点击显示 收起
1 From the Department of Internal Medicine, Cardioangiology, and Hepatology, University Hospital S Orsola-Malpighi, Bologna, Italy (GR, PF, FM, GB, MM, TT, LS, and MZ), and the Laboratory of Immunology and Genetics, Codivilla Putti Research Institute, Rizzoli Orthopaedic Institute, Bologna, Italy (EM)
2 Supported by grants from the Italian Ministry of Education, University and Scientific Research-Ministero dell’Istruzione, dell’Università e della Ricerca Scientifica, Rome (fund for basic oriented research) and from Rizzoli Orthopedic Institues-Istituti Ortopedici Rizzoli, Bologna (Current Research Fund). 3 Address reprint requests to GR, Department of Internal Medicine, Cardioangiology, and Hepatology, University Hospital S Orsola-Malpighi, Via Massarenti, 9, 40138 Bologna, Italy. E-mail: ravaglia{at}almadns.unibo.it.
ABSTRACT
Background: Plasma concentrations of several amino acids may affect the availability of important neurotransmitter precursors in the brain. Abnormalities in the plasma amino acid profile have been reported in elderly persons with cognitive impairment, but no data exist for the prodromal phase of Alzheimer disease (AD), which is characterized by amnestic mild cognitive impairment (aMCI).
Objective: The objective was to investigate whether the plasma amino acid profiles of elderly patients with aMCI or AD are abnormal.
Design: The plasma amino acid profile was assessed in 29 cognitively normal control subjects (age: 86.7 ± 5.9 y), 21 patients with aMCI (age: 84.9 ± 7.0 y), and 51 patients with AD (age: 86.7 ± 5.4 y). The participants were from the University of Bologna Research Center for Physiopathology of Aging, Italy.
Results: Higher plasma concentrations of the aromatic amino acid phenylalanine were found in the aMCI (68 µmol/L; 95% CI: 63, 73) and AD (62 µmol/L; 95% CI: 59, 65) patients than in the control subjects (54 µmol/L; 95% CI: 48, 61; P < 0.05). The ratio of arginine to other basic amino acids was also higher in the aMCI (0.31 ± 0.04) and AD (0.27 ± 0.08) patients than in the control subjects (0.21 ± 0.05; P < 0.05). Adjustment for differences in body composition, serum vitamin B-12 concentrations, and serum folate concentrations did not significantly affect the results.
Conclusions: The plasma amino acid profiles of elderly patients with aMCI or AD show abnormalities in aromatic and basic amino acids that potentially affect neurotransmitter biosynthesis.
Key Words: Plasma amino acids • amnestic mild cognitive impairment • Alzheimer disease • dementia • elderly • aromatic amino acids • arginine
INTRODUCTION
Several amino acids act as precursors of important neurotransmitters and neuromodulators, such as catecholamines (phenylalanine and tyrosine), histamine (histidine), nitric oxide (arginine), and serotonin (tryptophan). According to experimental results, plasma concentrations of amino acids may be a regulating factor of neurotransmitter biosynthesis in the brain (1).
The prevalence of cognitive disorders increases exponentially with age (2), but only a few studies have investigated how aging affects the plasma amino acid pattern of elderly persons (3-6). Even fewer data are available about plasma amino acid concentrations in elderly persons in relation to their cognitive status (7, 8).
In a previous study of selected persons aged >90 y, we found that aging is associated with several abnormalities in the plasma amino acid profile, and that abnormalities in the plasma availability of the aromatic amino acid phenylalanine are more pronounced in elderly patients with cognitive impairment (8). This finding might be relevant to cognitive function because aromatic amino acids act as cathecolaminergic precursors (1).
In recent years, research on cognitive disorders has focused on amnestic mild cognitive impairment (aMCI), defined by the Mayo Clinic researchers as an isolated memory disorder that can precede Alzheimer disease (AD) (9). Diagnostic criteria for aMCI include memory complaint (preferably corroborated by an informant) and objective memory impairment at neuropsychological testing in nondemented persons with preserved global cognitive function. No data, however, exist about plasma amino acid concentrations in elderly patients with aMCI. Therefore, in the present study we characterized the plasma amino profiles of elderly persons with no cognitive impairment and of elderly patients with aMCI and AD.
SUBJECTS AND METHODS
Subjects
Participants were sought among the ambulatory patients seeking medical advice at the University of Bologna Research Center for Physiopathology of Aging, from September 2002 to September 2003. All procedures were performed after informed consent was provided by the subjects or their next of kin and after approval of the Institutional Review Board of the Department of Internal Medicine, Cardioangiology, and Hepatology.
A research-trained geriatrician (FM) administered a semistructured questionnaire to each participant and to a knowledgeable informant (usually an immediate relative). Subjects for whom a reliable informant was not available were excluded from the study. When the status report by the informant was inconsistent with the data gathered by the physician, clinical judgment was used to weigh the information. The subjects also underwent a general and neurologic examination, conducted by the geriatrician, and the Italian version of the Mini-Mental State Examination (MMSE), for which specific age- and education-adjustments were validated (10).
The general eligibility criteria were as follows: 1) age >74 y; 2) no history, symptoms, signs, or laboratory evidence of malignant disease; severe cardiovascular, pulmonary, hepatic, renal, endocrine-metabolic, or hematologic disorders; or acute or chronic inflammatory conditions that could affect cognitive function or the plasma amino acid profile; 3) no evidence of protein-energy malnutrition, defined on the basis of clinical judgment and a body mass index (in kg/m2) <20 for women and <22 for men (11); 4) no depression (defined as a score <15 on the Geriatric Depression Screening Scale; 12) and no psychiatric disorders or delirium, sensory-motor deficits affecting performance at the time of neuropsychological testing, neurosurgical history or neurologic conditions (eg, Parkinson disease, epilepsy, and postencephalitic and postconcussional syndrome), or psychoactive substance use.
On the basis of the informant and participant interviews, the geriatrician assigned a Clinical Dementia Rating (CDR) score (13) to each participant after clinical evaluation of 6 areas of cognitive function (memory, orientation, judgment and problem solving, community affairs, home and hobbies, and personal care). A 5-point scale was used to rate function (0 = normal, 0.5 = questionable/very mild impairment, 1 = mild impairment, 2 = moderate impairment, and 3 = severe impairment). When functional status reported by the informant was inconsistent with the data gathered by the physician, clinical judgment was used to weigh the information. A brain scan was ordered as clinically indicated. Outside medical records were obtained whenever possible.
Within a few weeks of clinical assessment, a second geriatrician (PF) with expertise in the evaluation of elderly persons with cognitive disorders administered the Mental Deterioration Battery (MDB), an extensive neuropsychological battery validated and widely used in Italy for the cognitive assessment of elderly persons (14). The MDB comprises 7 neuropsychological tests, from which 8 performance scores can be derived. The tests provide information about different cognitive domains: verbal memory (immediate and delayed recall of Rey’s 15 words), visual memory (immediate visual memory), language (phonologic word fluency and sentence construction), abstract logical reasoning (Raven’s progressive colored matrices), and constructional praxis (freehand copying of drawings and copying of drawings with landmarks). Administration, scoring, and age- and education-specific norms for all of these tests were previously described (14). Data from the neuropsychological battery were not used for clinical assessment or CDR assignment by the first geriatrician. On the basis of a review of both the clinical and neuropsychological data, the second geriatrician—who was blinded to the clinical conclusion of the previous examiner—rendered a CDR rating for each participant.
A senior geriatrician clinician (GR) reexamined, in concert with the 2 previously mentioned geriatricians, the case records of all the subjects who met the general eligibility criteria, and a final diagnosis of normal cognitive function, aMCI, or AD was made by consensus.
Cognitively normal control subjects
The subjects received a diagnosis of normal cognitive function if they met all of the following requirements: 1) an MMSE score 24, 2) report of intact basic (15) and instrumental activities of daily living (16) by the informant (dependency due exclusively to physical impairment was not considered), and 3) a CDR of 0 and age- and education-adjusted scores in the normal range on all the neuropsychological tests of the MDB.
Amnestic mild cognitive impairment
According to Mayo Clinic criteria, subjects received a diagnosis of aMCI if they met all of the following requirements: 1) memory complaints corroborated by an informant, 2) MMSE score 24; 3) CDR score 0.5; 4) scores below the normal reference range (1.5 SD below the mean for age and education matched norms) on any of the memory tests included in the MDB, but scores in the normal range on all the other tests; and 5), did not meet DSM-IV (Diagnostic and Statistical Manual of Mental Disorders) criteria for dementia (17) on the basis of clinical judgment.
Alzheimer disease
Subjects received a diagnosis of AD if they met the following requirements: 1) had a CDR 0.5 and met DSM-IV clinical criteria for dementia (17), and 2) received a diagnosis of probable or possible AD according to NINCDS/ADRDA (National Institute of Neurological and Communicative Disorders and Stroke/Alzheimer’s Disease and Related Disorders Association) criteria (18).
Anthropometric measurements
For all eligible subjects, body mass index was calculated as weight (in kg) divided by the square of the height (in m). Arm muscle area and arm fat area were calculated from midarm circumference and triceps-skinfold thickness by using standard formulas (19). Upper arm anthropometric measures provides better body-composition indexes than does body mass index in elderly subjects because of the high frequency of motor and spinal impairments affecting stature and weight measurement (20).
Laboratory
A venous blood sample was taken from each participant between 0600 and 0900 after they had fasted overnight. Samples were put on ice and processed within 1 h of collection. Empty plastic tubes were used for serum collection. Evacuated tubes containing potassium EDTA were used for plasma collection. Supernatant fluid was carefully removed after centrifugation at 3000 x g for 30 min at 4 °C. This procedure, according to previous tests performed at our laboratory, avoids contamination of the supernatant fluid with platelets and leukocytes, which are very rich in taurine. Plasma aliquots were kept frozen at –70 °C until analyzed (usually within 6 mo), and concentrations of plasma amino acids were measured with a model 3A30 Carlo Erba Amino Analyser (Carlo Erba-Fisons, Rodano, Milan, Italy) as previously described (21). In our laboratory, the intraassay and interassay CVs in the measurement of plasma amino acids were ± 5% and ± 10%, respectively. Serum folate and vitamin B-12 concentrations were measured in fresh serum by immunoelectrochemiluminescence analysis (Elecsys Folate Immunoassay and Elecsys B12 Immunoassay for Elecsys 2010 System; Roche Diagnostics Italia S.p.A. Monza, Milano, Italy). Serum creatinine, plasma total cholesterol, serum albumin, and serum C-reactive protein concentrations were also assayed in fresh serum as previously described (8).
Calculations
Valine, leucine, and isoleucine were summed as branched-chain amino acids (BCAAs). BCAAs plus the aromatic amino acids tyrosine (Tyr) and phenylalanine (Phe) were summed as large neutral amino acids (LNAAs). BCAAs along with phenylalanine, methionine, threonine, lysine, and histidine were summed as essential amino acids (EAAs). Alanine, glycine, serine, glutamine, proline, arginine (Arg), taurine, glutamine and glutamic acid, asparagine and aspartic acid, ornithine, citrulline, and cystine were summed as nonessential amino acids (NEAAs). From a strictly metabolic perspective, many of these amino acids (particularly proline, arginine, taurine, and glutamine) are considered "conditionally essential" because, in critically ill adults, their consumption can exceed the biosynthetic capacity of the organism and the dietary supply may become crucial (22). However, because our study criteria excluded persons with malnutrition and severe diseases, for simplicity’s sake, we chose to consider all of these amino acids as nonessential. Arginine, ornithine, lysine, and histidine were summed as basic amino acids (BAAs).
The amino acid ratios EAAs:NEAAs and Phe:Tyr were calculated because they are typically low when there is a deficit in dietary protein (23); the ratios Tyr:other LNAAs, Phe:other LNAAs, and Arg:other BAAs were calculated because they are better predictors of phenylalanine, tyrosine, and arginine availability to the brain than are the individual plasma concentrations (1).The TSM ratio (defined as the ratio of 100 times the plasma taurine concentration and the product of serine and methionine plasma concentrations) was calculated because this ratio is a marker of the status of the amino acids involved in transmethylation processes; it has been reported to be elevated in AD (7).
Statistics
Data are reported as means ± SDs or as the number and percentage, except for variables that had skewed distributions (phenylalanine, tyrosine, Phe:Tyr, arginine, taurine, and the TSM ratio). These variables were normalized by log transformation and reported as geometric means and 95% CIs.
Sex-related differences among each study group were tested by Student’s t test. Because of the low number of subjects and the lack of sex-related statistically significant differences in any of the variables of interest, the men and women in each study group were pooled.
Differences across the study groups were evaluated by analysis of variance (Tukey’s test for all-pairwise multiple comparisons) or chi-square test as appropriate. B vitamins are known determinants of cognitive function in the elderly (24), and plasma concentrations of several amino acids are associated with muscle and fat mass as estimated by upper arm anthropometric measures (8). Therefore, a general linear model including serum vitamin B-12 and folate concentrations, arm muscle area, and arm fat area as covariates was used to further test differences in the plasma amino acid profile across the study groups (Tukey’s test for all-pairwise multiple comparisons).
The univariate associations of the plasma amino acid profile with the age- and education-adjusted MMSE and MDB scores of all participants taken as a whole were assessed with Pearson’s correlation coefficient. Statistically significant associations at univariate analysis were further tested by a multiple regression linear model including the same nutritional covariates entered in the general linear model. Statistical analyses were performed with SYSTAT10 (SPSS Inc, Chicago). All tests were two-tailed, and a P value < 0.050 was considered significant.
RESULTS
Of the 124 eligible subjects, 23 were excluded because they did not meet the diagnostic criteria for any of the conditions of interest (6 had nonamnestic MCI, 17 had non-AD dementia). The characteristics of the study groups are shown in Table 1. The groups did not differ by age, sex, or any of the biomedical variables considered in the table, except for arm fat area and serum folate. Arm fat area was lower in the AD patients than in the control subjects, and serum folate was lower in the aMCI and AD patients and than in the control subjects. CDR scores for patients with AD ranged from 0.5 to 3 (2 ± 1).
View this table:
TABLE 1. Clinical features of the control subjects and of the subjects with amnestic mild cognitive impairment (aMCI) or Alzheimer disease (AD)
The age- and education-adjusted scores of the study groups on the neuropsychological tests are shown in Table 2. Because of severe cognitive impairment (MMSE score <10), the MDB could not be administered to 6 patients with AD. Consistent with the selection criteria, the patients with aMCI did not differ significantly from the control subjects in terms of the MMSE score, taken as a measure of general cognition, but they performed more similarly to patients with AD than to control subjects on the MDB measures of verbal memory. Moreover, as previously reported by Petersen et al (9), even when scores were in the normal range, the aMCI patients performed worse than did the control subjects on measures of language and visuospatial praxis. As expected, the patients with AD had lower scores on the MMSE and the MDB (all tests) than did the control subjects.
View this table:
TABLE 2. Neuropsychological features of the control subjects and of the subjects with amnestic mild cognitive impairment (aMCI) or Alzheimer disease (AD)
The amino acid profiles of the study groups are shown in Table 3. For all groups, the average Tyr:other LNAA ratio fell above the upper limit of the normal reference range for a young population (data from 60 healthy and cognitively normal subjects, 30 men and 30 women, aged 25-50 y; 8). The aMCI and AD patients had higher plasma phenylalanine and Arg:other BAA ratios than did the control subjects. Patients with AD had higher taurine and TSM ratio values than did the control subjects. Moreover, their average EAA:NEAA ratios fell below the normal reference values for young persons. No statistically significant difference across the study groups was found for the EAA:NEEA, Phe:other LNAA, and Tyr:other LNAA ratios or for plasma concentrations of BCAAs, tyrosine, arginine, histidine, and methionine. The plasma amino acid profile did not significantly differ between patients with probable (n = 40) and possible (n = 11) AD. The results were not affected when other nutritional confounders were entered as covariates, except for differences in taurine and the TSM ratio, which were no more statistically significant (P > 0.100).
View this table:
TABLE 3. Fasting amino acid profile of the control subjects and of the subjects with amnestic mild cognitive impairment (aMCI) or Alzheimer disease (AD)1
In univariate analysis, the only statistically significant associations between the measures of cognitive function and the plasma amino acid profiles of all study participants considered as a whole were those between MMSE scores and the TSM ratio (r = –0.311, P = 0.003) and those between immediate memory recall of Rey’s 15 words and the Phe:other LNAAs (r = 0.261, P = 0.008). When other nutritional confounders were accounted for, however, only the inverse association of MMSE score with the TSM ratio remained statistically significant (P = 0.010). No association was found in AD patients between the plasma amino acid profile and the severity of dementia on the basis of CDR scores.
DISCUSSION
This is the first study to report specific abnormalities of plasma aromatic and BAAs in elderly patients with aMCI selected according to the diagnostic criteria proposed by the Mayo Clinic researchers (9). Another important strength of our study was the strict selection criteria adopted to avoid the confounding effect of any pathologic condition, other than cognitive problems, on plasma amino acid concentrations.
This study also had several important limitations. First, extrapolation to brain function of any abnormality in the plasma amino acid profile requires extreme caution. Second, the method used to assess the plasma amino acid profile did not allow a reliable measurement of free tryptophan. Therefore, we did not include tryptophan among the study variables, although this amino acid is an LNAA and a precursor of the neurotransmitter serotonin. Third, although diet can have marked effects on plasma amino acid concentrations (25), the diet of the study subjects was not controlled, and measures of dietary intake of energy and protein were not available. None of our study subjects, however, was following special dietary regimens, and the drawing of plasma samples before the morning meal may have attenuated diet-related differences in amino acid concentrations. Moreover, no subject had clinical, anthropometric, or laboratory signs of undernutrition. With respect to reference values for young persons, however, the average EAA:NEAA ratio was below the lower limit for patients with AD and borderline low for both cognitively normal persons and patients with aMCI. This finding agrees with previous results in elderly populations (3, 8) and is considered suggestive of replacement of dietary proteins with less expensive, more palatable, and more easily chewable carbohydrates (26).
Confirming previous results in elderly persons (8), the average Tyr:other LNAA ratio of all the study groups, independently of cognitive status, fell above the normal range for young persons. According to Rudman et al (27), aging might be associated with a slowing in the capacity of the homogentisic acid pathway, the metabolic pathway responsible for catabolism of phenylalanine and tyrosine. This impairment could potentially reduce dietary requirements and tolerance to aromatic amino acids in elderly persons.
In this study, plasma phenylalanine concentrations were higher in both the aMCI and AD patients than in the control subjects. Elevated phenylalanine concentrations may lead to brain damage through several mechanisms, including competition with other LNAAs (including tryptophan) for transport by the same carrier at the blood-brain barrier (1, 28), decreased brain protein synthesis, and increased myelin turnover (29). In inherited disorders of phenylalanine metabolism, however, neurologic complications are very uncommon at plasma phenylalanine concentrations <600 µmol/L (30). Data about the possible role of abnormalities in plasma aromatic amino acids in AD are scant and contrasting (7, 31), but an excess of phenylalanine has been consistently associated with delirium (32, 33).
Both the aMCI and AD patients also had Arg:other BAA ratios that were higher than those of the control subjects. Similarly to LNAAs, BAAs compete with each other for active transport across the blood-brain barrier (1). Arginine, in particular, is a precursor for polyamines and nitric oxide (34), both of which are important modulators of neuronal physiology and are thought to be involved in AD pathogenesis (35, 36).
Some caveats, however, must be made against an overinterpretation of these results in relation to neurotransmitter synthesis and cognitive function. First, in healthy adult subjects, convincing experimental evidence of associations between cognitive performance and changes in fasting plasma concentrations of specific amino acids has been reported for tryptophan only (37). Second, the differences observed in this study were of small magnitude, and further research is needed to evaluate their clinical significance. However, according to findings in elderly febrile patients, development of delirium was associated with differences in plasma phenylalanine as small as 15 µmol/L (32).
In agreement with the results of Fekkes et al (7), patients with AD were further characterized by an increased TSM ratio with respect to the control subjects. This alteration suggests a dysfunction in transmethylation reactions, fundamental in brain cells for the synthesis of myelin, neurotransmitters, and membrane phospholipids (24). Transmethylation dysfunction has been hypothesized to have a causal role in AD (7, 24). According to our results, however, the impaired transmethylation found in patients with AD was mainly related to their poorer nutritional status, especially with respect to folate, which is an essential cofactor for this metabolic pathway (38).
Finally, considering all of the study participants as a whole, we found a significant inverse relation between MMSE score and TSM ratio that, contrary to the higher TSM ratios of AD patients, was independent of differences in serum folate and body composition. This relation might mirror a similar one reported for MMSE and the sulfur amino acid homocysteine (39), a methionine derivate whose increased plasma concentrations in elderly persons may also be a marker of a folate deficit and dysfunctional transmethylation (38) and for which direct neurotoxic effects have been reported (40, 41).
Some results of the present study differ from our previous findings in a sample of persons aged >90 y with miscellaneous types of dementia and cognitive impairment without dementia (defined on the basis of an MMSE score below the normal range but with cognitive problems not severe enough to meet dementia clinical criteria) (8). Both diagnostic groups were characterized as having a higher Phe:LNAA ratio than that of the control subjects, a nonstatistically significant trend (P = 0.051) toward higher plasma arginine concentrations, but no abnormalities in plasma phenylalanine. These differences may have been due to the different ages of the study populations, the different criteria used to diagnose mild cognitive impairment, and the heterogeneous types of dementia included in the study of the oldest-old patients.
In conclusion, our findings confirm that aging is associated with several abnormalities in the plasma availability of amino acids that may act as neurotransmitter precursors. Our study also suggests that specific abnormalities in the plasma amino acid profile may occur in the early stages of cognitive impairment. Longitudinal studies in larger samples are required to confirm these findings and to clarify their physiopathologic significance in aMCI and AD.
ACKNOWLEDGMENTS
GR was the main contributor to the study design. PF and FM substantially contributed to data collection, data analysis, and preparation of the manuscript. EM, GB, and MZ contributed to expert methodologic advice and editing of the manuscript. MM, LS, and TT contributed to data collection and interpretation. None of the authors had a financial or personal interest in any organization sponsoring the research or advisory board affiliations.
REFERENCES