JOURNAL OF THE AMERICAN GERIATRICS SOCIETY Vol. XXVIII, No.

1
Copyright © 1980 by the American Geriatrics Society Printed in U.S.A.

Oral versus Intramuscular Vitamin

Supplementation for Hypovitaminosis in the
Elderfy
HERMAN BAKER, PhD*, OSCAR FRANK, PhD and SEYMOUR P. JASLOW, MD
Departments of Preventive Medicine and Community Health and Medicine, New Jersey Medical
School, East Orange; and Daughters of Miriam Center for the Aged, Clifton, New Jersey
ABSTRACT: Thiamin, folate, biotin, riboflavin, nicotinates, pantothenate,
carotenes, and vitamins B o, B 12 , A, E and C were measured in the blood of 228
elderly ambulatory residents of a nursing home. Their mean age was 87 years
(range, 60-102). None had undergone major surgical procedures; their diet was
good, and each had received at least one multivitamin pill every day for 3 to 5
months before the study. A comparison group of 204 healthy volunteers, aged
20-50, was also studied. Of the 228 elderly subjects, 88 (39 percent) showed
vitamin deficits despite oral vitamin supplementation. Single and multiple def-
icits of vitamin B o, nicotinate, vitamin B 12 , folate, and thiamin were found. Three
months after a single intramuscular injection of multivitamins (with no other
vitamin supplementation), these deficits were no longer detectable in the blood
of 89-100 percent of the vitamin-deficient elderly. Intramuscular rather than
oral vitamin supplementation is a more effective method for maintaining ade-
quate blood levels ofvitamins in the elderly; the intramuscular route apparently
promotes saturation of tissue stores with enough vitamins to meet the needs,
and thus obviates problems of vitamin malabsorption possibly due to drug
interference or small-bowel atrophy.

Elderly persons over 60 years of age often have
vitamin deficiencies caused by malabsorption (1,
2). Since there is little information concerning the
proper regimen for maintaining an adequate level
of circulating vitamins in the aged, we examined
the blood patterns of 12 vitamins in 228 elderly
residents of a nursing home and compared these
patterns with those in younger subjects aged 20-
50. Latent vitamin deficits were found in 88 (39
percent) of these elderly subjects despite ingestion
of multivitamin preparations. The administration
of multivitamins intramuscularly corrected these
vitamin deficits for at least three months even
though other vitamin supplementation was dis-
continued during this time.
MATERIALS AND METHODS
The elderly population of this study consisted
• Address for correspondence: Herman Baker, PhD, De-
partment of Preventive Medicine and Community Health, New
Jersey Medical School, 88 Ross Street, East Orange, NJ 07018.
42
of 228 persons living in a center for the aged in
New Jersey and provided with good nursing, med-
ical and dietary supervision. Their ages ranged
from 60 to 102 years (mean, 87). To qualify for the
initial vitamin screening, subjects had to be am-
bulatory with no history of gastrointestinal disease
or major surgical operations, and a history of daily
ingestion of a multivitamin pill. We discontinued
recording dietary histories in detail because the
subjects often gave unreliable accounts. The his-
tory of supplemental vitamin intake was obtained
from the medical charts and from the nursing
staff.
All 228 subjects ingested various vitamin sup-
plements in a dosage of 1 tablet daily for 3-5
months. Each tablet contained, on the average:
vitamin A palmitate (5,000 IU), vitamin E (30 IU),
vitamin C (160 mg), folic acid (0.2 mg), thiamin
HCl (1.5 mg), riboflavin (1.7 mg), nicotinate (20
mg), pyridoxine HCI (2 mg), and vitamin BIZ (5
ug), Vitamin supplementation was discontinued
for 48 hours before drawing blood for vitamin
analyses.
 January 1980 ORAL VS. INTRAMUSCULAR VITAMIN THERAPY

The comparison group consisted of 204 healthy When comparing values in the elderly to those in
volunteers whose ages ranged from 20 to 50 years the younger population (Table 1), we noted nei-
(mean, 34). They were selected from laboratory ther deficits of vitamin A [range 25-94 (mean 55
personnel and volunteer groups; none had a his- ± 16.6) ug/dl], carotene [range 44-273 (mean 147
tory of liver or gastrointestinal disease, or of hav- ± 70) ,ug/dl], vitamin E [range 0.5-1.7 (mean 0.8
ing undergone surgical procedures. All denied tak- ± 0.3) mg/dl], vitamin C [range 0.4-2.5 (mean 1.2
ing vitamin supplementation. ± 0.5) mg/dl], pantothenate [range 251-756 (mean
After an overnight fast, blood was drawn from 423 ± 115) ng/ml], riboflavin [range 164-539
an anticubital vein into Vacutainers (Becton, (mean 261 ± 80) ng/ml], nor of biotin [range 219-
Dickinson & Co.) containing 14 mg of disodium 625 (mean 410 ± 90) pg/rnl] in the 228 elderly
(ethylenedinitrilo)tetra-acetic acid powder. Blood subjects after the initial vitamin screening.
was analyzed with protozoan reagents for thiamin, Of the elderly group, 140 (61 percent) had no
biotin, nicotinates, pantothenate, vitamin B6 , ribo- deficits of vitamin B6 [range 30-83 (mean 40 ± 12)
flavin, vitamin B I2 (3) and with Lactobacillus casei ng/rnl], nicotinate [range 3.1-8.1 (mean 4.0 ± 0.9)
for folates (3); vitamins A, E, C (ascorbate) and ug/ml], vitamin B I2 [range 119-656 (mean 291 ±
carotenes were determined chemically (3). These 157) pg/ml], folate [range 6.9-53 (mean 14.1 ± 9.8)
methods have been used in various nutritional ng/ml] and thiamin [range 25-83 (mean 58 ± 11)
surveys (1, 4,...7). ng/ml]; these values were all within the range
Vitamin deficit (hypovitaminemia) was defined noted for the younger comparison population (Ta-
as a blood concentration less than the 95 percent ble 1). However 88 (39 percent) of the 228 elderly
confidence limit established on the data for the had single or multiple blood vitamin deficits in-
204 subjects in the comparison population (Table volving vitamin B6 , nicotinate, vitamin B 12 , folate,
1); this is a workable definition since subjects and thiamin (Table 2). Vitamin B6 hypovitami-
classified as hypovitaminemic often showed clini- nemia was the most common; it existed alone or
cal signs characteristic of vitamin deficits (5, 7, 8). in combination in 40 percent of the vitamin-defi-
As in other surveys (1, 4, 9), vitamin values for cient elderly (Table 2). As seen in Table 3, the
both populations were listed in ascending order range and mean of the vitamin deficits were sig-
and the frequency of occurrence in each popula- nificantly lower than those for the comparison
tion plotted. The mean and 95 percent confidence population or for the foregoing 140 elderly who
limits derive from cumulative frequency points on had no laboratory evidence of vitamin deficits;
semi-log coordinates. A straight line through the this group essentially constituted the elderly con-
intersect established the mean at the 50 percent trols since their vitamin titers were not signifi-
intersect, and 95 percent confidence limits at the cantly different from those of the younger com-
2.5 percent and 97.5 percent intersects. Student's parison population (Table 1).
t test served to estimate the significance of differ- A single intramuscular injection of vitamin sup-
ences (p) between means; results were expressed plement corrected the vitamin deficits listed in
as means ± S.D. (10). Table 2. Vitamin analyses, three months after the
After blood vitamin concentrations were deter- intramuscular dose, indicated that an adequate
mined, elderly subjects with evidence of hypovi-
taminemia were given an intramuscular dose of TABLE 1
multivitamins. The total dose contained: 10 mg Circulating Vitamins in 204 Healthy Volunteers (Aged 20-
50) Without Evidence of Malnutrition and Not Receiving
thiamine HCI, 10 mg riboflavin, 80 mg nicotin- Vitamin Supplementation
amide, 70 mg pyridoxine HCI, 23.2 mg calcium
Vitamin Mean Value Range"
pantothenate, 0.2 mg D-biotin, 100 mg ascorbic
acid, 5 mg folic acid, and 1.0 mg vitamin B 12 • All Vitamin B I2 pg/ml 210 105-660
Biotin pg/rnl 485 215-750
. other vitamin supplementation was discontinued Thiamin ng/rnl 40 25-71
for three months; after this period, blood vitamin Riboflavin ng/rnl 205 110-420
levels were re-estimated. Nicotinates /Lg/ml 4.0 3.4-6.4
Vitamin B6 ng/rnl 38 29-83
Folates ng/rnl 8.0 5.0-21.0
RESULTS Pantothenate ng/ml 340 205-780
Vitamin A /Lg/dl 45 25-84
Vitamin C mg/dl 0.9 0.4-1.8
Preliminary tabulations of the younger and el- Vitamin E mg/dl 0.8 0.5-1.6
derly groups showed that sex and race could be ,a-Carotene /Lg/dl 95 40-200
ruled out as factors influencing vitamin values (1). • 95% confidence limits.

43
 BAKER, FRANK AND JASLOW Vol. XXVIII

circulating vitamin pattern was maintained in 87- deficits in a high proportion (1). As shown in Table
100 percent of the deficient elderly despite discon- 2, a vitamin survey of the 228 institutionalized
tinuation of vitamin supplements (Table 3). Daily elderly in this study indicated that single and
oral vitamin supplementation failed to maintain multiple deficits of vitamins B 6 and B 12 , folate,
the levels of vitamins B6 and B 12 , folate, nicotinate, thiamin, and nicotinate existed in 39 percent of
and thiamin (Tables 2 and 3) according to the the subjects despite an ostensibly excellent diet
pattern in the younger comparison population and daily ingestion of a tablet containing more
(Table 1) or according to that in 140 of the 228 than the recommended daily allowances of vita-
elderly subjects. mins (13). It is possible that vitamin tablets are
not sufficiently broken down for absorption by the
DISCUSSION gut of the elderly patient and consequently are
evacuated unchanged.
Low vitamin titers have been noted in institu- Another possible cause of vitamin malabsorp-
tionalized elderly patients (1, 11). Sixty percent of tion is the large number of prescription drugs used
sheltered-dwelling residents showed subclinical by the elderly-the principal drug users in our
thiamin deficits and 20 percent had B6 deficits society (14). Vitamin antagonism is an undesirable
despite adequate diets and multivitamin therapy side effect of many medicines, e.g., hydralazine, L-
(11, 12). In our previous study of 473 elderly per- dopa, and penicillamine; this may contribute to B 6
sons, vitamin B 6 , nicotinate, vitamin B 12 , folate, deficiencies (15). It has been suggested that these
thiamin, and ascorbate emerged as commonest drug-induced vitamin deficits could be avoided by
use of high-potency vitamin supplements. How-
TABLE 2 ever high-potency vitamins may interfere with the
Frequency of Vitamin Deficits in 88 of 228 Elderly Nursing therapeutic effectiveness of ingested drugs. For
Home Subjects Ingesting Multivitamins
example, high intakes of folic acid depress the
% [and Number] effects of anticonvulsants such as diphenylhydan-
of the Elderly
toin and phenobarbital, and high doses of pyridox-
Single Vitamin Deficit ine can even overcome the L-dopa control of par-
B6 40 [31]
Nicotinate 14 [12] kinsonism if given concomitantly (15).
B 12 9 [8] To avoid vitamin malabsorption and drug an-
Folic acid 7 [6] tagonisms from any cause, we decided to admin-
Thiamine 1 [1]
ister a single intramuscular dose of vitamins; this
Multiple Vitamin Deficit bolus contained larger amounts of vitamins than
Folic acid + B'2 7 [6] do the oral vitamin preparations and supposedly
B 6 + Nicotinate 7 [6]
B 6 + B 12 6 [5] could saturate vitamin-binding sites in the tissues.
B 6 + Nicotinate + B l 2 3 [3] Saturated binding sites would then permit vita-
B6 + Folic acid 2 [2] mins to leak into the circulation and make vita-
B 6 + Thiamine 1(1]
Thiamine + B 12 1 [1] mins available as needed co-enzymes (3). Depleted
B 6 + Nicotinate + Thiamine 1 (1] vitamin stores would lead to lessened vitamin
Nicotinate + Thiamine 1 [1] leakage from tissues; this, in tum, would produce

TABLE 3
Comparison of Oral and Intramuscular Vitamin Supplementation in an Elderly Population
After 3-5 Months of Oral Vitamin Percent-
After 3 Months of Intramuscular
Supplementation (before Intra- Corrected
Vitamin Treatment
muscular Treatment) to Vitamin
Levels in
Reference
Range of Range of Population
Vitamin No. of Values (Mean ± S.D.)' No. of Values (Mean ± S.D.)t (Table 1)§
B 6 (ng) 51 13-25 (20 ± 3) 51 20-80 (47 ± 14) 90
Nicotinate (p.g) 23 1.9-3.0 (2.4 ± 0.2) 23 3.1-4.5 (3.5 ± 0.6) 87
B l 2 (pg) 23 9-111 (74 ± 28) 23 56-825 (232 ± 155) 91
Folic acid (ng) 14 1.6-4.0 (3.0 ± 1.0) 14 5.6-28 (13.3 ± 6.1) 100
Thiamine (ng) 4 15-21 (15 ± 5) 4 55-103 (71 ± 26) 100
, All values significantly depressed as compared to those in the reference population (Table 1); p < .001.
t Values significantly increased as compared to values before intramuscular treatment, p < .001.
§ See "Results" for 140 elderly control subjects with no vitamin deficits.

44
January 1980 ORAL VS. INTRAMUSCULAR VITAMIN THERAPY
low blood vitamin levels, i.e., hypovitaminemia. A
similar situation occurs with serum-ferritin, e.g.,
lack of storage iron means less ferritin leakage
from tissues and hence low levels of serum ferritin
(16).
Administration of oral vitamin preparations for
3-5 months did not permit maintenance of ade-
quate circulating vitamins compared to the
younger comparison population; thus hypovita-
minemias developed (Tables 2, 3). The fact that a
single intramuscular dose of multivitamins re-
stored adequate blood vitamin levels which were
maintained for at least three months (Table 3)
could also mean lessened C08ts for vitamins and
for the time consumed by staff. Apparently, intra-
muscularly administered vitamins saturate the tis-
sue binding sites with vitamins and thus obviate
vitamin malabsorption in the elderly. However,
vitamin saturation of tissues occurs only if tissue
binding sites are adequate. Binding affinities for
drugs do not seem to change with age. Recent
reports indicate that the number, but not the
affinity, of beta-adrenergic receptors decreases in
several areas of the brain in aging rats; younger
animals can increase receptor density but older
animals cannot (17, 18).
This may also be true for vitamins; the number
of vitamin binding sites in the liver-the main
organ for vitamin storage (3, 8)-may decrease
during aging, with production of an inadequate
number of binding sites for vitamin storage. Thus
vitamin saturation of the lessened number of bind-
ing sites is required to insure adequate vitamin
needs.
Our results provide a plausible answer to the
question raised earlier by us (1) and others (19) as
to why hypovitaminemia and associated clinical
symptoms may exist despite oral vitamin supple-
mentation; the absorption of small amounts of
vitamins may not suffice to saturate depleted tis-
sues (20), and thus vitamins are rapidly depleted
after discontinuation of the oral supplement (1,
21). Oral vitamin supplementation may not suffice
to obviate an existing hypovitaminemia in the
elderly (Table 3).
Intramuscular administration of vitamins may
therefore be the method of choice for maintaining
adequate vitamin levels in the elderly and for
correction of an existing hypovitaminemia. Intra-
muscular administration assures release of vita-
mins into the circulation from vitamin-saturated
tissues, even though all vitamin intake is diminish-
ing.
REFERENCES
1. Baker H, Frank 0, Thind IS et al: Vitamin profiles in
elderly people living at home or in nursing homes, versus
profile in healthy young subjects, J Am Geriatrics Soc 27:
444,1979.
2. Krehl WA: The influence of nutritional environment on
aging, Geriatrics 29: 65, 1974.
3. Baker H and Frank 0: Clinical Vitaminology: Methods and
Interpretation. New York, Interscience, 1968.
4. Baker H, Frank 0, Feingold S et al: Vitamins, total choles-
terol, and triglycerides in 642 New York City school chil-
dren, Am J Clin Nutr 20: 850, 1967.
5. Leevy CM, Cardi L, Frank 0 et al: Incidence and signifi-
cance of hypovitaminemia in a randomly selected munici-
pal hospital population, Am J Clin Nutr 17: 259,1965.
6. Baker H, Frank 0, Thomson AD et al: Vitamin profile of
174 mothers and newborns at parturition, Am J Clin Nutr
17: 259, 1965.
7. Dastur DK, Santhadevi N, Quadros EV et al: The B-
vitamins malnutrition with alcoholism, Brit J Nutr 36: 143,
1976.
8. Baker H and Frank 0: Vitamin status in metabolic upsets,
World Rev Nutr 9: 124, 1968.
9. Ziffer H, Frank 0, Cbristakis G et al: Data analysis for
nutritional survey of 642 New York City school children,
Am J Clin Nutr 20: 858, 1967.
10. Bancroft H: Introduction to Biostatistics. New York, Hoe-
ber-Harper, 1962, pp 78-79.
11. Vir SC and Love AHG: Nutritional evaluation ofB groups
of vitamins in institutionalized aged, Internat J Vit Nutr
Res 47: 211, 1977.
12. Vir SC and Love AHG: Vitamin B6 status of institutional-
ized and non-institutionalized aged, Internat J Vit Nutr
Res 47: 364, 1977.
13. Food and Nutrition Board, National Academy of Sciences:
National Research Council Recommended Daily Dietary
Allowances. Washington, DC, 1974.
14. US Task Force on Prescription Drugs: The Drug Users.
Washington, DC, US Dept of HEW, 1968, pp 126-129.
15. Roe DA: Drug-Induced Nutritional Deficiencies. Westport,
Conn., Avi Publishing Co, Inc, 1976, pp 154-186.
16. Editorial: Serum ferritin, Lancet 1: 533, 1979.
17. Weiss B, Greenberg L and Cantor E: Age-related altera-
tions in the development of adrenergic denervation super-
sensitivity, Fed Proc 38: 1915, 1979.
18. Cafruny EJ: Pharmacology and aging, Fed Proc 38: 1899,
1979.
19. Taylor GF: A clinical survey of elderly people from a
nutritional standpoint, in Vitamins in the Elderly, ed. by
AN Exton-Smith and DL Scott. Bristol, John Wright and
Sons Ltd, p 51.
20. Zamurowski MM and Wagner C: Effect of folacin defi-
ciency on folacin-binding proteins in the rat, J Nutr 107:
1037, 1977.
21. Barrows CH Jr and Roeder LM: Nutrition, in Handbook
of Biological Aging, ed. by CE Finch and L Hayflick.
Princeton, NJ, Van Nostrand, 1977, pp 561-581.
45