Download Use of a DASH Food Group Score to Predict Excess Weight Gain in

ARTICLE
Use of a DASH Food Group Score to Predict
Excess Weight Gain in Adolescent Girls
in the National Growth and Health Study
Jonathan P. B. Berz, MD, MSc; Martha R. Singer, MPH; Xinxin Guo, MPH;
Stephen R. Daniels, MD, PhD; Lynn L. Moore, DSc
Objective: To study the effects of selected dietary patterns, particularly a DASH (Dietary Approach to Stop Hypertension) eating pattern, on body mass index (BMI)
throughout adolescence.
Design: Prospective National Growth and Health Study.
Setting: Washington, DC; Cincinnati, Ohio; and Berkeley, California.
Participants: A total of 2327 girls with 10 annual visits starting at age 9 years.
Main Exposures: Individual DASH-related food groups
and a modified DASH adherence score.
Main Outcome Measure: The BMI value from measured yearly height and weight over 10 years.
and a DASH adherence score on BMI during 10 years of
follow-up, adjusting for race, height, socioeconomic status, television viewing and video game playing hours,
physical activity level, and total energy intake. Girls in
the highest vs lowest quintile of the DASH score had an
adjusted mean BMI of 24.4 vs 26.3 (calculated as weight
in kilograms divided by height in meters squared)
(P ⬍ .05). The strongest individual food group predictors of BMI were total fruit (mean BMI, 26.0 vs 23.6 for
⬍1 vs ⱖ2 servings per day; P ⬍.001) and low-fat dairy
(mean BMI, 25.7 vs 23.2 for ⬍1 vs ⱖ2 servings per day;
P⬍ .001). Whole grain consumption was more weakly
but beneficially associated with BMI.
Conclusions: Adolescent girls whose diet more closely
resembled the DASH eating pattern had smaller gains in
BMI over 10 years. Such an eating pattern may help prevent excess weight gain during adolescence.
Results: Longitudinal mixed modeling methods were
used to assess the effects of individual DASH food groups
O
Author Affiliations: Sections of
General Internal Medicine
(Dr Berz) and Preventive
Medicine and Epidemiology
(Ms Singer and Dr Moore),
Boston University Medical
Center, Boston, and Outcome
Sciences Inc, Cambridge
(Ms Guo), Massachusetts; and
Department of Pediatrics,
University of Colorado School
of Medicine, Anschutz Medical
Campus, Denver (Dr Daniels).
Arch Pediatr Adolesc Med. 2011;165(6):540-546
BESITY IS A MAJOR PUBLIC
health problem, with 17%
of American children
overweight and 67% of
adults either overweight
or obese.1-3 Excess weight during childhood leads to numerous health problems
and is even associated with premature death
as an adult.4,5 Few studies have examined
the relation of food-based dietary patterns
with weight gain, especially in children.
The examination of food-based dietary patterns acknowledges the synergistic effects on health that food nutrients may
have when eaten together.6 One example
is the DASH (Dietary Approach to Stop Hypertension) diet pattern. It emphasizes increased intakes of low-fat dairy products;
fish, chicken, and lean meats (to decrease saturated fat and increase calcium
levels); and nuts, fruits, whole grains, vegetables, and legumes (to increase potassium, magnesium, and dietary fiber
levels).7 The DASH eating pattern was
originally studied in clinical trials of adults
ARCH PEDIATR ADOLESC MED/ VOL 165 (NO. 6), JUNE 2011
540
as a treatment for hypertension8; these
clinical trials assessed the effects of increased fruit and vegetable intake, with or
without increasing the intake of low-fat
dairy products. In these studies,9-11 the
combined diet (rich in fruit, vegetables,
and low-fat dairy products) led to the
greatest reductions in blood pressure. The
DASH pattern has also been studied in relation to the metabolic syndrome and selected cardiovascular end points.8-14 Little
has been done, however, to examine the
effects of a DASH eating pattern on measures of excess weight, frequently a precursor of the aforementioned conditions.
In addition, the DASH eating pattern has
See also pages
567 and 580
been infrequently studied in children, and
the American Academy of Pediatrics states
that there is no reason to suggest that using
DASH would not be safe as long as proWWW.ARCHPEDIATRICS.COM
©2011 American Medical Association. All rights reserved.
Downloaded From: http://archfaci.jamanetwork.com/ on 08/03/2017
tein and calories are consumed in quantities adequate to
support child and adolescent growth and development
needs.15 In this study, we examined the effects of adherence to a DASH-style eating plan and its components on
the change in body mass index (BMI) in a racially diverse sample of adolescent girls.
METHODS
STUDY POPULATION
The National Growth and Health Study was initiated by the National Heart, Lung, and Blood Institute to investigate racial differences in dietary, physical activity, family, and psychosocial
factors associated with the development of obesity in black and
white girls. The National Growth and Health Study enrolled
2379 girls aged 9 and 10 years in 3 cities (Washington, DC;
Berkeley, California; and Cincinnati, Ohio) in 1987-1988 and
observed them for 9 years. Data were collected in a longitudinal manner on 10 occasions via follow-up at annual examinations. Height and weight were measured by trained study staff
using standardized assessment protocols at each examination.
Additional details of the methods used for ascertaining data are
described elsewhere.16 Almost 90% of the girls originally enrolled were observed through study year 10. This study was
approved by the Boston University institutional review board.
MAIN OUTCOME VARIABLE
The main outcome of interest was BMI (calculated as weight
in kilograms divided by height in meters squared) at each age
from 9 to 19 years.
DIETARY EXPOSURE VARIABLES
Dietary data were collected using 3-day diet records; the collection included 2 weekdays and 1 weekend day during each
of 8 examination years. Participants were trained by a study
nutritionist to record detailed dietary information using standard household measuring instruments for the estimation of
portion sizes. Standardized debriefing was performed, and diet
records that were considered unreliable by the research dietitian were excluded.
Dietary data were entered into the Nutrition Data System
of the University of Minnesota, Minneapolis, to estimate the
intake of total calories, macronutrients, and micronutrients.17
The Nutrition Data System also outputs food codes for each
food and each ingredient from composite foods (eg, from lasagna, macaroni and cheese, and even condiments). The Nutrition Data System food code data were combined with the US
Department of Agriculture’s survey food code database, the Food
and Nutrient Database for Dietary Studies, version 2.0.18 By
matching these food codes, the child’s average daily intake was
derived in each of the 5 major food groups and in all the subgroups as defined by Nutrition and Your Health: Dietary Guidelines for Americans by the US Department of Agriculture.19 Thus,
we derived total servings for each group and subgroup. For example, fruit servings included fruit from all sources, such as
whole fruit, fruit-based desserts, 100% fruit juice, and that portion of fruit drinks derived from fruit juice.
DASH FOOD GROUP SCORE
We created a modified DASH food group score based on a previous publication by Levitan et al.20 This original score was
designed to reflect adherence to a DASH eating pattern as de-
scribed in the 2005 Dietary Guidelines for Americans.21 Levitan
et al20 compared DASH scores for this scale with those of another DASH score by Fung et al13 and found them to be moderately well correlated (r = 0.61). Because the Dietary Guidelines for Americans differs across levels of energy intake, we used
energy-specific standards for intake in each food group. The
score contained 10 food groups or subgroups, 3 of which were
excluded from the modified score: added sugars, discretionary fats and oils, and alcohol. Added sugars were excluded because the high intake of sugar in this population led to almost
all the participants having a score of zero for this component.
Discretionary fats and oils contributed nothing to the prediction of BMI in this analysis, and the alcohol component was
excluded because of the ages of the girls. Therefore, we focused on the 7 DASH-related food groups in these analyses: fruits,
vegetables, low-fat dairy products, total and whole grains, lean
meats, and nuts, seeds, and legumes.
Low-fat dairy products were defined as those containing 2%
fat or less. Lean meat was defined as fish, eggs, beef, and poultry that was 85% lean or greater. To obtain more stable estimates of intake, we included only girls with 2 or more sets of
3-day diet records collected between ages 9 and 17 years (2330
of the original 2379 participants). One girl with an average intake of less than 1000 kcal/d and 1 with an average of more
than 3500 kcal/d at ages 9 to 17 years were excluded from the
study, as well as 1 girl with absent physical activity data, leaving a final sample of 2327 girls with available data who were
included in these analyses.
We followed the scoring protocol of Levitan et al.20 Each
food group was assigned a score ranging from 0 to 1. For total
grains, meats, low-fat dairy products, and nuts, seeds, and legumes, participants with intake meeting the guidelines were
assigned 1 point. Those with intake above the recommended
levels were assigned partial points as follows: 1 minus the percentage of intake over the guideline. For intake below the guideline level, partial points were assigned by dividing the actual
intake by the recommended intake. For fruits, vegetables, and
whole grains where optimal intake was deemed to be at or greater
than recommended, a full point was assigned to those who consumed the recommended level of intake or higher. Partial points
were given only for those who had less than the recommended intake. Because DASH recommends that most grains
be whole, we used 50% of the total grain recommended as the
goal for whole grain intake in accord with American Heart Association guidelines.15 The total DASH score was, thus, a sum
of the scores for each individual food group.
POTENTIAL CONFOUNDING VARIABLES
Potential confounding factors that were evaluated for inclusion in these analyses included race, height at each age, socioeconomic status (SES), physical activity level, television viewing and video game playing (hours per day), total energy intake,
and other dietary factors.22 The SES was classified as low, moderate, or high by combining information about parental income and education. Low-SES families were those with incomes of less than $10 000 per year or parental education level
of less than high school; high-SES families included those with
incomes of at least $40 000 per year and at least a high school
education. All the other participants were classified as moderate SES. Physical activity was measured at each visit using a
validated questionnaire that asked the participants to report the
frequency and duration per week (during the school year and
summer) of participation in a variety of structured physical activities in the past year.23 These data were combined with published information on metabolic equivalent levels to obtain a
final score estimating daily energy expenditure.24-26
ARCH PEDIATR ADOLESC MED/ VOL 165 (NO. 6), JUNE 2011
541
WWW.ARCHPEDIATRICS.COM
©2011 American Medical Association. All rights reserved.
Downloaded From: http://archfaci.jamanetwork.com/ on 08/03/2017
Table 1. Characteristics of the Study Population by Quintile of DASH Adherence Score a
Quintile
Characteristic
DASH adherence score, range b
Food group DASH score, mean (SD),
servings/d c
Total grains
Vegetables
Lean meats
Fruits
Low-fat dairy products
Whole grains
Nuts/seeds/legumes
Physical activity score, mean (SD)
Television viewing and video game
playing, mean (SD), h/d
Total energy intake, mean (SD), kcal/d
Race, row %
White (n = 1139)
Black (n = 1189)
SES, row %
Low (n = 548)
Middle (n = 996)
High (n = 784)
1
(n=465)
2
(n = 465)
3
(n = 466)
4
(n = 466)
5
(n = 465)
1.3-2.6
2.6-2.9
2.9-3.3
3.3-3.6
3.6-5.2
5.74 (1.72)
1.63 (0.65)
1.34 (0.71)
0.80 (0.53)
0.63 (0.40)
0.36 (0.28)
0.18 (0.23)
18.1 (9.6)
5.3 (2.1)
6.16 (1.53)
2.07 (0.85)
1.67 (0.82)
0.98 (0.59)
0.78 (0.51)
0.43 (0.31)
0.24 (0.24)
18.7 (10.1)
5.4 (2.1)
6.40 (1.48)
2.18 (0.82)
1.75 (0.91)
1.16 (0.69)
0.97 (0.57)
0.51 (0.32)
0.27 (0.20)
19.9 (10.0)
5.2 (2.1)
6.45 (1.41)
2.24 (0.88)
2.05 (1.00)
1.43 (0.80)
1.12 (0.67)
0.59 (0.36)
0.32 (0.20)
20.5 (10.7)
4.9 (2.2)
6.48 (1.24)
2.38 (0.86)
2.13 (1.07)
1.93 (1.05)
1.52 (0.76)
0.77 (0.43)
0.38 (0.20)
23.2 (10.5)
3.8 (2.1)
1686 (388)
1872 (369)
1912 (374)
1949 (354)
1944 (290)
14.8
24.9
15.3
24.5
19.2
20.8
20.6
19.4
30.0
10.4
26.9
21.4
13.4
27.2
20.1
14.8
21.0
22.3
16.5
16.3
19.4
23.5
8.6
16.9
31.9
Abbreviations: DASH, Dietary Approach to Stop Hypertension; SES, socioeconomic status.
a P ⬍ .001 for all.
b The maximum possible DASH score was 7.
c Dietary intakes are averages across approximately 20 days of diet records.
STATISTICAL ANALYSIS
Confounders were evaluated using a forward selection method.
Factors determined to be confounders and those that were independent predictors of the outcome were retained in the final models (ie, race, height, SES, physical activity level, television viewing and video game playing, and total energy).
Categories of average intake in each of the DASH food groups
were determined by balancing information about the distributions of the intake population (which affects study power) with
the recommended intake levels. For example, DASH recommended intake level of fruits is 4 to 5 servings per day, but the
category cutoff values used in the analysis were less than 1, 1
to less than 2, and 2 or more servings per day because few participants actually consumed 4 to 5 servings. Additional analyses were conducted to evaluate the sensitivity of the results to
subtle changes in category definitions.
To determine the association between level of DASH food group
intake and BMI over time, we used longitudinal data analysis methods, accounting for correlated observations from the repeatedmeasures data. In separate models, each categorical food group
variable was entered as the main exposure variable, with age as
an interaction factor, while controlling for fixed and changing potential confounders as previously described. This allowed us to
estimate the adjusted mean BMI at each age in each category of
intake for each food group. Analyses were conducted using an unstructured covariance matrix in Proc Mixed with the repeated option in SAS (SAS Institute Inc, Cary, North Carolina).22 The same
longitudinal mixed modeling methods were used to estimate the
adjusted mean BMI at each age according to quintile of DASH score.
In each model, the interaction of age and food group was
examined first. When a significant interaction was found, further testing was performed to evaluate differences between the
slopes, intercepts, and BMIs at the end of follow-up. When there
was no significant age–food group interaction (P⬎.05), the sta-
tistical significance of the fixed effects for the main dietary exposure variable was examined (using type III sums of squares).
Approximate linearity of the relationship between age and BMI
was assumed. All analyses were performed using a commercially available statistical software program (SAS, version 9.1).
RESULTS
STUDY CHARACTERISTICS
Dietary and population characteristics by quintile of DASH
score are presented in Table 1. The overall mean DASH
adherence score was 3.1, with a median of 3.1 (range,
1.3-5.2). Food group means in each quintile show that
higher DASH scores were associated with higher intake
in most food groups. Higher DASH scores were also associated with higher total energy intake. Black participants and those with lower SES were more likely to be
in a lower quintile of DASH scores. In addition, there was
higher mean physical activity and lower mean television viewing and video game playing hours in the highest quintile of the DASH score.
The distribution of actual intakes for each food component of the DASH score is given in Table 2, along with
the recommended DASH intakes. Even study participants in the 95th percentile of intake had relatively low
intake of fruits, vegetables, whole grains, and low-fat dairy
products compared with the DASH recommendations.
The average intake of added sugars was approximately
10 times higher than recommended. Discretionary fat intake was also relatively high, although no direct equivalent DASH recommendation applies.
ARCH PEDIATR ADOLESC MED/ VOL 165 (NO. 6), JUNE 2011
542
WWW.ARCHPEDIATRICS.COM
©2011 American Medical Association. All rights reserved.
Downloaded From: http://archfaci.jamanetwork.com/ on 08/03/2017
Table 2. Actual and Recommended Intakes of DASH Food Groups a
Servings per Day, Percentile b
Food Group
5th
50th
95th
DASH Recommendations,
Servings per Day
Discretionary fat
Grains
Added sugar
Vegetables
Lean meat
Fruits
Whole grains
Nuts/seeds/legumes
7.5
4.0
3.7
1.0
0.5
0.3
0.1
0
11.7
6.2
7.2
2.0
1.6
1.1
0.5
0.2
17.7
8.9
12.0
3.6
3.6
2.9
1.2
0.7
NA
6
0.7
3-4
1-2
4
3
0.5
Abbreviations: DASH, Dietary Approach to Stop Hypertension; NA, no DASH recommendation applies.
a Based on intake of 1600 kcal/d.
b Serving sizes for each food group: lean meat, 3 oz; low-fat dairy, 1 cup milk or yogurt or 1.5 oz cheese; nuts/seeds/legumes, 1⁄3 cup nuts, 2 tbsp seeds, and
1⁄2 cup cooked dry beans; (whole) grains, 1 slice of bread, 1 oz dry cereal, and 1⁄2 cup cooked rice, pasta, or cereal; vegetables, 1 cup raw/leafy, 1⁄2 cup cooked, and
6 oz vegetable juice; fruits, 6 oz fruit juice, 1 medium piece, 1⁄4 cup dried, and 1⁄2 cup fresh, frozen, or canned; added sugar, 1 tbsp; and discretionary fat, 5 g/1 tsp.
27
26
<1
≥2
1 to < 2
25
<1
2 to < 3
≥3
24
BMI
BMI
23
22
21
20
19
17
18
9
10
11
12
13
14
15
16
17
18
19
9
10
11
12
Age, y
Figure 1. Body mass index (BMI) (calculated as weight in kilograms divided
by height in meters squared) over 10 years associated with fruit
consumption (mean servings per day). Adjusted for race, height,
socioeconomic status, physical activity level, television viewing and video
game playing hours per day, added sugar, total dairy, vegetables, total
grains, nuts/seeds/legumes, processed and nonprocessed meat, and total
energy intake. Slopes: P⬍ .001 for differences between each line. Difference
in BMI at end of follow-up: less than 1 vs 1 to less than 2 servings, less
than 1 vs 2 to less than 8 servings, and 1 to less than 2 vs 2 to less than
8 servings, P ⬍.001 for all.
13
14
15
16
17
18
19
Age, y
Figure 2. Body mass index (BMI) (calculated as weight in kilograms divided
by height in meters squared) over 10 years associated with vegetable
consumption (mean servings per day). Adjusted for race, height,
socioeconomic status, physical activity level, television viewing and video
game playing hours per day, added sugar, total dairy, fruit, total grains,
nuts/seeds/legumes, processed and nonprocessed meat, and total energy
intake. For the overall difference between age and total grain interaction,
P = .16. For the overall difference between intake categories, P ⬎ .99. The
BMI at the end of follow-up did not statistically differ significantly among
the 3 consumption groups.
INTAKE OF
INDIVIDUAL FOOD GROUPS VS BMI
DASH FOOD GROUP SCORE
AS A PREDICTOR OF BMI
Figures 1, 2, 3, and 4 show the adjusted mean BMI at
each age associated with average intake in 4 DASH food
groups: fruits, vegetables, whole grains, and low-fat dairy
products. Participants who consumed 2 or more servings of fruit per day had the smallest gain in BMI over
time (P ⬍ .001) and the lowest BMI at the end of follow-up (23.6, 25.0, and 26.0 for low, moderate, and higher
intakes of fruit, respectively) (Figure 1 and Table 3).
No differences were noted in BMI according to intake of
vegetables (Figure 2 and Table 3). Highest (vs lowest)
whole grain intake conferred lower BMI increases over
time (P=.01) and a lower BMI at the end of follow-up
(Figure 3 and Table 3). Higher intake of low-fat dairy
products led to lower BMI gains over time (Figure 4 and
Table 3). In data not shown, we compared models including and excluding total energy and total and saturated fat as a percentage of energy intake. No substantive differences in the BMI trajectory were observed.
Figure 5 shows adjusted mean BMIs at each age associated with quintiles of the DASH score, averaged over
ages 9 to 17 years. Girls in the highest quintile had the
smallest gains in BMI over time and the lowest BMIs at
the end of follow-up (Table 3). In addition, at age 19 years,
those in the lowest quintile of the DASH score (compared with those in the highest quintile) had a mean BMI
that was greater than the threshold for overweight as defined by the 85th percentile for age.27
COMMENT
In this longitudinal cohort of adolescent girls, we found
that higher adherence to a DASH-style diet resulted in a
consistently lower BMI between the ages of 9 and 19 years.
These findings were stable over a 10-year follow-up and
ARCH PEDIATR ADOLESC MED/ VOL 165 (NO. 6), JUNE 2011
543
WWW.ARCHPEDIATRICS.COM
©2011 American Medical Association. All rights reserved.
Downloaded From: http://archfaci.jamanetwork.com/ on 08/03/2017
27
Table 3. BMI at Baseline and End of Follow-up for 4 DASH
Food Groups and DASH Adherence Score a
< 0.25
0.25 to < 1
≥1
25
BMI, Mean (SD)
BMI
23
Food Group and Intake,
Mean Servings per Day
21
19
17
9
10
11
12
13
14
15
16
17
18
19
Age, y
Figure 3. Body mass index (BMI) (calculated as weight in kilograms divided
by height in meters squared) over 10 years associated with whole grain
intake (mean servings per day). Adjusted for race, height, socioeconomic
status, physical activity level, television viewing and video game playing
hours per day, added sugar, total dairy, fruit, vegetables, nuts/seeds/legumes,
processed and nonprocessed meat, and total energy intake. Slopes: less than
0.25 vs 0.25 to less than 1 serving, P=.62; less than 0.25 vs 1 to less than
7 servings, P=.05; and 0.25 to less than 1 vs 1 to less than 7 servings, P=.01.
The BMI at the end of follow-up: less than 0.25 vs 0.25 to less than 1 serving,
P=.51; and less than 0.25 vs 1 to less than 7 servings, P=.002.
27
< 1.0
1.0 to < 2.25
≥ 2.25
25
BMI
23
21
19
17
9
10
11
12
13
14
15
16
17
18
19
Participants,
No.
Baseline
End of
Follow-up
1060
882
385
19.4 (0.16)
19.1 (0.16)
19.1 (0.22)
26.0 (0.19)
25.0 (0.21)
23.6 (0.32)
1192
833
302
19.3 (0.16)
19.2 (0.17)
19.2 (0.27)
25.2 (0.19)
25.3 (0.22)
25.1 (0.36)
531
1543
253
19.2 (0.20)
19.2 (0.14)
19.0 (0.26)
25.5 (0.26)
25.3 (0.16)
24.1 (0.38)
1363
834
130
19.3 (0.15)
19.2 (0.17)
18.6 (0.37)
25.7 (0.17)
24.9 (0.22)
23.2 (0.55)
465
465
466
466
465
19.7 (0.21)
19.0 (0.20)
19.3 (0.20)
19.0 (0.20)
19.1 (0.20)
26.3 (0.28)
24.9 (0.28)
25.2 (0.28)
25.3 (0.29)
24.4 (0.30)
Fruits
⬍1
1 to ⬍2
ⱖ2
Vegetables
⬍2
2 to ⬍3
ⱖ3
Whole grains
⬍0.25
0.25 to ⬍1
ⱖ1
Low-fat dairy products
⬍1.0
1.0 to ⬍2.25
ⱖ2.25
DASH adherence score
Quintile 1
Quintile 2
Quintile 3
Quintile 4
Quintile 5
Abbreviations: BMI, body mass index (calculated as weight in kilograms
divided by height in meters squared); DASH, Dietary Approach to Stop
Hypertension.
a The mixed models did not drop the girls below 1000 kcal/d (1 girl) and
above 3500 kcal/d (1 girl). Dropping them and rerunning the program gives
results that only differ at the second decimal place, so the data do not need
to be changed. However, data are from 2327 instead of 2328 girls because
one girl is missing activity at all ages.
Age, y
27
Quintile 1
Quintile 2
Quintile 3
Quintile 4
Quintile 5
25
23
BMI
Figure 4. Body mass index (BMI) (calculated as weight in kilograms divided
by height in meters squared) over 10 years associated with low-fat dairy
products intake (mean servings per day). Adjusted for race, height,
socioeconomic status, physical activity level, television viewing and video
game playing hours per day, added sugar, fruit, vegetables, total grains,
nuts/seeds/legumes, processed and nonprocessed meat, and total energy
intake. Slopes: less than 1 vs 1 to less than 2.25 servings, P=.001; less than
1 vs 2.25 to less than 5 servings, P ⬍ .001; and 1 to less than 2.25 vs 2.25 to
less than 5 servings, P =.02. The BMI at the end of follow-up: less than 1 vs
1 to less than 2.25 servings, P =.007; and less than 1 vs 2.25 to less than
5 servings, P ⬍.001.
21
19
P <.05
17
after controlling for nondietary factors associated with
eating patterns and excess weight gain.
Few studies have examined dietary patterns in children or used longitudinal data to examine their effects on
weight gain. A cross-sectional study28 of Korean preschool children found that a diet pattern that shares components of the DASH eating pattern was not associated with
measured weight status. One longitudinal study29 of women
showed that a pattern of intake lower in fruit, vegetables,
and low-fat foods was associated with a greater chance of
becoming overweight, and another study30 showed that a
diet consisting of many components present in the DASH
pattern resulted in smaller gains in BMI over time.
The present findings for the DASH score were mirrored by the effects of some of the individual food group
components. In particular, higher consumption of fruits,
whole grains, and low-fat dairy products led to less weight
9
10
11
12
13
14
15
16
17
18
19
Age, y
Figure 5. Body mass index (BMI) (calculated as weight in kilograms divided
by height in meters squared) over 10 years associated with DASH (Dietary
Approach to Stop Hypertension) adherence score (quintile). Adjusted for
race, height, socioeconomic status, physical activity level, television viewing
and video game playing hours per day, and total energy intake. Slopes:
quintiles 1 to 4 vs quintile 5, P ⬍ .05. The BMI at the end of follow-up:
quintiles 1, 3, and 4 vs quintile 5, P ⬍ .05.
gain. The observed fruit intake in this study was well below the DASH recommendation of 4 servings per day;
on average, at 9 to 17 years of age, only 15% of girls
reached this goal. In addition, higher vegetable consumption was not associated with decreased weight gain over
time. It is possible that relatively low intakes of vegetables and the narrow range of types of vegetables con-
ARCH PEDIATR ADOLESC MED/ VOL 165 (NO. 6), JUNE 2011
544
WWW.ARCHPEDIATRICS.COM
©2011 American Medical Association. All rights reserved.
Downloaded From: http://archfaci.jamanetwork.com/ on 08/03/2017
sumed (ie, a predominance of starchy vegetables) may
explain the absence of a protective effect. Indeed, in a
subanalysis, a higher intake of nonstarchy vegetables was
associated with a lower BMI at the end of follow-up (data
not shown).
The diet records used in the present study may provide more precise ascertainment of total fruit intake in
children than do Frequency Food Questionnaire methods because we extracted fruit servings from composite
dishes, as previously described. Thus, the present study
is likely to have less nondifferential misclassification of
diet exposures and a greater ability to detect meaningful
associations between diet and BMI.
Higher low-fat dairy product consumption resulted in
smaller increases in BMI during adolescence. Data on dairy
consumption and excess weight change during adolescence show mixed results in the larger literature. Two
small studies31,32 found no effect of dairy intake on weight
gain, and another study33 showed an adverse effect of
higher dairy consumption on weight gain, even for lowfat milk, although the effect seemed to be mediated by
excess energy intake. In contrast, 2 other studies,34,35 one
using diet records and another using the Frequency Food
Questionnaire, found that higher dairy intake protected
young adults from excess weight gain.
The present study may be the largest long-term study
using diet records that has shown a protective effect of
dairy intake on weight gain. Dairy may act to decrease
weight gain through a variety of possible mechanisms,
including an association with a healthier diet in general;
its protein content has been shown to increase satiety.36
Higher intakes of whole grains were associated with
decreased BMI gains during study follow-up. Although
there are few studies on grain intake and weight in children, the present finding is in line with other studies37-40
showing grain to be protective. Although the level of whole
grain intake across this study population was low and
well below the target threshold of 50% of total grains, it
was nonetheless protective. Whole grain intake may result in less weight gain via its higher fiber content and,
thus, higher satiety or as a marker for a healthier lifestyle, something we may not have been able to completely capture in the multivariate models.
The study strengths include use of a large socioeconomically and geographically diverse sample that incorporates more than 50% black girls, a population particularly beset by the obesity epidemic. An additional strength
is the availability of detailed dietary information that allows us to examine the change in BMI in relation to food
group exposure, a method that has seen little study in the
adolescent literature so far. Finally, the use of repeated measures collected in a longitudinal manner over 9 years of
study increased power substantially and likely decreased
random variation in exposure and covariate data.
A limitation of this study is the low level of intake in
certain food groups that may have restricted our ability
to detect true beneficial effects of these food groups. In
addition, there are food groups that other studies have
found to be important predictors of obesity, such as sugarsweetened beverages that are not a part of the DASH eating pattern and unavailable in this analysis, that may be
important predictors of excess weight gain.41,42 Finally,
it is possible that there is biased reporting for some dietary factors, especially in obese individuals, that could
affect the results of these analyses. However, the longitudinal nature of the study and the multiple measures
of dietary intake beginning in preadolescence suggest that
this explanation for these findings is unlikely. In conclusion, greater consistency with the DASH eating plan
resulted in lower excess weight gains in girls from early
adolescence to young adulthood.
Accepted for Publication: December 23, 2010.
Correspondence: Jonathan P. B. Berz, MD, MSc, Section of General Internal Medicine, Boston University Medical Center, 801 Massachusetts Ave, Second Floor,
Boston, MA 02118 ([email protected]).
Author Contributions: Drs Berz and Moore had full access to all the data in the study and take responsibility
for the integrity of the data and the accuracy of the data
analysis. Study concept and design: Berz and Moore. Acquisition of data: Singer, Daniels, and Moore. Analysis and
interpretation of data: Berz, Guo, and Moore. Drafting of
the manuscript: Berz and Guo. Critical revision of the manuscript for important intellectual content: Berz, Singer, Daniels, and Moore. Statistical analysis: Guo and Moore. Obtained funding: Daniels and Moore. Administrative,
technical, and material support: Singer and Moore. Study
supervision: Moore.
Financial Disclosure: None reported.
Funding/Support: This work was supported by grant
5R21DK075068 from the National Institute of Diabetes
and Digestive and Kidney Diseases.
Role of the Sponsor: No sponsors or funders had any role
in the design and conduct of the study; in the collection, analysis, and interpretation of the data; and in the
preparation, review, or approval of the manuscript.
Additional Contributions: Di Gao, AS, Caroline
Apovian, MD, Gheorghe Doros, PhD, and M. Loring
Bradlee, MS, assisted with statistical analysis, data presentation preparation, and insight.
REFERENCES
1. Chang VW, Lauderdale DS. Income disparities in body mass index and
obesity in the United States, 1971-2002. Arch Intern Med. 2005;165(18):
2122-2128.
2. Hedley AA, Ogden CL, Johnson CL, Carroll MD, Curtin LR, Flegal KM. Prevalence of overweight and obesity among US children, adolescents, and adults,
1999-2002. JAMA. 2004;291(23):2847-2850.
3. Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. Prevalence of overweight and obesity in the United States, 1999-2004. JAMA. 2006;
295(13):1549-1555.
4. van Dam RM, Willett WC, Manson JE, Hu FB. The relationship between overweight in adolescence and premature death in women. Ann Intern Med. 2006;
145(2):91-97.
5. Daniels SR, Arnett DK, Eckel RH, et al. Overweight in children and adolescents:
pathophysiology, consequences, prevention, and treatment. Circulation. 2005;
111(15):1999-2012.
6. Jacobs DR Jr, Tapsell LC. Food, not nutrients, is the fundamental unit in nutrition.
Nutr Rev. 2007;65(10):439-450.
7. Sacks FM, Obarzanek E, Windhauser MM, et al. Rationale and design of the Dietary Approaches to Stop Hypertension trial (DASH). A multicenter controlledfeeding study of dietary patterns to lower blood pressure. Ann Epidemiol. 1995;
5(2):108-118.
8. Appel LJ, Moore TJ, Obarzanek E, et al; DASH Collaborative Research Group.
A clinical trial of the effects of dietary patterns on blood pressure. N Engl J Med.
1997;336(16):1117-1124.
ARCH PEDIATR ADOLESC MED/ VOL 165 (NO. 6), JUNE 2011
545
WWW.ARCHPEDIATRICS.COM
©2011 American Medical Association. All rights reserved.
Downloaded From: http://archfaci.jamanetwork.com/ on 08/03/2017
9. Champagne CM. Dietary interventions on blood pressure: the Dietary Approaches to Stop Hypertension (DASH) trials. Nutr Rev. 2006;64(2, pt 2):S53S56.
10. Moore LL, Singer MR, Bradlee ML, et al. Intake of fruits, vegetables, and dairy
products in early childhood and subsequent blood pressure change. Epidemiology.
2005;16(1):4-11.
11. Moore TJ, Conlin PR, Ard J, Svetkey LP. DASH (Dietary Approaches to Stop Hypertension) diet is effective treatment for stage 1 isolated systolic hypertension.
Hypertension. 2001;38(2):155-158.
12. Folsom AR, Parker ED, Harnack LJ. Degree of concordance with DASH diet guidelines and incidence of hypertension and fatal cardiovascular disease. Am J
Hypertens. 2007;20(3):225-232.
13. Fung TT, Chiuve SE, McCullough ML, Rexrode KM, Logroscino G, Hu FB. Adherence to a DASH-style diet and risk of coronary heart disease and stroke in
women. Arch Intern Med. 2008;168(7):713-720.
14. Mellen PB, Gao SK, Vitolins MZ, Goff DC Jr. Deteriorating dietary habits among
adults with hypertension: DASH dietary accordance, NHANES 1988-1994 and
1999-2004. Arch Intern Med. 2008;168(3):308-314.
15. Gidding SS, Dennison BA, Birch LL, et al; American Heart Association. Dietary
recommendations for children and adolescents: a guide for practitioners [published correction appears in Pediatrics. 2006;118(3):1323]. Pediatrics. 2006;
117(2):544-559.
16. Obesity and cardiovascular disease risk factors in black and white girls: the NHLBI
Growth and Health Study. Am J Public Health. 1992;82(12):1613-1620.
17. Schakel SF, Sievert YA, Buzzard IM. Sources of data for developing and maintaining a nutrient database. J Am Diet Assoc. 1988;88(10):1268-1271.
18. Agricultural Research Service. USDA Food and Nutrient Database for Dietary Studies, 2.0. Beltsville, MD: Agricultural Research Service, Food Surveys Research
Group; 2006.
19. Nutrition and Your Health: Dietary Guidelines for Americans. 5th ed. Washington, DC: US Government Printing Office; 2000.
20. Levitan EB, Wolk A, Mittleman MA. Consistency with the DASH diet and incidence of heart failure. Arch Intern Med. 2009;169(9):851-857.
21. Dietary Guidelines for Americans. 2005. http://www.health.gov/dietaryguidelines
/dga2005/document/pdf/dga2005.pdf. Accessed October 17, 2008.
22. Littell RC, Pendergast J, Natarajan R. Modelling covariance structure in the analysis of repeated measures data. Stat Med. 2000;19(13):1793-1819.
23. Kimm SY, Glynn NW, Kriska AM, et al. Longitudinal changes in physical activity
in a biracial cohort during adolescence. Med Sci Sports Exerc. 2000;32(8):
1445-1454.
24. Harrell JS, McMurray RG, Baggett CD, Pennell ML, Pearce PF, Bangdiwala SI.
Energy costs of physical activities in children and adolescents. Med Sci Sports
Exerc. 2005;37(2):329-336.
25. Ridley K, Olds TS. Assigning energy costs to activities in children: a review and
synthesis. Med Sci Sports Exerc. 2008;40(8):1439-1446.
26. Ridley K, Ainsworth BE, Olds TS. Development of a compendium of energy expenditures for youth. Int J Behav Nutr Phys Act. 2008;5:45.
27. Statistics National Center for Health Statistics. BMI for age charts: CDC growth
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
charts: United States. May 30, 2000. http://www.cdc.gov/nchs/data/series/sr
_11/sr11_246.pdf. Accessed June 1, 2009.
Shin KO, Oh SY, Park HS. Empirically derived major dietary patterns and their
associations with overweight in Korean preschool children. Br J Nutr. 2007;
98(2):416-421.
Quatromoni PA, Copenhafer DL, D’Agostino RB, Millen BE. Dietary patterns predict the development of overweight in women: the Framingham Nutrition Studies.
J Am Diet Assoc. 2002;102(9):1239-1246.
Newby PK, Muller D, Hallfrisch J, Qiao N, Andres R, Tucker KL. Dietary patterns
and changes in body mass index and waist circumference in adults. Am J Clin
Nutr. 2003;77(6):1417-1425.
Phillips SM, Bandini LG, Cyr H, Colclough-Douglas S, Naumova E, Must A.
Dairy food consumption and body weight and fatness studied longitudinally over
the adolescent period. Int J Obes Relat Metab Disord. 2003;27(9):1106-1113.
Dixon LB, Pellizzon MA, Jawad AF, Tershakovec AM. Calcium and dairy intake
and measures of obesity in hyper- and normocholesterolemic children. Obes Res.
2005;13(10):1727-1738.
Berkey CS, Rockett HR, Willett WC, Colditz GA. Milk, dairy fat, dietary calcium,
and weight gain: a longitudinal study of adolescents. Arch Pediatr Adolesc Med.
2005;159(6):543-550.
Moore LL, Bradlee ML, Gao D, Singer MR. Low dairy intake in early childhood
predicts excess body fat gain. Obesity (Silver Spring). 2006;14(6):1010-1018.
Pereira MA, Jacobs DR Jr, Van Horn L, Slattery ML, Kartashov AI, Ludwig DS.
Dairy consumption, obesity, and the insulin resistance syndrome in young adults:
the CARDIA Study. JAMA. 2002;287(16):2081-2089.
Luhovyy BL, Akhavan T, Anderson GH. Whey proteins in the regulation of food
intake and satiety. J Am Coll Nutr. 2007;26(6):704S-712S.
Liu S, Willett WC, Manson JE, Hu FB, Rosner B, Colditz G. Relation between changes
in intakes of dietary fiber and grain products and changes in weight and development of obesity among middle-aged women. Am J Clin Nutr. 2003;78(5):
920-927.
Lutsey PL, Jacobs DR Jr, Kori S, et al. Whole grain intake and its cross-sectional
association with obesity, insulin resistance, inflammation, diabetes and subclinical
CVD: the MESA Study. Br J Nutr. 2007;98(2):397-405.
Newby PK, Peterson KE, Berkey CS, Leppert J, Willett WC, Colditz GA. Dietary
composition and weight change among low-income preschool children. Arch Pediatr Adolesc Med. 2003;157(8):759-764.
McKeown NM, Meigs JB, Liu S, Wilson PWF, Jacques PF. Whole-grain intake is
favorably associated with metabolic risk factors for type 2 diabetes and cardiovascular disease in the Framingham Offspring Study. Am J Clin Nutr. 2002;
76(2):390-398.
Ludwig DS, Peterson KE, Gortmaker SL. Relation between consumption of sugarsweetened drinks and childhood obesity: a prospective, observational analysis.
Lancet. 2001;357(9255):505-508.
Francis LA, Lee Y, Birch LL. Parental weight status and girls’ television viewing,
snacking, and body mass indexes. Obes Res. 2003;11(1):143-151.
Education is the ability to meet life’s
situations.
—Dr John Hibbon, former president of
Princeton
ARCH PEDIATR ADOLESC MED/ VOL 165 (NO. 6), JUNE 2011
546
WWW.ARCHPEDIATRICS.COM
©2011 American Medical Association. All rights reserved.
Downloaded From: http://archfaci.jamanetwork.com/ on 08/03/2017