Download The increasing prevalence of snacking among US children from

The increasing prevalence of snacking among
US children from 1977 to 1996
Lisa Jahns, RD, Anna Maria Siega-Riz, RD, PhD, and Barry M. Popkin, PhD
Objectives: To determine snacking trends and changes in nutrient contribution of snacking over time.
Study design: Nationally representative data from the 1977-78 Nationwide
Food Consumption Survey (NFCS77), 1989-91 Continuing Survey of Food
Intake by Individuals (CSFII89), and 1994-96 (CSFII96) were used. The
sample consisted of 21,236 individuals aged 2 to 18 years.
Methods: For each survey year, mean numbers of snacks consumed, mean
grams consumed per snack, and mean energy intake from snacks were computed, as was contribution of snacking to total energy intake and fat intake.
Snacking was self-defined, and a snacking occasion consisted of all snack
foods consumed during a 15-minute period. Differences in means between
age groups and across survey years were compared.
Results: The prevalence of snacking increased in all age groups. The average size of snacks and energy per snack remained relatively constant; however, the number of snacking occasions increased significantly, therefore increasing the average daily energy from snacks. Compared with non-snack
eating occasions, the nutrient contribution of snacks decreased in calcium
density and increased in energy density and proportion of energy from fat.
Conclusion: Snacking is extremely prevalent in our society. Healthy snack
food choices should be emphasized over high-energy density convenience
snacks for children. (J Pediatr 2001;138:493-8)
ty of higher energy density foods that
may be replacing nutrient-dense foods
as culprits in the obesity problem.4-5
Descriptive studies have shown that
snacking is currently an important
component of the diet of Americans as
measured by factors such as the snacking occasions per day and percent
of snacks eaten outside the home.6
Bigler-Doughton and Jenkins7 reported that 80% of adolescents consumed
snacks, and snacks made substantial
contributions to adolescents’ intakes of
several key nutrients. Cross et al8
found that <1% of children reported
never snacking.
We examined trends in snacking behaviors among children aged 2 to 18
years over the past 20 years to show
how behaviors have changed and the
effects of these changes on nutritional
intake of children.
CSFII
NFCS
The rapid increases over the past 3
decades in childhood obesity and related complications have led to increased
concern over the diets of American
children.1-3 Among school-aged children and adolescents, the proportion
From the Department of Nutrition, University of North
Carolina at Chapel Hill.
Submitted for publication Apr 7, 2000; revision received Aug 21, 2000; accepted Oct 5,
2000.
Reprint requests: Barry M. Popkin, PhD,
Carolina Population Center, University of
North Carolina at Chapel Hill, CB No. 8120
University Sq, 123 W Franklin St, Chapel
Hill, NC 27516-3997.
Copyright © 2001 by Mosby, Inc.
0022-3476/2001/$35.00 + 0 9/21/112162
doi:10.1067/mpd.2001.112162
classified as overweight has grown
from 8% to 14% for children and from
6% to 12% for adolescents between
1976 to 1980 and 1988 to 1994.3 The
consequences of childhood overweight
include abnormal lipid profiles, type 2
diabetes, and greater risk for adult obesity and its related comorbidities.1 Although major shifts in physical activity
patterns are associated with the increase in obesity, shifts in dietary pattern might also influence this outcome.
Public health researchers have pointed
to shifts in dietary behaviors such as
“supersizing” of the portions available
in many restaurants and fast food outlets, the rapid increase in soft drink
consumption, and increased availabili-
Continuing Survey of Food Intake by
Individuals
Nationwide Food Consumption Survey
METHODS
Survey Design and Sample
This study uses data on subjects aged
2 to 18 years from 3 nationally representative surveys of the US population. Of these 21,236 individuals,
12,231 participated in the 1977-78 Nationwide Food Consumption Survey
(NFCS77), 4008 participated in the
1989-91 Continuing Survey of Food
Intake by Individuals, and 4997 participated in the 1994-96 CSFII96. The
United States Department of Agriculture surveys from 1977 and 1989 contain stratified area probability samples
of noninstitutionalized US households
493
JAHNS, SIEGA-RIZ, AND POPKIN
THE JOURNAL OF PEDIATRICS
APRIL 2001
Table I. Proportion of snackers by sociodemographic characteristic in the continuing survey of food intake by individuals, 1994-96
Percent who snack*
Sociodemographic characteristics
Ages 2-5 y
(n = 2183)
Ages 6-11 y
(n = 1492)
Ages 12-18 y
(n = 1322)
94.3
93.4
94.9
90.2
93.9
91.5
96.6
91.7
96.9
92.6
90.0
94.4
86.2
82.6
86.9
95.7
88.0
94.7
88.4
86.6
89.9
78.8
85.8
83.7
91.5
84.9
91.0
Male (n = 2511)
Female (n = 2486)
Non-Hispanic white (n = 3424)
Non-Hispanic black (n = 760)
Hispanic (n = 813)
<185% of national poverty level (n = 2232)
>350% of national poverty level (n = 1291)
Mother’s education less than or equal to high school diploma (n = 2648)
Mother’s education more than high school diploma (n = 2392)
*Data are weighted to account for survey design effects and to be nationally representative.
Table II. Mean number of snacks per day, grams consumed per snack, kilocalories consumed per snack, and kilocalories obtained
per day from snacks by US children from the 1977, 1989, and 1996 CSFII surveys by age group
Ages 2-5 y
Snacks/day
Grams/snack
Kilocalories/snack
Kilocalories from snacks
Ages 6-11 y
Ages 12-18 y
1977
1989
1996
1977
1989
1996
1977
1989
1996
1.73
158
171
283
1.87
167
187
331*
2.29*†
153
175
378*†
1.56
200
231
347
1.59
205
250
387
1.99*†
195
243
462*†
1.60
275
296
460
1.62
298*
320
496
1.97*†
307
318
612*†
*Data point is significantly different from 1977 data point at the 0.01 level.
†Data point is significantly different from 1989 data point at the 0.01 level.
in the 48 contiguous states and in all 50
states from 1994 to 1996. These surveys
were self-weighting, multistage, stratified area samples of the US population.
Detailed information on the methodology pertaining to each survey has been
published previously.9-12
Dietary Data
The NFCS77 and CSFII89 surveys
collected 1 day of intake by in-home,
interviewer-administered, 24-hour recall and 2 additional days of selfadministered 1-day food records. The
CSFII96 collected 2 nonconsecutive,
interviewer-administered, 24-hour recalls. For children <12 years, information was obtained from the child’s caregiver. For our purpose of studying
snacking behavior over time, all the
days of dietary intake have been included in this analysis.
494
Definition of a “Snack”
For each food item reported, the respondent (or proxy) was asked “About
what time did you begin to eat (drink)
the (food item)?” Next, the respondent
was handed a card on which were
printed various eating occasions (ie,
breakfast, lunch, dinner, snack). The
respondent was then asked “Looking at
this card, please tell me what you
would call this occasion?” This was repeated for each food item reported (the
previous questions are verbatim from
CSFII96; however, NFCS77 and
CSFII89 asked similar questions).
Therefore our data contain each food
item a person consumed, along with the
self-reported eating occasion and beginning eating time for each food item.
Because people often consume more
than one food item when having a
“snack,” we chose not to consider each
self-reported snack food item as a separate snack. We arbitrarily grouped the
snack food items that had a beginning
eating time within 15 minutes of the first
snack food item reported as a discrete
snack. Once foods were categorized by
eating occasion (snacks vs non-snack
meals), the gram weight of each snack
and nutrients contributed by snacks and
non-snacks (meals) was computed for
each survey year by age group.
Statistical Analysis
For each survey year, the percentage
of individuals who consumed at least
one snack on the days covered was determined. Then for snackers, the mean
numbers of snacks consumed, mean
grams consumed per snack, and mean
energy intake from snacks were computed for each survey year. The contribution of snacking to total energy in-
JAHNS, SIEGA-RIZ, AND POPKIN
THE JOURNAL OF PEDIATRICS
VOLUME 138, NUMBER 4
take and total fat intake was also estimated. To examine differences of snacking behaviors by age, individuals were
stratified into 3 age groups; 2 to 5 years
old, 6 to 11 years old, and 12 to 18 years
old. The proportion of snackers by key
sociodemographic groups was compared in the 1996 survey. Differences
testing used SAS version 6.12 (SAS Institute, Cary, NC) and SUDAAN 7.51
(Research Triangle Institute, Research
Triangle Park, NC) statistical software
packages to weight the results and control standard errors for sample design
effects. A P value <.01 was used to denote statistical significance.
Fig 1. Percent of children consuming snacks by age group from the CSFII in 1977, 1989, and 1996.
Solid bars, 1977; open bars, 1989; hatched bars, 1996.
RESULTS
Who Is Snacking?
The prevalence of snackers among
children increased over all ages from
77% to 91% during the years covered.
The changes in the percent of snackers, overall and within age groups,
were minimal or nonexistent between
1977 and 1989; most of the increase occurred between 1989 and 1996 (Fig 1).
The percent of children consuming
snacks by sociodemographic characteristics in 1996 is shown in Table I. For
each sociodemographic characteristic
except Hispanic ethnicity, the oldest
children had the lowest proportion of
snackers. Slightly more boys than girls
snacked, and a greater proportion of
Hispanic and white children snacked
than black children in all age groups. A
higher proportion of snackers was
found at the higher income and education levels than at the lower levels.
Behavioral Changes in
Snacking Habits
The most significant change in
snacking behavior was the greater
number of snacking occasions per day,
not in the amount of energy consumed
per snack. Contribution of daily calories by snacks increased by 30% between 1977 and 1996, with large
changes occurring both between 1977
and 1989 and between 1989 and 1996
Fig 2. Contribution of snacking to total daily energy intake by year and age group for participants in
the CSFII surveys. Number above bar represents total caloric intake. Hatched bars, Energy from
meals; solid bars, energy from snacks.
(Table II). The mean number of snacks
per day increased by 24% to 32% (approximately 0.4 snacking events per
day) depending on age group, with the
major change occurring in the 1989 to
1996 period. Overall, the energy per
snack increased only 3%, and the
quantity in grams increased only 1%.
Nutritional Impact of Shifts in
Snacking Behavior
The contribution of non-snack meals
and snacks to total daily energy intake
is shown in Fig 2. The total energy intake overall and within each age group
increased from 1977 to 1996. Overall,
snacks contributed approximately one
fifth of energy intake in 1977, increasing to approximately one fourth of energy intake by 1996. The 1989 percent
energy from snacks values fell midway
between the 1977 and 1996 values in all
age groups. In every period examined,
the 12- to 18-year olds derived the highest proportion of calories from snacks.
Although total average fat intake de495
JAHNS, SIEGA-RIZ, AND POPKIN
THE JOURNAL OF PEDIATRICS
APRIL 2001
Table III. Key nutrient density of snacks and meals of US children in the CSFII 1977,
1989, and 1996 surveys by age group
Calcium
(mg/1000 kcal)
Age 2-5 y
1977 snack
1989 snack
1996 snack
1977 meal
1989 meal
1996 meal
Age 6-11 y
1977 snack
1989 snack
1996 snack
1977 meal
1989 meal
1996 meal
Age 12-18 y
1977 snack
1989 snack
1996 snack
1977 meal
1989 meal
1996 meal
Fiber
g/1000 kcal
Folate
(µg/1000 kcal)
557.0 (16.5)
500.0 (20.9)
488.9 (11.0)*
558.5 (6.4)
587.5 (12.5)
543.2 (5.0)†
6.8 (0.21)
6.5 (0.42)
7.7 (0.21)*†
6.5 (0.06)
6.9 (0.13)*
7.0 (0.07)*
94.5 (3.03)
103.6 (7.57)
101.6 (2.92)
152.8 (1.52)
162.2 (4.56)
154.0 (1.75)
503.7(11.2)
425.0 (19.4)*
399.8 (11.9)*
536.0 (4.0)
533.4 (8.8)
511.0 (7.1)*
7.2 (0.19)
7.0 (0.4)
7.1 (0.21)
6.7 (0.06)
7.0 (0.12)
7.0 (0.11)
90.4 (2.73)
99.2 (7.41)
94.7 (3.82)
148.0 (1.36)
153.8 (3.19)
146.3 (2.39)
476.3 (8.6)
457.2 (31.3)
380.1 (12.8)*
483.9 (4.6)
470.3 (10.2)
433.0 (5.7)*†
6.2 (0.15)
6.1 (0.25)
6.6 (0.25)
6.5 (0.07)
6.8 (0.14)
6.7 (0.09)
93.5 (3.14)
98.6 (8.49)
95.1 (5.49)
135.0 (1.37)
141.4 (3.34)
127.6 (2.87)†
*Data point is significantly different from 1977 data point at the 0.01 level.
†Data point is significantly different from 1989 data point at the 0.01 level.
creased for all age groups between
1977 and 1996 (there were small increases from 1989 to 1996 among 2- to
5-year-olds and 12- to 18-year-olds),
the absolute and relative proportions
of fat from snacks show an increasing
trend over the periods covered (Fig 3).
The largest increase in the proportion
of fat consumed as a snack occurred
among the 2 younger age groups, with
a smaller but statistically significant increase among 12- to 18-year-olds.
Micronutrient Contribution of
Snacks Compared with Meals
Table III presents mean nutrient intake from snacks and meals in each
survey year by age group. Regardless
of age, snacks were more energy dense
than meals, and the energy density of
snacks increased significantly over
time, whereas for meals it remained
stable (Table III). The increase in the
496
energy density of snacks over all age
groups was 0.14, equal to 0.2 kcal/g of
food. Fat density of meals decreased
significantly for all age groups, whereas the fat density of snacks did not
change for 2- to 5-year-olds or 6- to 11year-olds. Calcium is of concern because of the rapid displacement of milk
with sweetened beverages.13 These
data indicate a decline in calcium density in snacks and meals during the
past 20 years for all age groups. This
decline is greater in magnitude in
snacks than in meals. Iron density has
increased significantly in both snacks
and meals for all age groups, although
snacks are still less iron-dense than
meals. Although zinc density of snacks
is only approximately half that of
meals, there has been a decrease in
zinc density in meals for all age groups.
However, among 2- to 5-year-olds, the
zinc density of snacks increased. Fiber
increased in meals while remaining
fairly stable in snacks throughout the
years. Folate density of snacks and
meals increased in 1989 but then declined in 1996.
Several other nutrients were examined but not reported. Sodium and
niacin increased in snacks for all age
groups over time. Also, thiamin and
polyunsaturated fat increased in
snacks eaten by 2- to 5-year-olds and
6- to 11-year-olds, whereas vitamin B6
remained stable in snacks eaten by 6to 11-year-olds. Vitamin C and monounsaturated fats also increased in
snacks eaten by 2- to 5-year-olds. Saturated fat, cholesterol, B12, riboflavin,
and phosphorus declined in all age
groups’ snacks, whereas protein, magnesium, and potassium decreased in
snacks consumed by 6- to 11-yearolds. For all age groups, there was no
change in vitamin A, vitamin E,
carotene, or copper obtained from
snacks.
DISCUSSION
This study presents a picture of an
important increase in the prevalence
and dietary contribution of snacking in
the diets of US children. Currently,
snacking represents approximately a
fourth of the energy and a fifth of many
other nutrients, similar to any other
meal occasion. The increase in nutrient
contribution is primarily caused by an
increase in the frequency of snacking
and not the size of each snack. Moreover, energy density is much higher for
snacks than the non-snacking component of the diet, a component that has
increased dramatically between 1977
and 1996 (approximately .14 to .20
kcal/g more).
Given the public health concern of
increasing child obesity rates, there are
several interesting issues raised by this
analysis. Supersizing of portions is
raised as one of the causes of the excess
dietary intake of the American public.
These data showed no significant
JAHNS, SIEGA-RIZ, AND POPKIN
THE JOURNAL OF PEDIATRICS
VOLUME 138, NUMBER 4
changes in the reported size of snacks
consumed over the last 20 years. In
fact, the size of the average snack consumed remained almost constant between 1977 and 1996. What does
emerge as a potential contributor is the
increased energy density of snacks
amid a constant energy density of
meals. Energy density is a critical element in overconsumption.14-16 Some
link increased energy density with less
satiety per gram of food intake and
hence with passive overconsumption
and increased obesity.15-17
The potential limitations of this
study include bias in unknown directions by changes in survey methods
throughout the periods surveyed. The
nonlinearity of the decrease in fat intake shown in Fig 3 may be an artifact
of improved probing methods, as may
be the increase in the number of
snacks. However, no studies have explored the effect of these methodologic
changes on nutrient estimates in the
CSFII surveys. In addition to changes
in survey methodology, real changes in
the food supply occur, and therefore
the composition of the nutrient databases is critical for detecting changes.
Each survey used its own United
States Department of Agriculture
nutrient database. However, we have
created a way to link food codes from
previous surveys and then apply values
from the 1994 United States Dietary
Association nutrient database to the
1977 and 1989 surveys. The 1994-96
survey uses an updated database for
each year of collection (1994, 1995,
and 1996), but we believe that fewer
changes occurred between 1994 and
1996 than between 1977 and 1996.
Real changes in the food supply will
not be captured by this method, and
trends may be over- or underestimated. In weighing the advantages and
disadvantages of using different nutrient databases, we concluded that using
the 1994 nutrient data provides a conservative measure of some of the nutrient trends in food (eg, fat) and allows
for the examination of trends in several
Fig 3. Contribution of snacking to total daily fat intake by year and age group for participants in
CSFII surveys. Number above bar represents total fat intake (in grams). Hatched bars, Fat from
meals; solid bars, fat from snacks.
nutrients (eg, fiber) that would not
have been possible otherwise. Another
issue pertains to the number of days of
data collection. In the first 2 surveys, 3
days of data were used to identify
snackers, whereas in 1994-96 only 2
days were used. In 1977 and 1989 only
approximately 4% of the subjects
snacked only on the third day of data
collection. It was our decision to err on
the conservative side and include all
days of intake. Our measurement of a
snacking occasion is a more conservative one than others have used. In this
study all self-defined snack foods consumed within a 15-minute span were
grouped to represent one snacking occasion, whereas other researchers have
counted each snack food eaten at a
unique time interval as a snacking occasion.6,7 There is also a difference in
how snacks are defined. Time of day,
type of food, and foods not linked with
a meal have been used to define a
snack.8,18-21 To date, there is no consensus in the literature as to how to define what constitutes a meal or a snack.
It is clear that there is an increase in
the role of snacking and a strong possi-
bility that snacking has contributed to
the epidemic of obesity in US children.
Longitudinal studies are required to
examine this potential pathway. Meanwhile, there is a need in the food supply for less energy dense and more
healthful snacks that are appealing to
children.
REFERENCES
1. Pinhas-Hamiel O, Dolan LM, Daniels
SR, Standiford D, Khoury PR, Zeitler
P. Increased incidence of non-insulindependent diabetes mellitus among
adolescents. J Pediatr 1996;128:608-15.
2. Troiano RP, Flegal KM. Overweight
children and adolescents: description,
epidemiology, and demographics. Pediatrics 1998;101:497-504.
3. U.S. Department of Health and
Human Services. Update: Prevalence
of overweight among children, adolescents, and adults—United States 198894. MMWR Morb Mortal Wkly Rep
1997;46:199-202.
4. Harnack L, Stang J, Story M. Soft drink
consumption among US children and
adolescents: nutritional consequences. J
Am Diet Assoc 1999;4:436-41.
5. Jacobsen M, Nestle M. Halting the
obesity epidemic: a public health poli497
JAHNS, SIEGA-RIZ, AND POPKIN
6.
7.
8.
9.
10.
cy approach. Public Health Rep
2000;115:12-24.
Frazao E, editor. American’s eating
habits: changes and consequences.
Washington (DC): US Department of
Agriculture Information Bulletin
Number 750; 1999. p. 1-473.
Bigler-Doughton S, Jenkins R. Adolescent snacks: nutrient density and
nutritional contribution to total intake.
J Am Diet Assoc 1987;87:1678-9.
Cross A, Babicz D, Cushman LF.
Snacking patterns among 1,800 adults
and children. J Am Diet Assoc 1994;
94:1398-403.
Rizek RL. The 1977-78 nationwide
food consumption survey. Fam Econ
Rev 1978;4:3-7.
Tippett KS, Mickle SJ, Goldman JD,
Sykes KE, Cook DA, Sebastian RS, et
al. Food and nutrient intakes by individuals in the United States, 1 day,
1989-91. Continuing survey of food intakes by individuals 1989-91, Nationwide Food Surveys Rep. No 91-2. U.S.
Department of Agriculture, Agriculture Research Service, 1995.
THE JOURNAL OF PEDIATRICS
APRIL 2001
11. Tippett KS, Cypel YS. Design and operation: the continuing survey of food
intakes by individuals and the diet and
health knowledge survey 1994-96.
Continuing survey of food intakes by
individuals 1994-96, Nationwide Food
Surveys Rep. 96-1. U.S. Department
of Agriculture, Agriculture Research
Service, 1997.
12. Siega-Riz AM, Popkin BM, Carson T.
Trends in breakfast consumption for
children in the united states from 1965
to 1991. Am J Clin Nutr 1998;67
(suppl):748s-56s.
13. Cavadini C, Siega-Riz AM, Popkin
BM. US adolescent’s food intake
trends from 1965-1996. Arch Dis Child
2000;83:18-24.
14. Bray GA, Popkin BM. Dietary fat intake does affect obesity! Am J Clin
Nutr 1998;68:1157-73.
15. Drewnowski A. Energy density,
palatability, and satiety: implications
for weight control. Nutr Rev 1998;56:
347-53.
16. Rolls BJ, Kim-Harris S, Fischman
17.
18.
19.
20.
21.
MW, Foltin RW, Moran TH, Stoner
SA. Satiety after preloads with different amounts of fat and carbohydrate:
implications for obesity. Am J Clin
Nutr 1994;60:476-87.
Blundell JE, Macdiarmid JI. Passive
overconsumption. Fat intake and
short-term energy balance. Ann NY
Acad Sci 1997;827:392-407.
Devaney BL, Gordon AR, Burghardt
JA. Dietary intakes of students. Am J
Clin Nutr 1995;61(suppl):205S-12S.
Ezell JM, Skinner JD, Penfield MP.
Appalachian adolescents’ snack patterns: morning, afternoon, and evening
snacks. J Am Diet Assoc 1985;85:
1450-4.
McCoy H, Moak S, Kenney MA,
Kirby A, Chopin L, Billon W. Snacking patterns and nutrient density of
snacks consumed by southern girls. J
Nutr Epidemiol 1986;18:61-6.
Summerbell CD, Moody RC, Shanks
J, Stock MJ, Geissler C. Sources of
energy from meals versus snacks in
220 people in four age groups. Eur J
Clin Nutr 1995;49:33-41.
Receive tables of contents by e-mail
To receive the tables of contents by e-mail, sign up through our Web site at
http://www.mosby.com/jpeds
Choose E-mail Notification.
Simply type your e-mail address in the box and click the Subscribe button.
Alternatively, you may send an e-mail message to [email protected]. Leave the subject line blank and type
the following as the body of your message:
subscribe jpeds_toc
You will receive an e-mail to confirm that you have been added to the mailing list.
Note that table of contents e-mails will be sent out when a new issue is posted to the Web site.
498