Nutritional Status of Free-Living Alzheimer’s Patients BY MARIAN J. JOE W. RAMSDELL, MD,* RENVALL, MS, RD," AUDREY A. SPINDLER, PHD, RD,t MONICA. PASKVAN, MS, RD+ ABSTRACT: Self-reported, dietary intake and biochemical estimates of thiamine, riboflavin, folate, vitamin B-12, protein, and iron were compared in 22, free-living elders by individu- als who had senile dementia of the Alzheimer’s type (SDAT) and in 41 who were cognitively normal (CN). The two groups did not differ sig- nificantly in their intake of these nutrients or the number of deficiency states for intake (<67% RDA). Low serum transketolase (thia- min; p < 0.055), red blood cell (RBC) folate (p < 0.08), and serum vitamin B-12 (p < 0.05) lev- els occurred more often in SDAT patients than in CN subjects. Individuals in both groups who used multivitamin supplements had signifi- cantly higher biochemical values for thiamine (p <0.08), riboflavin (p < 0.01), and vitamin B- 12 (p < 0.003) than nonsupplement users. Be- cause of the differences in vitamin B-12 and RBC folate levels between groups, a retrospec- tive analysis was performed on a larger group of subjects drawn from a geriatric assessment clinic. Patients with SDAT had significantly lower serum vitamin B-12 (p < 0.01) and lower RBC folate (p < 0.03) values than GN subjects. When mean values for vitamin B-12 and RBC folate were grouped by degree of impairment in SDAT subjects, vitamin B-12 was significantly lower in mildly and moderately impaired sub- jects than in those with normal cognition. Mean, ‘From the "Department of Medicine, Division of General Internal ‘Medicine /Geriatres, University of California Medical Center, San ‘Diego, and the tchool of Family Studies and Consumer Sciences, San Diego State University, San Diego, California. ‘Funding was provided by BRSG $07 RROTOO4-10, NIH, SDSU PACAFBACT 1-1116, and SOCARE. his paper was a poster presentation atthe American Federation {for Clinical Research, San Diego, California, Mey 4, 1987, and a7 ‘oral presentation at the Second International Congress of Biomed (cal Gerontology, Hamburg, BRD, Jul) 16, 1987. ‘The authors wish to chank Dr. Claire Murphy for compensation. ofthe BOO subjects, and Shideh Mofdfortchnieal assistance, ‘Reprint requeste: Marian J. Renvall, MS, RD, Department of Medicine H-811-E, UCSD Medical Centr, 225 Diekinsan Street, ‘San Diego, CA 92108-1990. 20 values for both nutrients did not differ signift- cantly between severely impaired and CN sub- jects. There was a significant quadratic rela- tionship between cognitive impairment and biochemical values for vitamin B-12. The au- thors concluded that the self-reported dietary intake of free-living SDAT patients is similar to that of cognitively normal elderly people, and multivitamin supplementation significantly in- creases biochemical measurements for nutri- tional status. Further studies are needed on the importance of low serum vitamin B-12 and RBC folate, particularly as they relate to the degree of cognitive impairment. KEY INDEX- ING TERMS: Elders; Alzheimer’s tamins; Folate; Vitamin B-12; Nutritional Sta- tus. [Am J Med Sei 1989; 298(1):20-27.] S is imated to exist in 33-60% of the US elderly population. Precise identification of subclinical malnutrition is hampered by recommendations for intake of nutrients, which group together all people older than 51 years of age. ‘This problem is further confounded by poor agree- ment between dietary and biochemical data, wider variations of normal limits for biochemical values in older people compared to younger people, and the pos- sible impact of concomitant disease. Morgan et al’ re- ported litele difference in mean values for anthropo- metric, dietary, and biochemical measurements be- tween young controls and healthy, active, elderly subjects. However, mean values for all measurements declined directly with increasing severity of disease in five groups of elderly patients ranked by degree of illness. Values for psychogeriatric patients fell mid- way on the continuum, ‘The availability of information on the prevalence of subclinical malnutrition in elderly patients with cognitive impairment is limited, In studies of dietary intake of nursing home residents, no significant differences were found between patients with senile July 4989 Volume 298 Numbor +dementia of the Alzheimer’s type (SDAT) and cogni- tively normal (CN) patients." However, Litchford and Wakefield® found that SDAT patients consumed significantly less energy (Keal), thiamine, and ribo- flavin than CN patients, In addition, Singh et al‘ found no differences in hemoglobin, hematocrit, albu- min, serum vitamin B-12, and folate levels, but pa- tients with SDAT patients weighed 14% less than those with multi-infarct dementia, and 21% less than nondemented patients. These findings suggest that subclinical malnutrition may be significant in cogni- tively impaired individuals. Furthermore, elderly peo- ple who have nonreversible dementias may be at greater risk for nutritional deficiencies than mentally normal elders" Cognitive impairment ranging from mild memory loss to severe dementia may be produced by deficiency of one or a combination of several vitamins.® Even within groups of active, healthy elderly people, those in the lower 10th percentile for intake (vitamin C and folate) and blood levels (vitamins C, B-2, B-12, and folate) scored significantly lower on tests for cognitive function than individuals within the 90th percentile. ‘The authors here tested the hypotheses that dietary intake and biochemical values of vitamin B-12, folate, thiamine, and energy would be lower in SDAT pa- tients than in elderly patients with normal cognition. Methods Subjects. Blderly people who wore at least 60 years of age and free from acute illness or recent hospital- ination were selected from SOCARE, a comprehen sive outpatient geriatric assessment program at the University of California, San Diego Medical Center and from Educational Growth Opportunities (EGO), fa program at San Diego State University for elderly individuals. ‘The study was approved by the human subjects committees at the University of California, San Diego, and San Diego State University, and sub- jects or their caregivers gave written, informed con- sent for voluntary participation. ‘Subjects were grouped as CN or having SDATT. The type of demen- tia!” and degree of impairment’ were determined by consensus of the SOCARE medical team composed of internists, psychiatrists, and neurologists. All EGO subjects were found to be cognitively normal as tested by the Blessed examination." Sixty-nine subjects met the selection criteria. Six subjects who did not complete the 3-day dietary rec- ord were excluded from the study. Biochemical val- ‘ues, age, Mini-Mental State Examination (MMSE) scores, and living situation did not differ between the excluded and included subjects. Mean body weight of the excluded subjects was significantly (p < 0.006) lower than that of the participants. Individual records were reviewed, and the lower weight was not caused by recent weight loss within the last year. THE AZERICAN JOURNALOF THE MEDICAL SCIENCES Renvail et al Not all biochemical values were obtained from the 63 subjects completing the dietary record. Informa- tion for each parameter is presented only for those subjects from whom both dietary and biochemical data wore available. Diotary Status. Subjects or their caregivers were given oral and written instructions in keeping written food consumption records for 3 days. The records ‘were analyzed by computer program (Ohio State U: versity) and missing values were added from standard handbooks of food composition. Intake of nutrients was converted to percentage of the Recommended Di etary Allowances (RDA) by sex to estimate the ad quacy of the subjects’ diets. Multivitamin supple- ments were taken sporadically by many of the sub- jects; therefore, values from supplements were not ‘added to those of the diets. Nutritional status. Blood samples drawn from non- fasted subjects were analyzed for protein and albumin in serum and for standard indices (hematocrit, hemo- globin, mean corpuscular volume, red blood cell {RBC} count) by automated methods (Beckman, Brea, CA). A radioassay method," that used separate isotopes was used to determine serum vitamin B-12 and RBC folate (Quantaphase B-12/Folic Acid, Bio- Rad; Richmond, CA). The activity of RBC transketo- lase was used to estimate thiamine status.” ‘The transketolase assay is the most widely used for esti- mation of thiamine status. However, the ways of ex- pressing the activity vary, and there is little agree- ment among them."* Activity herein is expressed as micromoles of either sedoheptulose-7-phosphate (BK'TAS) produced-ml blood-hr™," without the thiamine pyrophosphate effect or as pentose-6-ph phate (EKTAPH) used-hr"?-g hemoglobin." The boflavin content of casual urine samples was mea- sured fluorometrically and corrected by creatinine coneentration."” Stalistical Analyses, All analyses used the BMDP statistical software package."® Analysis of variance (ANOVAR) followed by Bonferroni tests, when ap- propriate, was used to measure differences among means for dietary and biochemical parameters. For this analysis, subjects were grouped by 1) cognition; 2) sex and cognition; or 3) supplement use and eogni- tion. The effects of age on MMSE scores, dietary in- take, and biochemical values were estimated by cl square analyses. Differences in the number of defi- ciencies by nutrient were calculated by chi-square analyses between SDAT and CN groups. All data in the tables are reported as means + stan- dard deviations (SD), with the number of subjects in parentheses. Ninety-five percent confidence intervals (CI) are reported in brackets. The exact significance of the analysis of variance for each parameter is given, and symbols are used to denote significance by the Bonferroni tests. 24Nutrition in Alzheimer's Patients ‘Table 1. Description of Bidrly Subjects with SDAT or Normal Cognition SDAT. Noniat Men ‘Women Men ‘Women Number 15 18 28 Ager) Tesi mas 5° nat 36t p<0001 Range 66-83 0-78 65-80 61-83 Number of modications ats 10 ‘Weight (ke) 636104 ‘Multivitamin users 50% 42% “p< 00s, tp 80 years. Protein, Energy, and Iron Status. There were no significant differences among groups in dietary or bio- chemical measures of protein, energy, or iron status (Table 2). For these nutrients, mean consumption levels exceeded 67% of the RDA. Although mean en- ergy (Keal/kg) consumption was higher in the SDAT. ‘men (30.0 # 9.0) and women (26.0 + 9.0) than in CN ‘men (26.0 + 8.0) and women (25.0 + 7.0), body weight (kg) was lower in the SDAT men (71.8 + 8.2) and ‘women (60.9 + 9.1) than in the CN men (75.9 = 12.7) and women (63.6 + 10.4). Mean values for iron and protein status measures in blood were normal and did not differ between cognitive groups. Hemoglobin and hematocrit values, which were lower in women, showed the typical significant differences according tosex. ‘Table 2. Protein, Bnergy, and Iron Status of Bderly Subjects ‘with SDAT or Normal Cognition Normal SDAT Subjects Subjects Distary values Protein, g/kg 102 0200)" — 1040 28) (as)tp <0.62 (os-a}t (09-12) Energy, Keal/kg 289105(10) 5.88.1 (28) (@8)p <035 (214-264) (22.7-289), Tron, mg/day 116 69() 11123204) (8) p <0.73, (8a-143) (00-123), Biochemical values ‘Sorum protein ¢/ 71 0700) 6.7=0.4(28) (G0) p <0.10 (60-76) (65-68) Serumalbumin,e/dl 40 0600) 4.20.3 (28) (8) p <049 (36-45) (40-43) Hematocrit, % 408+ 3300) 999436(04) (ip <034 (29.2-42.4) [as6-41.1) Hemoglobin, g/d 1372 13@0) 18613 (34) Gap<08 tis1-143) * Numbers in parentheses indicate number of subjects +t Numbers in brochets indicate 95% Cl. ly 1989 Volume 298 Number 4‘Table 3. Vitamin Status of Ferly Subject with SDA'T ‘or Normal Cognition SDATSubiecte Normal Subjects 14 oss 122 0400) (as)tp <034 {o-taif {ial-ta) Rite; mg/day ae aan gs Og (Gp 2088 2-13) ea) Potten 19'S ean 188 = TA 0) (snp cost 40-105) ise-226) ‘iin B12 pa/day eas isan 22 230) (Gai <03i ssi) 20-49} DBigchemeal ales BIKAS IUmlhe! 43 1805) a9 16@9) {4a} < 049 1452), (33-46), BEKAPH, IU. ming "b* 1s2 0500 18s 048) a2} <0085 (8-18) i720) Ribovin ato isa sas 80@n Gp <04, 108.33) tics} RBC folte n/m aos "Sais en) 2 12380. (0) (np 608 (seo-s12) sic) Serum vitamin B12, esa 42 197 nea 2178 a2) (hp zoo 1980-805) Wéa-s20) * Number in parentheses inleate number of subjects + Nubers in brackets indice 95% CL Vitamin Status. The self-reported, dietary intake of vitamins was similar in both groups of subjects (Table 3), Except for folate, the means for intake exceeded. 67% of the RDA. Folate levels in food composition tables are frequently missing, and low folate intake needs to be evaluated in terms of the individual's health status and other nutrient intake. There was no significant difference in biochemical status for ribo- flavin between groups, but SDAT patients had lower thiamine (p < 0.055), RBC folate (p < 0.06), and se- rum vitamin B-12 (p < 0.047) values than those of CN individuals. ‘Number of Deficiencies. ‘The number of subjects with deficient intake (<67%) of nutrients was similar in both subject groups. The number of low biochemi- cal values did not differ significantly between the two groups, except for albumin (data not shown). Albu- min levels ranged from 3.2-4.1 g/dl. Levels below 3.4 g/dl for free-living, healthy adults are considered low. ‘Two SDAT subjects had low levels, while subjects in the control group all had values above 3.4 (chi-square 9; p < 0.02). ‘Multivitamin Supplementation. Use of vitamin sup- plements (Table 4) increased values for thiamine (BTKAPH only; p < 0.03), riboflavin (p < 0.01), and vitamin B-12 (p < 0.003). Supplementation resulted in excretion of excess riboflavin because values were 40-50 times greater than normal. Significance was lost with the post hoe tests, except that the normal subjects, who used supplements, had higher values for EKTAPH (p < 0.05) and serum vitamin B-12 (p ‘< 0.001) than SDAT subjects who took none. "HE AMERICAN JOURNAL OFTHE MEDICAL SCIENCES Renvail et al Relationships among Measurements. The self-re~ ported, dietary intake and biochemical values were weakly correlated, a frequent observation in elderly people. Correlation coefficients (data not shown) be- tween the measures of nutritional status were higher in CN than in SDAT subjects. Retrospective Study. Population Characteristics. De- mographie data for age, sex, living situation, marital status, and number of medications did not differ sig- nificantly from those in the initial study. ‘The mean age (75.8 & 9 years) in the 208 subjects was similar, and the sex distribution was 66% female. One hun- dred and fifty-four subjects met the DSM II criteria, for SDAT, and 49 subjects were diagnosed as having no dementing illness. ‘Mean body weight grouped by sex and cognitive status differed significantly (p < 0.0065). SDAT women weighed significantly less than the CN fe- males (p <0.01), but there was no difference in weight, between the two groups of men (Table 6). Vitamin Status. RBC folate and serum vitamin B- 12 levels were lower for subjects with SDAT com- pared to normal controls (p < 0.08 and p < 0.004, re- spectively), as presented in Table 6. RBC folate and serum vitamin B-12 values ini declined directly with MMSE scores, but the mean values for both nutrients in the most severely im- paired subjects, whose MMSE scores were <11, dem- onstrated an upward inflection. Equations for @ qua- dratie offect wore significant for vitamin B-12 only (p ‘<0.023). Analysis of covariance showed that multivi- tamin supplementation had a significant interaction on serum levels of vitamin B-12 (p < 0.06), while age did not affect the serum levels significantly. Folate had a similar quadratic curve, with the group scoring lowest on the MMSE (<11' points) displaying in- creased mean values for RBC folate, and the group scoring between 19 and 25 points having the lowest mean values; however, the ANOVA was not signi cant (Table 6). Discussion ‘The self-reported intake of nutrients did not differ between the patients with SDAT and normal elderly, individuals, and nutrients were not consumed at lev- els low enough to be the primary cause for nutrient, deficiencies. For both studies, the use of multivitamin supplements was associated with higher biochemical measures of nutritional status. Serum vitamin B-12 and RBC folate levels were significantly lower in the SDAT subjects than in CN subjects. Other major findings in the retrospective study include 1) signi cantly lower means for body weight in SDAT than in CN subjects; 2) significantly lower means for serum vitamin B-12 values in all moderately and mildly i paired subjects than in CN subjects; and 3) the al sence of statistical difference for mean values of se- 23Nuttition in Alzheimer’s Patients ‘Table 4. The Eiect of Supplementation on Vitamin Status in Elderly Subjects ‘Thiamine BTKAS1U-mi"-he"* 4o2 20 38217 ox 12 402 16 (8) p <048 o (i) @) 5) ETKAPHIU/e/hgb 17s 08 19s 04" Laz 06" aie 03 (2) p <0.03 © (5) ®) (as) Riboflavin ratio. are 21 ars 76 oss. 04 10 06 Gi)p < 001 ® cy Oo (ds) RBC folate, ng/za 488 245 551 245 sm 2169 61s 298 (1p <007 a3) © as) © (86) p 0.038 500 = 148" 542 2253 354 = 147" 485 207 a5) © a) aa) Serum vitamin B-12, pg/ml 519 = 122 500 170° 313 Ells" 498 161 (43) p< 0.008. ao) as) a @ (203) p <0.02 499 E148 59 206° 435 = 198°t 599 ast ey Go) casa) (8) * Significantly different means by Bonferroni correction: p < 0.01. 1 Significantly different means by Bonferroni: p< 0.05 rum vitamin B-12 between the severely impaired and CN groups. Blass and Zemcov" have emphasized that SDAT patients in terminal stages often develop cachexia, despite the fact that they receive more than adequate feeding by their caregivers. The subjects in this study. were all free living; no one was cachectic, but some biochemical values for nutrients differed from those expected from the dietary intake. The RDA values were established to ensure optimal nutrition in healthy populations. Because SDAT subjects are ill, it may be that higher levels of specific nutrients are needed to provide adequate nutrition. Recording of dietary data by proxy is often unreliable” and, in ad- dition to exclusion of the contribution to intake made by supplements, proxy recording may have contril uted to the weak correlations observed between di- etary and biochemical data in the subjects. Alll groups had adequate riboflavin, iron, and pro- tein intake. All individuals had adequate dietary sta- tus for vitamin B-2, while a few subjects had values ‘Table 5. RBC Folate, Serum Vitamin B-12, and Weight from the Retrospective tidy SDATSubjccla Normal Subjects ‘Blochemia eaton "RBC flat g/ml 980 £158. (es)* 52 2242 18) Coren (sricaasl iene] SermmvieminB int 40 ia) toy oo iam (dbistoo} 08-647 Bex mss) sex 194) Sire 10306) TS 630) * Number parentheses Indicate number sat 1 Munters in brackets indicate 95% Cl 24 suggestive of incipient iron or protein malnutrition. ‘Morgan ot al? found that indices for protein nutriture progressively deteriorated as illness and physical de- pendency increased. Values given for albumin by ‘Morgan et al* for psychogeriatric patients, who were attending a day hospital on a regular basis, do not differ from those of the SDAT subjects in the present study. This finding is relevant to the present study in that individuals with dementia become more depen- dent as their disease progresses and, therefore, may become more likely to exhibit concomitant decline in nutritional status, Many studies"! report pro- tein and/or protein energy malnutrition in SDAT pa- tients based on biochemical or anthropometric mea- surements, despite presumed adequate intake of pro- tein and/or Keal. ‘Thiamine nutriture, as assessed by means for di- etary intake and the transketolase enzyme assay, was adequate in most subjects. However, 28% (6 SDAT. and 8 CN) of the 48 elders had thiamine intakes of <1 mg/day, which is the minimum recommended for people consuming diets that contain fewer than 2,000 Keal/day. Normal values for transketolase activity are unclear. In normal subjects before a 14-day treat- ment with 100 mg vitamin B-1, Graudal"® found a mean of 1.03 (0.78 1.98) for ETKAPH and a mean of 1.37 (1.01 + 2.30) after treatment. Normal ETKAS. values were 2.44 + 0.54."" The authors found higher mean values for ETKAS and ETKAPH in all groups than those previously reported." Kjosen and Seim"* reported ETKAS activity in patients with pernicious anemia (82.5 + 29.6) nearly twice that of controls (46.2 9.5); however, measurements in this case were ETKAS with hematocrit adjusted to 35%. Because as serum vitamin B-12 values were not given and ET- KAS activities are expressed differently in the study July $989 Volume 298 NumberRenvall et al ‘Table 6, Red Blood Col Folate and Serum Vitamin B-12 Grouped by Mini Mental State Examination Scores MMSE Scores 0-10 18 19-25 26-90 [RBC folate, ng/ml ‘Mean 4782318 4352218 395+ 220, 5612 202 95% C1 + (28-606) [359-514] [en186), [420-682), Nambor 18 37 15 18 ‘Namber < 250 ng/a 4 6 3 1 ‘Serum vitamin B-12, p/m "Mean 47e201 4342 198" aus 1731 556-4 218"4 85% C1 [12-51] (87-490) [392-497], [490-623], ‘Number 4 0 6 a ‘Number < 250 ng/ml 7 10 8 3 “+p < 0.05 by Bonferroni test. bby Kjosen et al, itis difficult to compare those findings to the present study. However, BTKAS was high SDAT subjects than CN individuals, and serum vita- min B-12 was lower in SDAT than GN subjects. ‘The exact mechanism for greater sedoheptulose formation (allowing greater values for ETKAS) in SDAT subjects than in CN subjects is not clear. It hhas been suggested that in Alzheimer’s disease, the pyruvate dehydrogenase complex is damaged,"® which may account for more available substrate for metabo- lism through the hexose monophosphate pathway. ‘The SDAT subjects who used multivitamin supple- ments had the highest ETKAS values, potentially al- lowing more thiamine pyrophosphate to be available. Because hemoglobin values were similar in SDAT and CN subjects, itis not known why the effect of supple- mentation observed with EKTAPH is not observed with ETKAS. Further investigation is needed to clar- ify the meaning of elevated transketolase activity found among elderly subjects in this study to deter- mine whether it occurs in-all subjects consequent to vitamin B-1 supplementation or in SDAT subjects in conjunction with low vitamin B-12 and deranged car- bohydrate metabolism. Forty-four percent (10 SDAT and 9 CN) of the pop- ulation had poor intakes of vitamin B-12 (<50% RDA, and 67% Recommended Daily Intake [RDI]). Herbert suggests a RDI for vitamin B-12 of 2 ug/ day. In the prospective study, no subject had a serum level indicative of deficiency (<150 pg/ml), but two subjects (one from each group) had low normal levels (<250 pg/ml). In the retrospective study, 25 (16%) of 155 SDATT sind 3 (7%) of 49 CN subjects had very low normal values (<250 pg/ml). Because pernicious ane- mia takes 8 to 10 years of inadequate vitamin B-12 intake to develop, the frequency of poor dietary vita- min B-12 consumption should be of concern. In both studies, subjects who voluntarily consumed a vitamin supplement had higher serum vitamin B-12 values than those who did not consume a supplement. An THE AMERICAN JOURNAL OF IHEMEDICAL SCIENCES inability to cleave vitamin B-12 from protein in foods has been reported to exist in elderly patients who have intrinsic factor.”* Because vitamin B-12 supplements contain the vitamin unbound to protein, this may ac- count for higher serum values in subjects consuming vitamin supplements. Marcus et al also have de- tected a decline in the delivery of vitamin B-12 to tis- sues in elderly people resulting from a change in transcobalamin II structure and function. The rel vance of poor intake coupled with losses in bioaval ability are ill defined at present. Whether or not higher levels of serum vitamin B-12 have a physio- logic benefit for elderly individuals, especially those with SDAT, needs to be determined, Lower levels of vitamin B-12 in the sera of SDAT. patients than in those with other dementing illnesses, or in mentally unimpaired elderly people have been reported.**” In cases of SDAT, the authors found that the mean for serum vitamin B-12 (pg/ml) was similar to the value of 472 + 270 reported by Karnaze and Carmel,” but higher than the level of 219 found by Cole and Prchal.* The mean level of 367 pg/ml for se- rum vitamin B-12 in CN elderly people" was also lower than that found for CN subjects in this study. Poor intake does not explain the difference between the vitamin B-12 status of SDAT subjects and elders with other types of dementia who would be expected to have equally poor dietary intakes. Whether the higher mean serum vitamin B-12 value in the severely impaired subjects occurs as a consequence of receiv- ing the social support that is needed by patients with this degree of impairment to survive outside of @ nurs- ing home, or reflects a metabolic block associated with the appearance of “irreversible neurological damage” is unclear and remains a problem for future study. ‘The authors concur with Karnaze and Carmel” that low-normal serum vitamin B-12 is a real but atypical deficiency that requires more attention than it generally receives.”® These data also support a sim- ilar relationship with respect to folate.’ Normal val- 25Nutrition In Alzheimer’s Patients ues for serum vitamin B-12 (pg/ml) and RBC folate (ng/ml) range widely and may not be the best indica- tors of metabolic insufficiency."™° Elevations of methylmalonic acid and homocysteine in plasma and urine precede the appearance of deficient levels of vi- tamin B-12 and folate. Measurements of these metab- olites could be used to identify patients in the early stages of vitamin B-12 deficiency. Treatment can then be initiated to possibly prevent the irreversible nourologic damage caused by vitamin B-12 defi ciency." Lindenbaum et al" recently indicated that vitamin B-12 supplementation benefited patients with low levels of serum vitamin B-12 (<200 pg/ml), even when anemia was not present. Overall, the self-reported, dietary intakes of folate ‘were low in the whole population. Herbert™ advocates an RDI offolate of 3 ug- kg-day”*, which is consider- ably lower than the current RDA of 400 g/day. Con- sumption of 98% and 88% of the folate RDI by CN and SDAT groups, respectively, supported mean val- ues for RBC folate within normal ranges for both groups and further defends Herbert's position. In the small study, three SDAT subjects and one CN subject had low RBC folate values (<200 ng/ml). All SDAT patients with low RBC folate values consumed diets below the RDI, but the GN subject had an intake of folate in excess of 112% of the RDA. Multivitamin supplementation markedly improved RBC folate val ues of users with SDAT over nonusers. Whether higher serum levels of RBC folate benefits patients with SDAT remains to be studied. ‘The significantly lower body weight in SDAT pa- tients refutes the position taken by numerous workers in the field who have concluded that: because kiloca- loric intake meets a standard established for healthy elderly individuals, that energy needs are being met. If that were true, body weight would be maintained and not lost. Loss of body weight, coupled with ele- vated ETKAS and low RBC folate and serum vitamin B-12 values, strengthen the authors’ speculation that the metabolisin of mildly and moderately impaired SDAT patients is altered, and that this alteration can be partially compensated for by multivitamin supple- ments. However, adequate energy intake is critical for sustaining life and should be the primary focus for treatment of individuals. In conclusion, the authors’ findings confirm pre- viously reported lower levels of serum vitamin B-12, RBC folate, and body weight found in SDAT subject compared to CN elderly subjects. The study also found lowered ETKAPH in SDATT patients. In addi- tion, when subjects are grouped by MMSE scores, sig- nificantly higher serum vitamin B-12 and RBC folate levels occur in elderly individuals who score higher (26-80 points) on the MMSE compared to subjects scoring in the middle ranges (11-25 points); however, no difference occurred when these were compared to 26 severely impaired SDAT subjects (scores < 11 points). Finally, use of multivitamin supplements in- creased biochemical values for all nutrients exam- ined. Further clarification is needed to determine the physiologic benefits of such supplementation, espe- cially of vitamin B-12, in patients with SDAT. These data suggest that SDAT patients exhibit a form of atypical cobalamin deficiency, which in other pa- tients, has responded to vitamin B-12 therapy. References 1 Phillips P: Protein turnover in th elderly: A comparison be- ‘svoenil patients and normal controls, Human Nutrition: Clin foal Nutrition 37:839-344, 1983, 2, Morgan DB, Newton HMV, Schorah CJ, Jowitt MA, Hancock. (MR, Hullin RP; Abnormal indices of nutrition in the elderly ‘study of different clinical groups. Age Ageing 15:65-76, 1986 43, Sondmen PO, Adolfazon R, Nygron C, Hallmans G, Winblad B: Nutritional status and dietary intake in instittionalized patients with Alzheimer's disease and multiinfarct dementia Am Geriatr Soe 24:31-28, 1081, 4. Singh S, Mulley GP, LocoWksy MR: Why are Alzheimer's pa- tients thin? Age Ageing 17:21—28, 1988 5. Litchford MD, Wakefield LM: Nutrient intakes and nergy ex- pendituresofresidente with senile dementia ofthe Alzheimer's ype. Am Diet Asse 87:211-213, 1987 6, Stahlin HB: Senile emer fon to nutritional factors Bib Nutr Dieta 38:136-148, 1986, 11, Goodiwin J8, Goodwin JM, Garry PJ: Association between nu- {ritonal status and cognitive functioning ina healthy eldecly Doptlation. JADA 248:2917-2921, 1983, ‘8 Cole MC, Prehal JP: Low serum vitamin B-12 in Alzheimer type dementia, Age Ageing 13:101-105, 1984, 8, Renvoize BB, Gaskell RK, Klar HM: Results of investigations patients consecutively admitted toa peyehia Br J Psychiatr 147-204-208, 1985. ‘Diagnostic and Statitical ‘Manual of Mental Disorders ed.3, Washington, DC, American Prychintrie Association, 1960, 1, Folstein MF, Folstain SE, McHugh PR: Mini-mental tate: A practical method for grading the cognitive state of patients for the clinician. J Payehiatr Res 12:189-108, 1975, 12, Anthony J, LaResche L, Niaz U, Von Korf M, Folstein M: Limite of the "Mini-Mental State” ae a screen mentia and delirium among hospital patients. Psychiatric ‘Medicine 12:397-408, 1982, 13, Blessed GB, Toralingon Ey Roth Mi The ass quantitative measures of dementia and of seni rey matter of elderly subjects, Br J Prychiatr 114797-811, ‘oea. 14, Lau KS, Gotlie C, Wasserman LR, Herbert V: Measurement of serum vitamin B-12 level using radioiotope dilation and coated charcoal Blood 26:902-214, 1985, 15, Warnock LG, Frattll V, Preston AMC Transketolase activity of blood hemolyaate, a useful index for diagnosing thiamine Seficiency. Clin Chem 21:432430, 1975. 16. Graudal N, Torp-Pedersen K, Hane! H, Kristensen M, ‘Thom. sen AC, Norgard G: Assessment of the thiamine nutritional status. int J Vitom Nutr Rew $5:209-403, 1986, 137, Burch HB, Bessey OA, Lowry OH: Fluorometric measure- avin and ie natural derivatives in mall quanti- tes of blood serum an eel 38, Dixon Wd, Brown MB, Software Manual. Betkeley, CA, Unive ress, 1985, 1, Blass JP, Zomeoy A: Alzhelmer's disease: A metabolic aystems ‘dageneration? Neurochemical Pathology 2109-118, 1984, uly 1989 Volume 2 Number20. a. 2 2a, 2. 2%, 28. Pekkarinen Mt: Methodology in collection of food consumption data, World Rev Nutr Diet 12:145-171, 1970. Asplund K, Normark K, Petersson V: Nutritional assessment lofpsychogeriatric patients. Age Ageing 1087-96, 1981 Kjosen B, Seim SH. The transketolase assay of thiamine in some diseases. Am J Clin Nutr 30:1591-1596, 1977. ‘Horbert V: Recommended dietary intekes (RDI) of vitamin Be 12 in humans. Am J Clin Nutr 45:671-678, 1987 Carmel R, Sinow RM, Karnaze DS: Atypical cobalamin de- sleney.J Lab Clin Med 108454463, 1987. sous DL, Shadick N, Crantz J, Gray M, Hernandez F, Freedman NL: Low serum B-12 level in a hematologically normal elderly subpopulation. J Am Geriatr Soe 35:696-698, 1987, Herbert V: Recommended dietary intakes (RDI) of folate in humans. An J Clin Nut? 45:661-670, 1987, {HE AMERICAN JOURNAL OFTHE MEDICAL SCIENCES m1 28, 29, 20, a Renvall et al Karneze DS, Carmel R: Low serum eobalamin levels in pi rary degenerative dementia. Arch Intern Med 147-429-431, {o8r. arse! R, Karngze DS: Physician response tolow serum cobe- lamin levels. Arch Intern Med 146:1161-1165, 1986, ‘Norman EJ, Martolo 03, Denton MD: Cobalamin (vitamin B- 12) deficiency detection by urinary methylmalonte acid quanti- tation. Blood 69:1128-1181, 1982, Stable: SF, Marcell PD, Bodell ER, Allen RH, Savage DG, Lindenbaum J: Blevation of total hotioeysteine in the serum ‘ofpatients with eobalamin or folate deficiency detected by eap- lary gus chromatograpihy-mass spectrometry. Clin Invest ‘31s466-474, 1088. Lindenbaum J, Healton EB, Savage DG, Brust JC, Garrett 3, Podell ER, Marcell PD, Stabler SP, Allen RH: Nosropsyehiat ric disorders caused by cobalamin deficiency in the absence of| ‘anemia or macrocytosis. N Engl J Med 18:1720-1728, 1988, 27