{"subscriber":false,"subscribedOffers":{}} Consumption Of Specific Foods And Beverages And Excess Weight Gain Among Children And Adolescents | Health Affairs

Consumption Of Specific Foods And Beverages And Excess Weight Gain Among Children And Adolescents

Affiliations
  1. Di Dong is a PhD student in health services and systems research at the Duke–National University of Singapore (NUS) Graduate Medical School.
  2. Marcel Bilger is an assistant professor in the Health Services and Systems Research Program at the Duke-NUS Graduate Medical School.
  3. Rob M. van Dam is an associate professor in the Saw Swee Hock School of Public Health, National University of Singapore.
  4. Eric A. Finkelstein ( [email protected] ) is a professor in the Health Services and Systems Research Program at the Duke-NUS Graduate Medical School and a research professor at the Duke Global Health Institute, Duke University, in Durham, North Carolina.
PUBLISHED:Free Accesshttps://doi.org/10.1377/hlthaff.2015.0434

Abstract

Efforts are under way to identify successful strategies to reduce long-term childhood obesity risk, such as ways to improve diet quality. To identify foods and beverages associated with excess weight gain, we used cohort data from the Avon Longitudinal Study of Parents and Children in the United Kingdom. We quantified the associations between changes in or levels of consumption of twenty-seven food and beverage groups and excess weight gain in three-year periods among youth ages 7–13. When we considered all dietary factors and physical activity levels simultaneously, we found that foods with the largest positive associations with three-year excess weight gain were fat spread (butter or margarine), coated (breaded or battered) poultry, potatoes cooked in oil (French fries, roasted potatoes, and potato chips), coated fish, processed meats, other meats, desserts and sweets, milk, and sugar-sweetened beverages. Foods associated with weight loss were whole grains and high-fiber cereals. These results provide evidence for targeting specific food and beverage groups in efforts to influence weight outcomes.

TOPICS

The childhood obesity epidemic is a global public health concern. More than 31 percent of children in the Americas, 38 percent in Europe, 27 percent in the western Pacific, and 22 percent in Southeast Asia are now overweight or obese. 1,2 Childhood obesity persists into adulthood and is associated with higher risk of hypertension, type 2 diabetes, cancer, and cardiovascular diseases; reduced life expectancy; and greater lifetime medical expenditures. 35 As a result, efforts are under way to identify successful long-term risk reduction strategies, including efforts to modify dietary intake.

Dariush Mozaffarian and coauthors analyzed the intake of thirteen groups of foods and beverages among 120,877 adults in the United States and found that within each four-year period studied, those who increased their daily servings of potato chips, potatoes, sugar-sweetened beverages, red meats, and processed meats had the greatest weight gain, whereas those who increased their daily servings of vegetables, whole grains, fruit, yogurt, and nuts had the most weight loss. 6 To date, no similar study has been conducted among youth.

Children’s metabolism, dietary patterns, and energy expenditure differ from those of adults, and children are expected to gain weight as they age. Therefore, it is not clear whether foods strongly associated with excess weight gain or loss among adults would have the same effects among youth. Addressing this question is the primary goal of this study.

Using a unique longitudinal data set that tracked food consumption and children’s weight and height over time, we quantified the extent to which changes in a child’s body mass index (BMI) z-score (a measure of relative adiposity in children that is described in more detail below) over time was associated with changes in the consumption of specific food and beverage groups over the same time interval (the change-change model) and with the average levels of consumption of these foods and beverages (the change-level model). The change-change model is designed to explore whether those who increase their consumption of sugar-sweetened beverages, for example, regardless of their consumption at the start of the time interval, show greater weight gain than those who do not increase their consumption of the beverages. In contrast, the change-level model is designed to explore whether those who consume larger amounts of the beverages show greater weight gain.

Previous studies have typically focused on the change-change model. 68 However, that model makes the simplifying assumption that a given change will have the same influence on weight regardless of the underlying levels of consumption. This may or may not be true in reality. It likely depends on whether a person is at his or her steady-state weight—that is, weight at which only increased energy intake can lead to weight gain. Given their levels of growth and development, most youth are unlikely to be at their steady-state weight, which suggests that the change-level model may be more appropriate than the change-change model for research on youth. 9 Regardless, to reveal the impact of both levels of consumption and change in consumption on excess weight gain, both models were used and expected to show results similar to those in the article by Mozaffarian and coauthors. 6

Study Data And Methods

Survey Data

The data came from a study conducted in the United Kingdom: the Avon Longitudinal Study of Parents and Children. The study is an ongoing prospective birth-cohort study whose primary aim is to investigate the influence of genetic and environmental factors on health and development during childhood and beyond. 10 This cohort consisted of 15,444 children born in 1991 and 1992 in the area centered on the city of Bristol, in southwest England. 11

The study provides comprehensive information on dietary intake and body weight and has several desirable features. First, food intake was recorded using a three-day food diary, which is a commonly accepted dietary assessment method that records both consumption frequencies and quantities. Second, height and weight were measured in clinics, which reduced potential errors associated with self-reporting. And third, physical activity was measured using an accelerometer, which can precisely measure the duration and intensity of physical activity, at ages eleven and thirteen.

The 4,646 children who completed a three-day food diary and attended research clinics to have their height and weight measured at ages seven, ten, and thirteen were included in our analysis. 12 More details about the sample are in the online Appendix. 13

Dependent Variable

Height and weight were measured during clinic visits around ages seven, ten, and thirteen. 12 The relationship between BMI and adiposity may differ according to age and sex. 14,15 We therefore used age- and sex-adjusted BMI z-scores as a measure for adiposity and change in adiposity over time. 2,16 We calculated the BMI z-score for each child as the number of standard deviation units that a child’s BMI was away from the mean BMI on the 1990 UK age- and sex-specific growth reference charts. 14 An increase in BMI z-score over a three-year period indicates an increase in adiposity compared to the average adiposity of the reference cohort—in other words, it indicates excess weight gain.

Independent Variables

At ages seven, ten, and thirteen, children recorded their food consumption using a diary for three days with the help of their parent or guardian. 12 Intakes of over a hundred commonly consumed foods and beverages were quantified in grams consumed per day. Fish, poultry, and potatoes were divided into subcategories based on the cooking method, since this was expected to influence calories and nutrient quality. We classified the items reported in the diaries into twenty-seven groups (see Appendix Exhibit A1) 13 and summed intakes within each group.

We then calculated change in intake (for the change-change model) as the difference between the intake at the end and at the start of each three-year interval, and we calculated levels of intake (for the change-level model) as the average consumption based on the reported consumption at the start and end of each interval. Both change and levels of consumption were modeled as continuous variables.

Control Variables

We controlled for two other influential factors on adiposity among youth: physical activity and puberty status. Minutes per day of physical activity of moderate-to-vigorous intensity is a commonly used measure for physical activity. It takes into account both the duration and the intensity of activity, and it has been shown to be correlated with weight change. We calculated minutes per day of this activity from children’s accelerometer outputs at ages eleven and thirteen. 17 These data were not available at age seven. Correlations between early puberty and obesity have also been reported. 18,19 Therefore, we accounted for puberty status by including the self-reported Tanner stage at ages eight and thirteen in the analyses. 18,19

We also included the highest education level for each child’s mother, to control for the effect of unobserved differences in socioeconomic factors on food intake, such as food quality and eating habits. Several other socioeconomic variables were available in the data set, such as mother’s class (a classification based on occupation), father’s class, and housing status. We did not include them in our regression models because once mother’s education was included, the inclusion of these variables had a negligible influence on the results.

Statistical Analyses

Two multivariate linear regressions models were used. As explained above, the change-change model estimated the association between the change in BMI z-score in a three-year interval with changes in food and beverage consumption within the same period. The change-level model analyzed the correlation between changes in BMI z-score in a three-year interval and the average level of consumption over that period, based on the diary data. For additional details on the model specifications, see the Appendix. 13

Data from the two three-year periods (ages 7–10 and 10–13) and by sex were pooled in the regression models, since we expected the relationship between dietary intake and excess weight gain to be similar across age and sex groups. Robust variance estimation was used to account for repeated observations from individuals over time. As a sensitivity analysis, we also explored whether there were differential effects by age or sex group (see Appendices A2 and A3). 13

To allow for easier interpretation of the results, we used the regression coefficients to estimate excess weight gain for each food or beverage group based on the assumption that a 0.01 increase in BMI z-score corresponded to a 50 gram increase in weight. For both models, we report the three-year excess weight gain due to a 100 gram per day increase in intake and the gain due to a one serving per day increase in intake over three years. We assumed that the serving size for each food or beverage was equal to the portion size for school meals recommended by the Scottish government. 20

All statistical analyses were conducted using Stata, version 11.2.

Strengths And Limitations

This study had several strengths, including a large cohort, height and weight measured in clinics, validated food diaries, the availability of accelerometer-measured physical activity, the use of different models to examine the effect of both changes in and levels of dietary intake on excess weight gain, and the evaluation of all major food and beverage groups simultaneously to assess the relative effects on weight gain. However, it also had limitations.

First, as with any dietary assessment, measurement error is a significant concern. Portion size was roughly documented in the diaries, and food intake in grams per day was imputed. There are likely to be individual variations in portion size and eating habits. However, measurement error would bias our estimates toward no effect, which suggests that measurement error is not responsible for the observed associations. Moreover, the change-level model considers average consumption levels across time and is less influenced by random measurement errors in self-reported food consumption, compared to the change-change model.

Second, pubertal stages and physical activity were not measured at the same time as height, weight, and dietary intake. However, the data were collected within a year, and we expect the puberty status and physical activity levels at the time of assessment to be close to the levels at the time the other data were collected.

Third, as is the case in all observational studies, we could not completely address residual confounding by other unobserved lifestyle factors. Nor could we exclude the possibility of reverse causality, in which body weight influences food choice. Randomized controlled trials would provide the best evidence for causality and should be implemented to confirm our results.

A fourth limitation concerns the generalizability of our results. We chose the Avon Longitudinal Study of Parents and Children data set because of the study’s prospective cohort design and high-quality measurements of dietary intake, height, and weight. However, there is potential selection bias resulting from loss of follow-up on dietary and clinic assessments for some participants. Our results reflect the average effect of dietary intake on excess weight gain among those participants with complete data in the cohort. If consumption patterns or dietary habits differed greatly in other countries or populations, our results might not be generalizable to them. Nevertheless, our findings were generally consistent with those of studies in other Western youth and adult populations. 6,7,21

Study Results

Characteristics

Of the 4,646 individuals analyzed in this study, the mean BMI was 16.2 at seven years and six months, 18.2 at ten years and seven months, and 20.2 at thirteen years and nine months, based on the height and weight measurements and ages at the time of measurement (data not shown). The characteristics by sex are shown in Exhibit 1 . The mean change in BMI z-score was 0.18 units from age seven to age ten and 0.02 units from age ten to age thirteen (data not shown).

Exhibit 1 Characteristics Of A Sample Of 4,646 Children From The Avon Longitudinal Study Of Parents And Children

Age 7
Age 10
Age 13
Boys
Girls
Boys
Girls
Boys
Girls
CharacteristicMeanSDMeanSDMeanSDMeanSDMeanSDMeanSD
Age7 y 6 m4 m7 y 6 m4 m10 y 7 m3 m10 y 7 m3 m13 y 10 m3 m13 y 10 m3 m
Weight (kg)25.74.325.74.737.48.038.18.654.311.654.310.5
Height (cm)126.25.5125.45.6143.86.3143.96.8164.98.6162.06.3
BMI16.01.916.32.118.03.018.33.219.83.320.63.5
BMI z-score0.11.00.11.00.41.10.21.20.31.10.31.1

SOURCE Authors’ analysis of data from the Avon Longitudinal Study of Parents and Children (see Note  10 in text). NOTES SD is standard deviation. Body mass index (BMI) z-score is the number of standard deviation units that a child’s BMI varies from the mean BMI on the 1990 UK age- and sex-specific growth reference charts (see Note  14 in text). Y is years. M is months.

Dietary Intake And Physical Activity

Measured in grams per day, the solid foods with the highest daily intake at age seven years were desserts and sweets (including sweet biscuits, or cookies; pudding; cake; ice cream; and confectionery, or candy), refined grains, vegetables, fruit, potatoes cooked in oil (French fries and roasted potatoes), meats, potatoes not cooked in oil (referred to in the study as “boiled/mash/jacket potato”), yogurt, and uncoated (neither breaded nor battered) poultry ( Exhibit 2 ).

Exhibit 2 Average Dietary Intake And Physical Activity Levels Of Children In The Sample, By Age

Daily intake (g/day)
Age 7
Age 10
Age 13
Food or beverageMeanSDMeanSDMeanSD
Diet soda133.3163.8101.6155.0102.7175.5
Full-fat milk133.3182.074.0146.748.3138.4
Desserts and sweets131.562.7128.869.0107.579.1
Low-fat milk125.4167.1144.6170.7168.5195.1
Refined grains110.961.193.676.3115.498.6
Juices94.4138.1124.0154.7164.7208.2
Sugar-sweetened beverages82.2132.8106.9155.9131.9212.8
Fruit79.977.568.673.976.291.5
Vegetables74.352.590.268.5105.283.6
French fries and roasted potatoes46.939.458.151.257.161.4
Yogurt37.346.334.350.527.453.6
Meats34.637.748.052.252.867.2
Potatoes not cooked in oil30.235.434.044.339.354.6
Processed meats17.220.819.623.618.629.0
Potato chips17.113.218.515.215.016.2
High-fiber breakfast cereal17.024.215.124.919.131.6
Uncoated poultry16.323.523.533.433.646.8
Non-high-fiber breakfast cereal15.217.712.416.511.119.1
Whole grains11.614.912.817.215.721.1
Fat spread11.423.712.026.522.236.8
Cheese11.48.113.09.511.810.8
Coated poultry10.318.510.121.37.422.6
Coated fish7.815.26.817.16.018.4
Eggs7.615.69.019.29.623.3
Uncoated fish6.216.77.118.59.926.0
Canned fruit2.913.03.315.71.912.3
Nuts1.85.51.76.01.86.9
Physical activity (min/day)Age 11Age 13
Mean moderate to vigorous activity a22.915.424.117.2

SOURCE Authors’ analysis of data from the Avon Longitudinal Study of Parents and Children (see Note  10 in text). NOTES “Uncoated” is neither breaded nor battered. “Fat spread” is butter or margarine. SD is standard deviation.

a Among those with available physical activity data ( n = 3,790 for age 11; n = 3,337 for age 13).

Children at all ages consumed a significant amount of beverages other than water. Yet the types of beverages consumed changed with time. At age seven, the most highly consumed beverages were full-fat milk and diet soda, whereas at age thirteen, low-fat milk, juices, and sugar-sweetened beverages were the most commonly consumed drinks.

The average amount of time per day of physical activity of moderate-to-vigorous intensity was 22.9 minutes at age eleven and 24.1 minutes at age thirteen ( Exhibit 2 ).

Excess Weight Gain And Food And Beverage Consumption

In the change-change model, three-year excess weight gain was significantly positively associated with increased intake of fat spread (butter and margarine), coated poultry, potato chips, coated fish, processed meats, French fries and roasted potatoes, meats, desserts and sweets, full-fat and low-fat milk, and sugar-sweetened beverages ( Exhibit 3 ). Increased intake of whole grains and high-fiber breakfast cereal was significantly associated with weight loss. Increases in all other dietary factors were not significantly associated with weight change at the observed levels of changes in consumption levels.

Exhibit 3 Associations Between Estimated Three-Year Excess Weight Gain And Intake Of Each Food And Beverage Among Children In The Sample

Food or beverageEstimated 3-year excess weight gain (g) per 100 g increase in daily intake (change-change model)Estimated 3-year excess weight gain (g) per 100 g daily intake (change-level model)
Fat spread 1,245 *** 1,230 ***
Nuts495 1,145 *
Coated poultry 380 ***305
Potato chips 375 ** 930 ***
Coated fish 290 ** 655 ***
Processed meats 240 ** 555 ***
Uncoated fish140210
French fries and roasted potatoes 135 ** 200 **
Meats 115 **50
Uncoated poultry105 −275 **
Canned fruit100−340
Eggs85−75
Desserts and sweets 65 * 140 **
Full-fat milk 65 *** −40 *
Low-fat milk 40 ** −40 *
Sugar-sweetened beverages 35 **−20
Fruit2515
Potatoes not cooked in oil15−75
Juices10−15
Diet soda1030
Vegetables0 −100 *
Yogurt−5−65
Cheese−40−115
Refined grains−50 145 ***
Non-high-fiber breakfast cereal−175 435 *
Whole grains −260 *** −380 ***
High-fiber breakfast cereal −260 **−195

SOURCE Authors’ analysis of data from the Avon Longitudinal Study of Parents and Children (see Note  10 in text). NOTES All dietary intake was evaluated simultaneously. Levels of moderate-to-vigorous activity and pubertal stages were included in the regression to control for their influence on adiposity, but coefficients are not shown here because of differences in measurement units. Weight gain was estimated assuming that a 0.01 increase in body mass index (BMI) z-score (explained in the notes to Exhibit 1 ) corresponded to 50 grams of excess weight gain. The change-change model (left column) shows associations between changes in consumption of food and beverages and changes in a child’s BMI z-score over the same time period. The change-level model (right column) shows associations between average levels of consumption of the food and beverages and changes in BMI z-score in the same period. Data from the two three-year periods (ages 7–10 and ages 10–13) were pooled in the regression, with robust variance estimated to account for repeated observations from the same individual. “Uncoated” is neither breaded nor battered. “Fat spread” is butter or margarine.

*p<0.10

**p<0.05

***p<0.01

Moderate-to-vigorous physical activity was associated with weight loss. A thirty-minute increase per day in such activity in a three-year interval was associated with a significant ( p<0.01 ) 0.39 kg reduction in weight.

As in the change-change model, in the change-level model excess weight gain was significantly positively associated with intake of fat spread, potato chips, coated fish, processed meats, French fries and roasted potatoes, and desserts and sweets ( Exhibit 3 ). In addition, in the change-level model excess weight gain was significantly positively associated with increased intake of nuts, refined grains, and low-fiber breakfast cereal. And in the change-level model, higher intake of whole grains, uncoated poultry, vegetables, and full-fat and low-fat milk was significantly associated with weight loss. All other foods were not significantly associated with excess weight gain at the observed levels of consumption.

There are different serving size conventions for various foods and beverages. 20 To take into account variations in serving sizes, serving size–adjusted excess weight gains based on the change-change and change-level models are shown in Exhibit 4 . After we adjusted for portion sizes, we found that the foods and beverages most associated with excess weight gain were coated poultry and fish, processed meats, nuts, and French fries and roasted potatoes.

Exhibit 4 Associations Between Estimated Three-Year Excess Weight Gain And Intake Among Children In The Sample, By Type Of Food And Beverage, Adjusted For Serving Size

Exhibit 4
SOURCE Authors’ analysis of data from the Avon Longitudinal Study of Parents and Children (see Note  10 in text). NOTES All dietary intake was evaluated simultaneously. Weight gain was estimated assuming that a 0.01 increase in body mass index (BMI) z-score (explained in the notes to Exhibit 1 ) corresponded to 50 grams of excess weight gain. The change-change model shows associations between changes in a child’s BMI z-score and change in consumption of foods and beverages over the same time period. The change-level model shows associations between BMI z-score and average levels of consumption of the foods and beverages in the same period. Serving sizes were based on recommendations for ten-year-old children made by the Scottish Government (see Note  20 in text). “Uncoated” is neither breaded nor battered. “Fat spread” is butter or margarine. * p<0.10 ** p<0.05 *** p<0.01

Discussion

Main Findings

This study examined the association between excess weight gain and changes in or levels of consumption of twenty-seven food and beverage groups. Although the two models tested different hypotheses, the results were generally consistent across models. Both models showed the consumption of fat spread, potato chips, coated fish, processed meats, French fries and roasted potatoes, and desserts and sweets to be associated with excess weight gain and the consumption of whole grains and greater levels of physical activity to be associated with weight loss. When differential serving sizes were accounted for, those foods consumed in larger portions—such as coated poultry and fish, processed meats, and French fries and roasted potatoes—showed larger associations with excess weight gain, compared to foods consumed in smaller portions.

The relationship between the consumption of specific foods and beverages and excess weight gain is complex. Energy density (calories in a given weight of food) is one factor that is commonly considered to be a contributor to overconsumption. 2225 Indeed, our results generally show that compared to foods with a lower energy density, those with a higher energy density are associated with greater weight gain.

However, we cannot exclude the possibility that other characteristics of these foods are responsible for some of the effects. Emerging evidence has also highlighted the importance of diet composition on weight change, regardless of calories. 2630 For example, liquid calories have been shown to be more obesity-promoting than calories from solid food, 31,32 which may be because beverages are less satiating than food. 33 Food that is highly satiating may generate weight loss by reducing subsequent calorie intake. 34

Foods of high glycemic index have also been associated with excess weight gain. The possible mechanisms involved include inducing a faster rise in glucose level and insulin response, compared to foods with lower glycemic index; altering lipid metabolism; stimulating brain regions associated with reward and craving; and changing resting energy expenditures. 2830

In our study, the foods negatively associated with excess weight gain were generally high in protein and fiber—two dietary components that increase satiety and contribute to a lower glycemic index. 35 More research is needed to better quantify the relative influence of each of these factors on weight outcomes among children and adults, as well as to clarify the underlying physiological mechanisms.

Policy-Relevant Foods And Beverages

Certain potato-based foods, including French fries and potato chips, are increasingly being targeted by policy makers as foods to be restricted in school lunch programs and other venues because of their potential obesity-promoting effects. 3638 Our results and those from other studies 6,8,3941 support these policies.

In fact, we found potato chips to be one of the most obesity-promoting foods for youth to consume. Potato chips are very high in energy density (383–574 kcal/100g) and have a low satiety index, yet they are commonly consumed as snacks. 35,42 Our results showed that an increased intake of potatoes not cooked in oil—in contrast to French fries and potato chips—was not significantly associated with excess weight gain. Potatoes cooked without oil and with no dressing added have a low energy density (63–104 kcal/100g) and are highly satiating. In a field experiment, Susanna Holt and coauthors found that boiled or mashed potatoes were the most filling food, with a calculated satiety index three times that of white bread. 35 In addition, potatoes are high in potassium, vitamins, and other essential nutrients. 43

Therefore, targeting all potato-based foods may not be appropriate. Based on our results, a policy that encourages switching to healthier methods of preparing and consuming potatoes, perhaps in school meals programs, should be considered.

Like Mozaffarian and coauthors 6 and other researchers, 7,44 we found a positive correlation between excess weight gain and intake of foods with added sugar (desserts, sweets, and sugar-sweetened beverages), which supports policies to limit the consumption of these foods among youth. 37,45,46 Compared to youth in the United Kingdom, young people in the United States consume more sugar-sweetened beverages. Therefore, reducing the consumption of these beverages in the United States is likely to have a greater positive impact on weight outcomes than in the United Kingdom.

Many studies have shown an association between greater meat consumption and weight gain. 6,44 However, some have shown an inverse association between that consumption and waist circumference, which is another indicator for body fat distribution and obesity status. 7,41 Nonetheless, the different types of meats and different cooking methods were often not separated in the studies, and these differences can lead to large variations in nutrient quality.

In our study, the consumption of processed meats and coated fish and poultry was found to be positively associated with excess weight gain, whereas the consumption of uncoated poultry and fish was inversely or not significantly associated with that gain. Processed meat and coated fish and poultry are high in fat and energy and may be less satiating, compared to the less-processed or uncoated versions. 35,47 In contrast, uncoated poultry and fish are high in protein, which may be protective against obesity. 30,48 This suggests that reducing the coating of meat, poultry, and fish may be effective in reducing weight gain.

Consistent with many other studies, 6,7,49,50 we found positive correlations between weight gain and increased intake of refined high-carbohydrate foods (refined grains and desserts) and an inverse association between weight gain and increased intake of whole grains and high-fiber breakfast cereal. Fiber can slow digestion and absorption, which may not only augment satiety and reduce subsequent caloric intake but can also result in lower insulin and glucose responses, which in turn may favor the use instead of the storage of fat. 35,49

In 2012 the US Department of Agriculture’s Nutrition Standards for National School Meals recommended increased offerings of whole grains. 51 Similar recommendations have been made by the Department for Education in England 37 and the Ministry of Education in Ontario, Canada. 38 Our results provide additional evidence in support of these and similar recommendations.

The success of interventions designed to reduce young people’s intake of obesity-promoting foods will depend on what they eat instead of the displaced foods. A strategy that encourages them to eat foods that are associated with weight loss, such as whole grains, uncoated poultry, and vegetables may be effective in controlling hunger and restricting total energy intake.

Conclusion

In summary, our findings support policies that aim to reduce the intake of fat spread, potato chips, French fries and roasted potatoes, sugar-sweetened beverages, desserts and sweets, and refined grains, since these foods have a sizable influence on three-year excess weight gain and are major dietary components for youth in Western countries. Our results also support efforts to change methods of cooking and processing food by reducing oil and coatings. For potatoes, poultry, and fish, simply changing the cooking methods may be an effective strategy to reduce calorie intake and weight gain.

ACKNOWLEDGMENTS

The authors thank the Avon Longitudinal Study of Parents and Children team for sharing data.

NOTES

  • 1 Ogden CL , Carroll MD , Kit BK , Flegal KM . Prevalence of childhood and adult obesity in the United States, 2011–2012 . JAMA . 2014 ; 311 ( 8 ): 806 – 14 . Crossref, MedlineGoogle Scholar
  • 2 Han JC , Lawlor DA , Kimm SY . Childhood obesity . Lancet . 2010 ; 375 ( 9727 ): 1737 – 48 . Crossref, MedlineGoogle Scholar
  • 3 Finkelstein EA , Graham WCK , Malhotra R . Lifetime direct medical costs of childhood obesity . Pediatrics . 2014 ; 133 ( 5 ): 854 – 62 . Crossref, MedlineGoogle Scholar
  • 4 Franks PW , Hanson RL , Knowler WC , Sievers ML , Bennett PH , Looker HC . Childhood obesity, other cardiovascular risk factors, and premature death . N Engl J Med . 2010 ; 362 ( 6 ): 485 – 93 . Crossref, MedlineGoogle Scholar
  • 5 Reilly JJ , Kelly J . Long-term impact of overweight and obesity in childhood and adolescence on morbidity and premature mortality in adulthood: systematic review . Int J Obes (Lond) . 2010 ; 35 ( 7 ): 891 – 8 . Crossref, MedlineGoogle Scholar
  • 6 Mozaffarian D , Hao T , Rimm EB , Willett WC , Hu FB . Changes in diet and lifestyle and long-term weight gain in women and men . N Engl J Med . 2011 ; 364 ( 25 ): 2392 – 404 . Crossref, MedlineGoogle Scholar
  • 7 Fogelholm M , Anderssen S , Gunnarsdottir I , Lahti-Koski M . Dietary macronutrients and food consumption as determinants of long-term weight change in adult populations: a systematic literature review . Food Nutr Res . 2012 ; 56 . CrossrefGoogle Scholar
  • 8 Halkjaer J , Sørensen TI , Tjønneland A , Togo P , Holst C , Heitmann BL . Food and drinking patterns as predictors of 6-year BMI-adjusted changes in waist circumference . Br J Nutr . 2004 ; 92 ( 4 ): 735 – 48 . Crossref, MedlineGoogle Scholar
  • 9 Hall KD , Butte NF , Swinburn BA , Chow CC . Dynamics of childhood growth and obesity: development and validation of a quantitative mathematical model . Lancet Diabetes Endocrinol . 2013 ; 1 ( 2 ): 97 – 105 . Crossref, MedlineGoogle Scholar
  • 10 Golding J , Pembrey M , Jones R . ALSPAC—the Avon Longitudinal Study of Parents and Children. I. Study methodology . Paediatr Perinat Epidemiol . 2001 ; 15 ( 1 ): 74 – 87 . Crossref, MedlineGoogle Scholar
  • 11 Boyd A , Golding J , Macleod J , Lawlor DA , Fraser A , Henderson J , et al. Cohort profile: the “children of the 90s”—the index offspring of the Avon Longitudinal Study of Parents and Children . Int J Epidemiol . 2013 ; 42 ( 1 ): 111 – 27 . Crossref, MedlineGoogle Scholar
  • 12 Ness AR . The Avon Longitudinal Study of Parents and Children (ALSPAC)—a resource for the study of the environmental determinants of childhood obesity . Eur J Endocrinol . 2004 ; 151 ( Suppl 3 ): U141 – 9 . Crossref, MedlineGoogle Scholar
  • 13 To access the Appendix, click on the Appendix link in the box to the right of the article online.
  • 14 Cole TJ , Freeman JV , Preece MA . Body mass index reference curves for the UK, 1990 . Arch Dis Child . 1995 ; 73 ( 1 ): 25 – 9 . Crossref, MedlineGoogle Scholar
  • 15 Ellis KJ . Human body composition: in vivo methods . Physiol Rev . 2000 ; 80 ( 2 ): 649 – 80 . Crossref, MedlineGoogle Scholar
  • 16 Kipping RR , Jago R , Lawlor DA . Obesity in children. Part 1: epidemiology, measurement, risk factors, and screening . BMJ . 2008 ; 337 : a1824 . Crossref, MedlineGoogle Scholar
  • 17 Mattocks C , Leary S , Ness A , Deere K , Saunders J , Tilling K , et al. Calibration of an accelerometer during free‐living activities in children . Int J Pediatr Obes . 2007 ; 2 ( 4 ): 218 – 26 . Crossref, MedlineGoogle Scholar
  • 18 Kaplowitz PB , Slora EJ , Wasserman RC , Pedlow SE , Herman-Giddens ME . Earlier onset of puberty in girls: relation to increased body mass index and race . Pediatrics . 2001 ; 108 ( 2 ): 347 – 53 . Crossref, MedlineGoogle Scholar
  • 19 Ahmed ML , Ong KK , Dunger DB . Childhood obesity and the timing of puberty . Trends Endocrinol Metab . 2009 ; 20 ( 5 ): 237 – 42 . Crossref, MedlineGoogle Scholar
  • 20 Scottish Government . Hungry for success—a whole school approach to school meals in Scotland [Internet]. Edinburgh : Scottish Government ; [updated 2006 Mar 22; cited 2015 Sep 8 ]. Available from: http://www.scotland.gov.uk/Publications/2003/02/16273/17578 Google Scholar
  • 21 Berkey CS , Rockett HR , Field AE , Gillman MW , Colditz GA . Sugar‐added beverages and adolescent weight change . Obes Res . 2004 ; 12 ( 5 ): 778 – 88 . Crossref, MedlineGoogle Scholar
  • 22 Bes-Rastrollo M , van Dam RM , Martinez-Gonzalez MA , Li TY , Sampson LL , Hu FB . Prospective study of dietary energy density and weight gain in women . Am J Clin Nutr . 2008 ; 88 ( 3 ): 769 – 77 . Crossref, MedlineGoogle Scholar
  • 23 Ledikwe JH , Blanck HM , Kettel Khan L , Serdula MK , Seymour JD , Tohill BC , et al. Dietary energy density is associated with energy intake and weight status in US adults . Am J Clin Nutr . 2006 ; 83 ( 6 ): 1362 – 8 . Crossref, MedlineGoogle Scholar
  • 24 Rolls BJ . Energy density of foods influences satiety & total caloric intake: consumption of foods low in energy density may facilitate healthy eating & weight loss [Internet]. Silver Spring (MD) : Obesity Society ; 2014 Sep 17 [cited 2015 Sep 8 ]. (Position Statement). Available from: http://www.obesity.org/publications/energy-density-of-foods-influences-satiety-a-total-caloric-intake.htm Google Scholar
  • 25 Ledikwe JH , Rolls BJ , Smiciklas-Wright H , Mitchell DC , Ard JD , Champagne C , et al. Reductions in dietary energy density are associated with weight loss in overweight and obese participants in the PREMIER trial . Am J Clin Nutr . 2007 ; 85 ( 5 ): 1212 – 21 . Crossref, MedlineGoogle Scholar
  • 26 Ebbeling CB , Swain JF , Feldman HA , Wong WW , Hachey DL , Garcia-Lago E , et al. Effects of dietary composition on energy expenditure during weight-loss maintenance . JAMA . 2012 ; 307 ( 24 ): 2627 – 34 . Crossref, MedlineGoogle Scholar
  • 27 Browning JD , Baker JA , Rogers T , Davis J , Satapati S , Burgess SC . Short-term weight loss and hepatic triglyceride reduction: evidence of a metabolic advantage with dietary carbohydrate restriction . Am J Clin Nutr . 2011 ; 93 ( 5 ): 1048 – 52 . Crossref, MedlineGoogle Scholar
  • 28 Lennerz BS , Alsop DC , Holsen LM , Stern E , Rojas R , Ebbeling CB , et al. Effects of dietary glycemic index on brain regions related to reward and craving in men . Am J Clin Nutr . 2013 ; 98 ( 3 ): 641 – 7 . Crossref, MedlineGoogle Scholar
  • 29 Ludwig DS , Friedman MI . Increasing adiposity: consequence or cause of overeating ? JAMA . 2014 ; 311 ( 21 ): 2167 – 8 . Crossref, MedlineGoogle Scholar
  • 30 Smith JD , Hou T , Ludwig DS , Rimm EB , Willett W , Hu FB , et al. Changes in intake of protein foods, carbohydrate amount and quality, and long-term weight change: results from 3 prospective cohorts . Am J Clin Nutr . 2015 ; 101 ( 6 ): 1216 – 24 . Crossref, MedlineGoogle Scholar
  • 31 Chen L , Appel LJ , Loria C , Lin PH , Champagne CM , Elmer PJ , et al. Reduction in consumption of sugar-sweetened beverages is associated with weight loss: the PREMIER trial . Am J Clin Nutr . 2009 ; 89 ( 5 ): 1299 – 306 . Crossref, MedlineGoogle Scholar
  • 32 Hu FB , Malik VS . Sugar-sweetened beverages and risk of obesity and type 2 diabetes: epidemiologic evidence . Physiol Behav . 2010 ; 100 ( 1 ): 47 – 54 . Crossref, MedlineGoogle Scholar
  • 33 Drewnowski A , Bellisle F . Liquid calories, sugar, and body weight . Am J Clin Nutr . 2007 ; 85 ( 3 ): 651 – 61 . Crossref, MedlineGoogle Scholar
  • 34 Rolls BJ . Plenary lecture 1: dietary strategies for the prevention and treatment of obesity . Proc Nutr Soc . 2010 ; 69 ( 1 ): 70 – 9 . Crossref, MedlineGoogle Scholar
  • 35 Holt SH , Miller JB , Petocz P , Farmakalidis E . A satiety index of common foods . Eur J Clin Nutr . 1995 ; 49 ( 9 ): 675 – 90 . MedlineGoogle Scholar
  • 36 Cullen KW , Watson KB . The impact of the Texas Public School Nutrition Policy on student food selection and sales in Texas . Am J Public Health . 2009 ; 99 ( 4 ): 706 – 12 . Crossref, MedlineGoogle Scholar
  • 37 UK Department for Education . School food in England: departmental advice for governing bodies [Internet]. Manchester : The Department ; [revised 2014 Jul; cited 2015 Sep 8 ]. Available from: https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/344684/School_food_in_England-_June_2014-_revised_August_14.pdf Google Scholar
  • 38 Ontario Ministry of Education . School food and beverage policy: quick reference guide [Internet]. Toronto : The Ministry ; 2010 [cited 2015 Sep 8 ]. Available from: http://www.edu.gov.on.ca/eng/healthyschools/PPM150Quick_Reference_Guide_2010.pdf Google Scholar
  • 39 Pereira MA , Kartashov AI , Ebbeling CB , Van Horn L , Slattery ML , Jacobs DR , et al. Fast-food habits, weight gain, and insulin resistance (the CARDIA study): 15-year prospective analysis . Lancet . 2005 ; 365 ( 9453 ): 36 – 42 . Crossref, MedlineGoogle Scholar
  • 40 Lin B-H , Morrison RM . Higher fruit consumption linked with lower body mass index. FoodReview [serial on the Internet]. 2002 [cited 2015 Sep 8 ]. Available from: http://webarchives.cdlib.org/sw12j6951w/http:/www.ers.usda.gov/publications/FoodReview/DEC2002/frvol25i3d.pdf Google Scholar
  • 41 Halkjaer J , Tjønneland A , Overvad K , Sørensen TI . Dietary predictors of 5-year changes in waist circumference . J Am Diet Assoc . 2009 ; 109 ( 8 ): 1356 – 66 . Crossref, MedlineGoogle Scholar
  • 42 Nguyen V , Cooper L , Lowndes J , Melanson K , Angelopoulos TJ , Rippe JM , et al. Popcorn is more satiating than potato chips in normal-weight adultsNutr J2012 ; 11 : 71 . Crossref, MedlineGoogle Scholar
  • 43 Zaheer K , Akhtar MH . Recent advances in potato production, usage, nutrition—a review . Crit Rev Food Sci Nutr . 2014 Jun 12 . [Epub ahead of print]. Google Scholar
  • 44 Malik VS , Schulze MB , Hu FB . Intake of sugar-sweetened beverages and weight gain: a systematic review . Am J Clin Nutr . 2006 ; 84 ( 2 ): 274 – 88 . Crossref, MedlineGoogle Scholar
  • 45 Powell LM , Chriqui J , Chaloupka FJ . Associations between state-level soda taxes and adolescent body mass index . J Adolesc Health . 2009 ; 45 ( 3, Suppl ): S57 – 63 . Crossref, MedlineGoogle Scholar
  • 46 Nestle M . Food marketing and childhood obesity—a matter of policy . N Engl J Med . 2006 ; 354 ( 24 ): 2527 – 9 . Crossref, MedlineGoogle Scholar
  • 47 Rolls BJ , Drewnowski A , Ledikwe JH . Changing the energy density of the diet as a strategy for weight management . J Am Diet Assoc . 2005 ; 105 ( 5 Suppl 1 ): 98 – 103 . Crossref, MedlineGoogle Scholar
  • 48 Paddon-Jones D , Westman E , Mattes RD , Wolfe RR , Astrup A , Westerterp-Plantenga M . Protein, weight management, and satiety . Am J Clin Nutr . 2008 ; 87 ( 5 ): 1558S – 61S . Crossref, MedlineGoogle Scholar
  • 49 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 – 7 . Crossref, MedlineGoogle Scholar
  • 50 Koh-Banerjee P , Franz M , Sampson L , Liu S , Jacobs DR , Spiegelman D , et al. Changes in whole-grain, bran, and cereal fiber consumption in relation to 8-y weight gain among men . Am J Clin Nutr . 2004 ; 80 ( 5 ): 1237 – 45 . Crossref, MedlineGoogle Scholar
  • 51 US Department of Agriculture . Nutrition standards in the National School Lunch and School Breakfast Programs . Federal Register [serial on the Internet]. 2012 Jan 26 [cited 2015 Sep 8 ]. Available from: http://www.gpo.gov/fdsys/pkg/FR-2012-01-26/pdf/2012-1010.pdf Google Scholar
Loading Comments...