Escolar Documentos
Profissional Documentos
Cultura Documentos
Current Research
Even so, less than 50% of both users and nonusers met
ABSTRACT the EAR for folate, vitamin E, and magnesium from food
Objective To measure nutrient intake adequacy of vitamin/ sources alone. Overall, supplements improved the nutri-
mineral supplement users and nonusers aged 51 years ent intake of older adults. After accounting for the con-
and older, determine the efficacy of supplement practices tribution of supplements, 80% or more of users met the
in compensating for dietary deficits, and identify predic- EAR for vitamins A, B-6, B-12, C, and E; folate; iron; and
tors of supplement use.
zinc, but not magnesium. However, some supplement
Design Analyses of two 24-hour recalls, demographic vari-
users, particularly men, exceeded Tolerable Upper Intake
ables, and attitude questions collected during the Con-
Levels for iron and zinc and a small percentage of women
tinuing Survey of Food Intakes by Individuals and Diet
exceeded the Tolerable Upper Intake Level for vitamin A.
and Health Knowledge Survey in 1994 to 1996. Data were
Significant sociodemographic factors related to supple-
weighted to be representative of older Americans.
ment use for older men were age group, metropolitan
Subjects Four thousand three hundred eighty-four adults
aged 51 years and older (1,777 daily supplement users, area, and educational status. Race, region, smoking sta-
428 infrequent users, and 2,179 nonusers) residing in tus, and vegetarian status were significant factors for
households in the United States. women. Attitude about the importance of following a
Statistical analyses Usual nutrient intake distributions healthful diet was a consistent predictor of supplement
were estimated using the Iowa State University method. use for both men and women.
The Estimated Average Requirement (EAR) cutpoint Conclusions A large proportion of older adults do not con-
method was applied to determine the proportion of older sume sufficient amounts of many nutrients from foods
adults not meeting requirements before and after ac- alone. Supplements compensate to some extent, but only
counting for nutrient intake from supplements. Student t an estimated half of this population uses them daily.
tests were used to assess differences between users and These widespread inadequacies should be considered
nonusers. Logistic regression was used to determine so- when developing recommendations for supplement use
ciodemographic and attitudinal predictors of supplement for clients in this age group. Modifying dietary attitudes
use. may result in a higher rate of supplement use in this
Results For one or more of the sex-age groups studied, a at-risk population.
significantly smaller proportion of supplement users than J Am Diet Assoc. 2007;107:1322-1332.
nonusers had intakes from food alone below the EAR for
vitamins A, B-6, and C; folate; zinc; and magnesium.
A
nalyses of nationwide survey data have shown that
a large percentage of older adults do not receive
R. S. Sebastian is a nutritionist, J. D. Goldman is a recommended amounts of many nutrients from food
statistician, and A. J. Moshfegh is research leader, US alone (1-11). Other measures, such as the Healthy Eating
Department of Agriculture, Agricultural Research Ser- Index, also indicate that the diets of older adults need
vice, Food Surveys Research Group, Beltsville, MD. improvement and may leave them susceptible to nutri-
L. E. Cleveland is retired; at the time of the study, she tion-related problems (12,13). This is of particular con-
was a supervisory nutritionist, US Department of Agri- cern because conditions prevalent in this population, in-
culture, Agricultural Research Service, Food Surveys cluding chronic diseases and absorption problems, can
Research Group, Beltsville, MD. compromise nutritional status (14-16). At the same time,
Address correspondence to: Rhonda S. Sebastian, MA, a growing proportion of older adults are using vitamin
Nutritionist, US Department of Agriculture, Agricul- and mineral supplements, which can substantially in-
tural Research Service, Food Surveys Research Group, crease nutrient intake and counter some of these short-
10300 Baltimore Ave, Bldg 005, Room 102, BARC-West, falls (17-19). Little is known about the effectiveness of
Beltsville, MD 20705. E-mail: Rhonda.Sebastian@ars. vitamin/mineral supplements in appropriately compen-
usda.gov sating for dietary deficits.
Published by Elsevier Company on behalf of the With the introduction of the Dietary Reference Intakes
American Dietetic Association. (DRIs), new standards are available for the assessment of
0002-8223/07/10708-0002$0.00/0 nutrient intakes. The establishment of Estimated Average
doi: 10.1016/j.jada.2007.05.010 Requirements (EARs) makes it possible, for the first time, to
Identification of Factors Related to Supplement Use Comparison of Nutrient Adequacy from Food between Users and
Sociodemographic variables for this analysis included Nonusers
variables shown in previous studies to be related to sup- Tables 3 and 4 present the mean usual daily nutrient
plement use, including sex, age, race, poverty status, and intake and intake at the 25th, 50th, and 75th percentiles
education level (17,27-31,33-40,42,44). In addition, two for supplement users and nonusers. For supplement us-
attitude scales, composed of 11 questions each, were de- ers, estimates are presented for intake from food sources
rived from Diet and Health Knowledge Survey data. The only and for the combined intake from both food and
first scale was developed from questions on perceived supplements. Also shown are percentages of the popula-
Table 2. Frequency of use of selected supplements by adults aged 51 years and older, based on data from the Continuing Survey of Food
Intakes by Individuals, 1994-1996
Men Women
51-70 y (%) >71 y (%) 51-70 y (%) >71 y (%)
Supplement Supplement Supplement Supplement
Supplement All users All users All users All users
Multivitamin/mineral 34 75 33 72 43 70 37 70
Vitamin C 15 33 14 30 21 34 15 28
Vitamin E 13 29 13 28 16 27 10 19
Calcium 4 8 5 11 16 26 9 18
Vitamin B/B complex 6 12 7 16 9 14 6 11
Zinc 3 7 3 6 3 5 2 3
Iron 2 4 2 4 3 5 2 4
Vitamin A 2 5 2 5 4 6 2 4
tion with intakes below the EAR (ie, percentages with This was true for men and women in both age groups,
inadequate intakes). Data are shown separately for men regardless of supplement use.
(Table 3) and women (Table 4).
When considering nutrient intake from food sources
alone, generally smaller proportions of supplement users Contribution of Supplements to Nutrient Intake
than nonusers had inadequate intakes. Intakes of vita- Regular supplement use, in effect, reduced the percent-
mins A and C were significantly higher for users than age of older adults with inadequate intakes by at least
nonusers in both age groups of men. Vitamin B-6 and three fourths for most nutrients (Table 3 and 4). Folate
magnesium were significantly higher for users as com- and vitamin E intakes were particularly improved by
pared to nonusers in both age groups of women. More supplements. Food-only intakes of these nutrients were
than 50% of both supplement users and nonusers had
inadequate for 57% to 96% of supplement users. After
inadequate intakes of folate, vitamin E, and magnesium
accounting for supplements, only 6% to 17% were classi-
from food. In addition, for most of the sex/age groups
studied, more than 25% had inadequate intakes of vita- fied as inadequate.
mins A, B-6, and C from food. Iron was the only nutrient When the nutrient contribution of supplements was
studied for which the prevalence of inadequate intakes considered, significant differences were found between
among older adults was consistently low; 3% or fewer users and nonusers in percentages with inadequate in-
older adults had intakes below the EAR. takes for nearly all the nutrients studied. Mean intakes of
It was not possible to determine proportions of the vitamin B-12 and iron were boosted by supplements, but
population with inadequate calcium intakes because the the percent of this population with inadequate intakes
DRI is expressed as an Adequate Intake and not as an was largely unaffected, because most older adults met the
EAR. Although mean intakes from food alone were higher EAR from food intake alone. An exception was that sig-
among supplement users than nonusers (Table 5), the nificantly more nonuser women aged 71 years and older
only statistical difference between users and nonusers had inadequate intakes of vitamin B-12 compared to the
was among men aged 51 to 70 years. total intake of supplement users in this sex/age group.
The nutrient intake from food alone was not sufficient Although for the most part the percent meeting the EAR
to exceed the ULs for any of the nutrients in this analysis. was not significantly increased for vitamin B-12 and iron,
supplements did affect intake. Mean intakes from food Supplements and Overconsumption of Nutrients
and supplements for users were significantly different
from total intake of nonusers. Supplements more than Fewer than 3% of men in both age groups had total
doubled vitamin B-12 intake for men and women, and intakes that exceeded the UL for vitamins B-6, and C,
this may be of benefit because the absorption of naturally and calcium, indicating that risk of adverse effects from
occurring vitamin B-12 may be problematic in older excesses of these nutrients is low in this population group
adults (55). (Table 6). Supplement use did, however, result in intakes
Older men and women who consumed supplements had that exceeded the ULs for iron and zinc for 10% or more
significantly higher total calcium intakes than nonusers of older men.
of these products (Table 5). The proportion reaching or Women in both age groups exceeded the UL for iron
exceeding the Adequate Intake increased from 15% to and zinc, but they did so in smaller proportions than the
29% for men aged 51 to 70 years. Large improvements men. Unique to the women was that a higher proportion
were also found for older women and men aged 71 years (5% to 9%) exceeded the UL for vitamin A than the
and older. proportion of men (4%) who did so.
Demographic and Attitudinal Predictors of Supplement Use nificant predictors of supplement use. White women
Predictors of supplement use were different for older were more likely to take supplements than all other
men than for older women (Table 7). Men aged 71 years racial groups combined, and were twice as likely to
and older and those living in metropolitan areas were take supplements as Hispanic women. Residence in the
significantly more likely to consume supplements com- western region of the United States indicated greater
pared to men in the 51- to 70-year age group and those prevalence of supplement use than in all other regions
in nonmetropolitan areas, respectively. Education sta- together; those in the northeast and midwest were
tus was also a predictor of supplement use for men. about half as likely to be users. Women who were
Contrasts showed that significant differences were nonsmokers were more likely to consume supplements
found between the least educated men (less than high than women who smoked, and the small percentage of
school completion attained) and those with either a older women who were vegetarians (4%) were twice as
high school education or formal education beyond high likely to be supplement users as their nonvegetarian
school. No differences were noted between the two counterparts.
higher education categories. For the women, race, re- Of the two attitude variables included in the logistic
gion, smoking status, and vegetarian status were sig- regression model, the personal importance of consuming
Males 51-70 y
Nonusers: Food only 726⫾20cd 681 513, 889 7⫾1.2
Users: Food only 871⫾24c 827 638, 1,056 15⫾2.1
Food and supplement 1,038⫾25d 989 768, 1,256 29⫾2.5
Males >71 y
Nonusers: Food only 677⫾24d 636 475, 834 5⫾1.1
Users: Food only 810⫾34 773 598, 983 10⫾2.1
Food and supplement 981⫾32d 941 731, 1,187 24⫾3.0
Females 51-70 y
Nonusers: Food only 573⫾16d 544 416, 699 ⬍3
Users: Food only 630⫾16 599 459, 767 ⬍3
Food and supplement 892⫾19d 849 629, 1,109 19⫾1.4
Females >71 y
Nonusers: Food only 540⫾21d 513 388, 662 ⬍3
Users: Food only 645⫾29 611 473, 779 ⬍3
Food and supplement 856⫾36d 808 617, 1,042 14⫾3.1
a
SE⫽standard error.
b
The Adequate Intake for men and women aged 51 y and older is 1,200 mg.
c
Significant difference between nonusers and users (food only) (P⬍0.001).
d
Significant difference between nonusers and users (food and supplements) (P⬍0.001).
Table 6. Percent of supplement users⫾standard error aged 51 years and older exceeding the Tolerable Upper Limit (UL), based on data from
the Continuing Survey of Food Intakes by Individuals, 1994-1996a
% Supplement Users Above the UL
Men Women
b c
Nutrient UL 51-70 y >71 y 51-70 y >71 y
Attitudinal variables
Perceived adequacy of dietb 0.74 (0.45-1.23) 0.95 (0.70-1.29)
Importance of healthy dietc 1.65 (1.25-2.17) *** 1.62 (1.10-2.38) *
Sociodemographic variablesd
Age group (51-70 y, ⱕ71 y) 0.68 (0.49-0.94) * 1.22 (0.88-1.69)
Race (Non-Hispanic whiteⴝ1.00) **
Hispanic 1.20 (0.49-2.91) 0.45 (0.24-0.82)
Non-Hispanic Black 0.65 (0.33-1.28) 0.63 (0.35-1.11)
Non-Hispanic other race 0.53 (0.14-2.04) 0.33 (0.09-1.18)
Region (Westⴝ1.00) **
Northeast 0.70 (0.40-1.22) 0.46 (0.30-0.69)
Midwest 0.93 (0.51-1.71) 0.59 (0.40-0.87)
South 0.96 (0.57-1.62) 0.66 (0.43-1.02)
Metropolitan area (yes/no) 1.65 (1.09-2.49) * 1.26 (0.94-1.69)
Poverty statuse (350% and over⫽1.00)
⬍130% 0.83 (0.52-1.32) 0.65 (0.40-1.08)
131%-350% 0.86 (0.60-1.24) 0.83 (0.58-1.18)
Educational status (education
beyond high schoolⴝ1.00) **
⬍High school completed 0.50 (0.33-0.76) 0.72 (0.46-1.15)
High school completed 0.76 (0.51-1.13) 0.75 (0.52-1.08)
Current smoker (yes/no) 0.92 (0.62-1.37) 0.67 (0.46-0.97) *
Weight status (overweight/not
overweight) 0.73 (0.52-1.03) 0.82 (0.60-1.13)
Exercise (ⱕ1 time/wk, ⱖ2
times/week) 0.96 (0.69-1.33) 0.78 (0.57-1.07)
Television watching (ⱕ1 h/d, 1⫹ h/d) 0.78 (0.53-1.16) 0.79 (0.58-1.06)
Follows special diet (yes/no) 1.36 (0.93-1.99) 1.28 (0.95-1.72)
Vegetarian (yes/no) 1.14 (0.44-2.92) 2.03 (1.18-3.49) **
a
CI⫽confidence interval.
b
11 item scale querying respondent if their diet was too high, too low, or about right in (nutrient).
c
11 item scale asking respondent if it was very important, somewhat important, not too important, or not at all important to follow (1990 Dietary Guidelines [60]).
d
For dichotomous variables, latter choice was base for comparison.
e
Poverty categories expressed as percent of the Federal poverty level.
*P⬍0.05.
**P⬍0.01.
***P⬍0.001.
supplements. To our knowledge, this is the first study to or more met EARs with their total intake for all nutrients
use DRIs in combination with estimates of usual total studied except magnesium.
nutrient intake to evaluate intakes in this population. The folate estimates here underestimate current in-
The results are consistent with those from previous takes because the US Food and Drug Administration did
studies (27,32,38,40,41,43-45) in showing that, generally, not mandate fortification of grain products until 1998,
supplement users consume more nutritious diets than after data for this study were collected; however, they
nonusers. We found that supplement users were more provide a benchmark for comparison to post-fortification
likely to have adequate nutrient intakes from food alone levels. Fortification was expected to add 70 to 130 g to
than nonusers for many of the nutrients studied. How- the daily diet, but subsequent studies have shown that
ever, both supplement users and nonusers had large intakes have been increased by 200 g or more
shortfalls in intake from food alone for several nutrients, (48,49,61,62). Based on our study, most older adults
most notably folate, vitamin E, and magnesium. Supple- would meet adequacy standards if diets provided a net
mentation was most beneficial in improving intakes for increase of 200 g/day. Nevertheless, recent analyses
these nutrients. Nutrients for which little benefit from using NHANES 2001-2002 data and the accompanying
supplementation was observed were vitamin B-12 (as- US Department of Agriculture’s Food and Nutrient Da-
suming adequate absorption) and iron because most older tabase for Dietary Studies 1.0, which accounts for current
adults met the EARs from foods alone. Among users, 80% fortification levels found that depending on the sex/age