be completely or substantially cured. Multimorbidity was defined as the co-occurrence of multiple diseases within one person. Patients younger than 65 years or patients for whom blood samples for Hcys measurements were unavailable were excluded. According to the International Classification of Diseases and Related Health Problems (ICD-10), patients with Alzheimer’s disease (G30.-) were excluded, as well.

The study was approved by the local ethics committee of Charité–Universitaetsmedizin Berlin, and every patient or person in charge provided written informed consent.

Short Cognitive Performance Test

The Short Cognitive Performance Test (Syndrom Kurz-Test, SKT) introduced by Erzigkeit et al. is a diagnostic tool that can be administered in ~15 min and is practical for serial and routine use (22). Especially in German-speaking countries, the SKT has been used extensively in clinical trials to evaluate the severity of impairment in attention and memory, with an emphasis on speed and execution time. The SKT is a brief battery that defines two factors of memory and attention deficit. The test consists of nine performance subtests, each limited to 1 min, including naming objects and numbers, immediate and delayed recall, recognition memory, arranging and replacing numbers, counting symbols, and a Stroop task. Scores range from 0 (without cognitive impairment) to 27 points. Patients are divided into the following score groups: 0-4, no cerebral cognitive dysfunction; 5-8, suspected incipient cognitive dysfunction; 9-13, mild cognitive dysfunction; 14-18, moderate cognitive dysfunction; 19-23, severe cognitive dysfunction; and 24-27, very severe cognitive dysfunction.

Mini-Mental State Examination

The “Mini-Mental State Examination” (MMSE) is a brief bedside screening test for cognitive function in the elderly (23). The items cover orientation, immediate and delayed recall, attention and calculation, naming, reading, writing, and drawing. The maximum score of the test is 30. A score of 24-30 indicates no or only mild cognitive impairment, 17-23 indicates moderate cognitive impairment, and 0-16 indicates severe cognitive impairment. Administration of the test requires 10 min. The MMSE is practical to use routinely for the elderly and serially for people with dementia.

Nutritional status

Nutritional status was assessed by two trained investigators according to a Mini Nutritional Assessment (MNA) and anthropometric measurement 48 h after hospital admission. Body weight was measured to the nearest 0.1 kg with a seat scale (Seca, Hamburg, Germany) in light indoor clothing without shoes, and height was measured to the nearest 0.1 cm with a stadiometer. The height of bedridden patients was estimated by knee-height measurement with the calculation by Chumlea and Guo (24). The knee height was measured in the supine position as the distance between the knee and the sole of the foot when the leg forms a 90° angle with the thigh. BMI was calculated from weight and height (weight/[height]²).

The MNA was developed by Guigoz et al. for the assessment of nutritional status in elderly people (25). The MNA covers 18 items dealing with anthropometric assessment (BMI, calf circumference, mid-upper arm circumference), general assessment (medication, acute disease, neuropsychological problems, pressure ulcer, independent living), dietary assessment (number of meals; everyday consumption of protein-containing food, vegetables, fruits, beverages), and self-assessment (consideration of health status, self-perceived view of nutritional status). The nutritional status of the patient was classified as “well-nourished” (24-30 points, MNA-A), “moderately malnourished or at risk of malnutrition” (17-23.5 points, MNA-B), or “malnourished” (<17 points, MNA-C).

Oral nutritional intake was measured with a 3-day nutrition diary that was initiated after the identification of at-risk patients. For analysis, we used the nutritional software program DGE-PC, professional version 3.1.1.042. The basal metabolic rate (BMR) was calculated with the Harris-Benedict formula (26). According to this formula, energy expenditure was calculated with an activity score of 1.2 for bedridden and 1.4 for active patients.

Activities of daily living

The activities of daily living (ADL) were assessed by the Barthel Index (BI). The first version of BI was developed by Mahoney and Barthel (27). The BI is a proxy questionnaire that measures three categories of function: self-care, continence of bowel and bladder, and mobility. The following single items were assessed: drinking from a cup; eating; dressing the upper and lower body; grooming; washing; bowel and bladder status; transfers in and out of a chair, toilet, tub, or shower; and walking and climbing stairs. Scores range from 0 (totally dependent on another’s help) to 100 (totally independent) points.

Biochemical parameters

After an overnight fast, venous blood samples were obtained during the first 24 h of admission. The samples for homocysteine (Hcys) measurement were collected in EDTA tubes, and blood was centrifuged at 3000 x g for 10 min at 6 °C. The plasma was portioned into 1-mL plastic tubes and stored at –80 °C until analysis. The Hcys level was determined by a competitive immunoassay with direct chemiluminescent detection by the Siemens/Bayer Centaur Immunoanalyzer.

Hemoglobin and mean corpuscular volume (MCV) were measured with an XE2100 analyzer (Sysmex). Albumin, creatinine, and ferritin were analyzed with a Modular P800 system. Folate and vitamin B₁₂ were measured by electrochemiluminescence with an Elecsys 2010 (Roche) system. Biochemical parameters were analyzed at the Charité-Universitaetsmedizin Berlin, Germany.

Individuals were classified as being at high risk of vitamin deficiency if they had low folate (<2.8 μg/L) or vitamin B₁₂ (<199 ng/L) concentrations or elevated Hcys concentrations (>15 μmol/L).

Statistics

Statistical tests and analyses were performed with SPSS software, version 12.0 (SPSS, Inc., Chicago, IL). Results were considered statistically different if p < 0.05, and data were represented as mean ± SD or median [25^th percentile;75th percentile]. To identify differences in biochemical parameters, nutritional status, age, length of stay, activities of daily living, multimedication, and multimorbidity according to SKT as well as MMSE, a one-way analysis of variance and Bonferroni multiple comparison test were used. Differences in nutritional intake from recommendations were assessed by using the Student’s t-test of one sample. The Spearman correlation coefficient was used for bivariate analysis. Significant differences are indicated in Tables 1 and 2.

Table 1.

Patient Characteristics According to SKT

Number of patients	All 189	0-41 48	5-82 40	9-133 24	14-184 17	Dementia 60
% female	69.3	72.9	70.0	62.5	82.4	65.0
Number of medications	9 [7;12]	8 [6;11]	10 [8;12]	9 [6;12]	9 [7;13]	9 [7;11]
Number of diseases	9 [7;12]	8 [6;10]	9 [7;12]	10 [8;13]	10 [8;13]	10 [8;13]
Length of hospital stay (d)	21 [15;26]	17 [14;26]	21 [19;24]	21 [16;24]	23 [20;28]	21 [15;27]
Age (years)	78.6 ± 7.3	75.9 ± 6.9	77.5 ± 5.8	78.0 ± 7.6	81.4 ± 6.7	81.8 ± 7.5
Activities of daily living (BI)	55 [35;65]	60 [55;79]	63 [46;70]	55 [40;65]	55 [33;63]	30 [5;45]
Nutritional status
BMI (kg/m2)	25.7 ± 4.6	26.9 ± 4.8	25.7 ± 3.8	26.2 ± 5.6	25.1 ± 5.0	24.6 ± 4.0
Albumin (g/L)	37 [34;40]	38 [35;41]	38 [34;40]	37 [33;39]	36 [34;41]	36 [33;39]
Well-nourished (%)	15.9	27.7	15.0	4.2	5.9	-
At risk of malnutrition (%)	69.6	66.0	72.5	62.5	76.5	-
Malnourished (%)	14.5	6.4	12.5	33.3	17.6	-
Cognition
MMSE	25 [19;27]	27 [25;28]	26 [25;28]	24 [21;26]	22 [18;25]	15 [13;19]
None/mild cognitive impairment (%)	52.5	89.1	86.8	68.2	37.5	0
Moderate cognitive impairment (%)	14.9	10.9	13.2	31.8	62.5	0
Severe cognitive impairment (%)	32.6	0	0	0	0	100.0
Depression (%)	5.4	4.8	2.9	10.5	0	6.8
Biochemistry
Folate (μg/L)	6.1 [4.7;9.2]	7.2 [5.0;9.9]	6.0 [3.8;9.6]	6.5 [5.3;9.3]	5.9 [4.6;7.7]	5.9 [4.5;9.1]
Homocysteine (μmol/L)	18.8 [15.8;24.9]	19.2 [14.3;23.5]	18.8 [15.2;30.2]	20.7 [16.4;27.9]	21.9 [17.6;29.5]	18.1 [15.4;24.1]
Vitamin B12 (ng/L)	418 [298;562]	461 [378;610]	380 [275;513]	366 [283;544]	466 [386;804]	400 [257;595]
Creatinine (mg/dL)	0.95 [0.77;1.28]	0.92 [0.75;1.19]	0.91 [0.71;1.16]	1.14 [0.85;1.30]	1.03 [0.83;1.67]	0.95 [0.77;1.30]
Hemoglobin (g/dL)	11.8 [10.8;13.3]	11.9 [10.2;13.4]	11.8 [10.2;13.3]	11.3 [10.9;12.3]	11.9 [10.6;13.7]	12.2 [10.9;13.5]
Ferritin (mg/L)	236 [112;428]	238 [91;355]	191 [97;377]	248 [151;438]	228 [96;825]	244 [146;462]
MCV (fL)	91 [88;94]	91 [89;93]	92 [88;93]	92 [88;96]	92 [88;95]	91 [86;94]

BMI, Body Mass Index; ADL, activities of daily living; BI, Barthel Index; MMSE, Mini-Mental State Examination. 1. No cerebral cognitive dysfunction; 2. suspected incipient cognitive dysfunction; 3. mild cognitive dysfunction, 4moderate cognitive dysfunction. Results are given as mean ± standard deviation or median [25th;75th percentile].

Table 2.

Nutritional Intake

	Recommendation	Daily intake	P
Energy (kcal)	1611	1090 [951;1290]	<0.001
Folate (μg/d)	400	97 [80;128]	<0.001
Vitamin B12 (μg/d)	3	3.29 [2.63;4.27]	0.006
Vitamin B6 (mg/d)	1300	950 [806;1298]	0.020

Results

Data from 38 patients were discarded before analysis because of unavailable serum Hcys levels. Therefore, 189 (131 female and 58 male) geriatric patients with a mean age of 78.6 ± 7.3 years were included between October 2005 and December 2006. Moderate depression was found in only 9 patients and did not influence the data; thus, these patients were included in the study. The clinical and nutritional characteristics of the study population are presented in Table 1. All included patients were characterized by acute medical conditions and multimorbidity. Admission diagnoses included fractures (27.0%), neurological problems (20.1%), cardiovascular disease (15.9%), and connective tissue disease (11.1%). Approximately 70% of the population had a secondary school certificate and professional training.

Cognitive dysfunction in relation to Hcys and vitamin levels

None of the patients consumed multivitamin, folate, or vitamin B12 supplements. Mild-to-moderate elevation of plasma Hcys levels (>15 μmol/L) was observed in 78.8% of patients. None of the patients had severe hyperhomocysteinemia (>100 mmol/L). Median serum levels of folate (range 2.8-13.5 μg/L) and vitamin B12 (range 199-730 ng/L) were within the normal ranges. Only 2.1% and 8.5% of the patients showed reduced levels of serum folate and vitamin B12, respectively.

According to the SKT, 25.4% of patients had no cerebral cognitive dysfunction, 21.2% had suspected incipient cognitive dysfunction, 12.7% had mild cognitive dysfunction, and 9.0% had moderate cognitive dysfunction. Approximately 31.7% of patients exhibited a type of dementia according to ICD-10 and were classified in a special “dementia” group. None of the patients was categorized in the SKT group of 19-23 or 24-27 points. Plasma Hcys, serum folate, and serum vitamin B12 levels did not differ between SKT and demented groups. Significant differences in Hcys, folate, or vitamin B12 levels could not be found between demented patients and those without cognitive dysfunctions. Furthermore, we did not find significant differences in nutritional status, activities of daily living, numbers of diseases or medications, or selected biochemical parameters between SKT groups.

To verify the results of the SKT, patients were also divided into MMSE groups. Again, there were no significant differences between groups. Hcys, folate, and vitamin B12 levels were inadequate parameters to assess cognitive dysfunction. Weak significant correlations between serum Hcys and plasma folate (r = -0.298) and between plasma folate and plasma vitamin B₁₂ (r = 0.302) were found, but no correlation between serum Hcys and plasma vitamin B₁₂ levels was detected. Furthermore, we did not find any correlation between plasma folate and SKT or MMSE score or between serum Hcys and SKT or MMSE score.

In this study group, cognitive dysfunction increased significantly with age, but vitamin and Hcys levels did not change significantly with age. Gender, multimorbidity, and multimedication did not influence serum levels of Hcys.

Nutritional intake

Nutritional intake was assessed in 41 patients (Table 2). The median total energy amount (1090 kcal) was 32.3% below the recommendation (1611 kcal). Most of the elderly patients in the sample had inadequate intakes of folate (97.4% of patients) and of vitamin B6 (75%), but vitamin B₁₂ was within the recommended range. These results are comparable with unpublished data of 340 multimorbid geriatric patients at the same hospital.

We did not find significant correlations between vitamin intake and serum or plasma levels of Hcys, folate, or vitamin B₁₂.

Discussion

In Alzheimer’s disease, neither recommendations for prevention nor an effective cure exists. Therefore, screening for possible risk factors, e.g., low folic acid concentration in serum or pre-existing hyperhomocysteinemia, is crucial. Several clinical studies have investigated the relationship between Hcys levels and cognitive function in elderly individuals, but data on multimorbid patients have been lacking.

Multimorbid elderly patients differ from other patients by the presence of pre-existing multiorgan insufficiency that requires multidrug therapy. High Hcys levels have been reported with the use of anticonvulsants such as phenytoin, carbamazepine, and valproic acid (46, 47, 48). Several other drugs, such as theophylline, isoniazid, and hydralazine, may increase plasma Hcys levels by inhibiting the synthesis of vitamin B₆ (18). Another important factor associated with elevated Hcys levels is a lack of physical exercise because of chronic pain or other reasons. In addition to gender, age, and folate, vitamin B₁₂, and serum albumin levels, renal function and the use of diuretics are possible factors that influence Hcys levels (5, 6, 29, 39)

We did not investigate these factors, with the exception of blood creatinine, which is an indicator of renal function. However, the median creatinine levels of our investigated patients did not exceed 1.14 μmol/ml.

The vitamin status of healthy elderly people is controversial, and little is known about the multimorbid elderly (14, 31, 36, 37). The Hcys level might be a more significant parameter to identify folate deficiencies than serum folate concentration (38). Several investigators have reported that polymorphism of the enzyme 5-10-methylenetetrahydrofolate reductase is related to higher Hcys levels, especially in combination with an impaired folate status (40, 41).

Vitamin deficiencies in the transmethylation pathway might cause elevated Hcys levels due to impaired transformation of Hcys to methionine. Plasma/serum folate and vitamin B₁₂ can be limiting factors in this pathway. Although almost all patients (97.4%) showed impaired dietary folate intake, the majority showed increased Hcys with normal serum folate and vitamin B₁₂ concentrations, independent of cognitive dysfunction. Selhub et al. (14) and Saw et al. (15) also reported increased Hcys and normal vitamin levels. These results conflict with the low intake of folate because dietary folate strongly influences serum folate levels. Our diet protocols included only 50% coverage of the recommended folate dosage per day according to the U.S. and German recommendations. The food composition in our clinic corresponds to an average folic acid provision of 273 μg/day without fortification. Considering the half-life of folic acid (100 days), it would make sense to reconstitute folic acid storage in patients who show a high risk for malnutrition because of multimorbidity during a hospital stay.

Increased Hcys levels are commonly associated with cognitive dysfunction, Alzheimer’s disease, dementia, and memory problems (1, 2, 10, 32, 33, 34). In contrast to other findings, (18, 21, 35) cognitive performance tests were not related to total serum Hcys levels in our study population. Budge et al. (17) and Kalmijn et al. (20) obtained similar results in non-multimorbid patient populations. The available data show that in the multimorbid patient, the Hcys value is increased by 20% over normal levels, even if the folic acid lies in the normal range. As described in previous works, Hcys levels can be decreased by folate fortification, (7, 28) and also small improvements in cognitive efficiency and memory function might be achieved with folic acid supplementation of 800 μg/day (30, 45). The recently published data of Kim et al. (44) that indicate a three-fold increased risk of dementia associated with folic acid deficiency in South Korea might support the recommendation of prophylactic folate supplementation.

In the present study, the observed Hcys levels appeared to be independent of dietary or serum vitamin concentrations, multimorbidity, multimedication, age, gender, or cognitive dysfunction. Upon adjustment for gender, age, multimorbidity, and multimedication, none of the univariate associations of SKT or MMSE score with folate, vitamin B₁₂, or Hcys level was confirmed. Our results in multimorbid patients with an average of nine diagnoses, nine different medications, and a median age of 79 years are important despite multiple influencing factors and different illness profiles. This population reflects patients with the highest risk profile, who are commonly encountered in geriatric medicine. Nevertheless, the question of when preventive folic acid substitution should be initiated for the reduction of Hcys levels remains unanswered. Hospital food might not be appropriate to fulfill the recommendations of daily folate intake in multimorbid elderly patients. Because recommendations for folate and vitamin B₁₂ determination are controversial, (42, 43) further studies in multimorbid patients are needed to analyze parameters that influence Hcys levels and cognitive function in this population.

In conclusion, we report for the first time elevated serum Hcys levels in multimorbid geriatric patients with normal plasma folate and vitamin B₁₂ concentrations. There were no significant differences in Hcys levels among different SKT score groups. Therefore, plasma Hcys does not appear to be an important biological marker of cognitive dysfunction in multimorbid geriatric patients.

Financial disclosure: None of the authors had any financial interest or support for this paper.