TABLE 2.

Overview of studies that assessed snack foods and body weight¹

Study	Target population	Study design	Intervention type	Comparison group	Results
Evans et al. (65)	Urban schoolchildren, ages 9–15 y	Cross-sectional and prospective study	Eating frequency, the average number of reported daily eating occasions, was assessed by using 2 weekday 24-h diet recalls.	Normal BMI z score	From baseline to 6 mo, BMIz increased by 0.03 units for each additional reported eating occasion. This relation was no longer significant at 1 y.
Field et al. (5)	Boys and girls between ages 9 and 14 y	Prospective cohort study	Intake of snack foods was assessed in 1996–1998 with a validated FFQ designed specifically for children and adolescents.	No control group	There was no relation between intake of snack foods and weight gain.
Hendriksen et al. (53)	Normal-weight and overweight Dutch adults	Prospective cohort study	Intake of EDS foods (sweets, cakes, and pastries and savory snacks) was assessed at baseline by using a validated FFQ.	Lowest tertile of EDS food consumption	Inconsistent evidence of an association of EDS food consumption with annual weight change.
Jodkowska et al. (63)	Polish overweight and normal-weight adolescents, aged 13–15 y, of whom 953 were overweight and 953 had normal body mass	Cross-sectional analysis	Self-reported questionnaire containing questions regarding how often selected food products were usually consumed during the week, how regularly basic meals (breakfast, lunch, supper) were eaten, and data on snacking.	“Selection in pairs,” each overweight pupil was paired with a pupil with normal body weight	Overweight adolescents consumed unhealthy products such as sweets and crisps significantly less often than their peers with appropriate body mass. Overweight girls ate dark bread significantly more often, and consumed soft drinks less often, than their peers with normal weight. Nevertheless, overweight teenagers snacked significantly less often than young people with normal body mass.
Johnstone et al. (56)	8 British men with ad libitum access to a diet of fixed composition	Randomized controlled crossover design	1) Mandatory snacks compared with snacks and 2) the composition of isoenergetically dense snacks high in protein, fat, or carbohydrate on food and energy intakes.	No snack group	Body weight was not affected by snacking.
Moreno and Rodriguez (64)	Children and adolescents	Review	Dietary aspects influencing obesity development	Review	Longitudinal studies have only found a clear and positive association between obesity development and sugar-sweetened beverage consumption; this is not the case with snacking, fast food, or food portion sizes.
Muñoz-Pareja et al. (66)	10,791 persons representative of the Spanish population who were ≥18 y old in 2008–2010	Cross-sectional study	Self-reported information was collected on 12 OREBs	Participants with ≤1 OREB	OREBs were associated with higher food energy density and higher consumption of sugary and alcoholic beverages.
Nederkoorn et al. (57)	Normal-weight undergraduate female students over a 1-y period	Longitudinal prospective cohort study	Implicit preference for food, response inhibition and BMI were measured.	Low implicit food preference	Participants with strong implicit preferences for snack foods and low inhibitory capacity gained the most weight.
Nicklas et al. (55)	1562 children aged 10 y (65% EA, 35% AA) over a 21-y period	Cross-sectional analysis	No intervention was delivered.	No comparison group	Food consumed from snacks was positively associated with overweight status.
Njike et al. (59)	34 overweight adult participants	Randomized controlled trial	Participants were provided nut-based snack bars to consume daily for 12 wk.	Conventional snack bar, specifically empty-calorie snack bar	Body fat and visceral fat were reduced and there were no adverse effects on weight, blood pressure, lipid profile, satiety, or quality of life.
Oellingrath et al. (61)	Primary-school children from Norway; middle childhood (fourth grade, 9–10 y old) to early adolescence (seventh grade, 12–13 y old).	Prospective cohort, cross-sectional, and longitudinal analysis	Categorization into 4 eating patterns: snacking, junk or convenient, varied Norwegian, and dieting	Unchanged normal-weight child	Children with stable or increased “varied Norwegian” pattern scores had a lower risk of remaining overweight over time than children with decreased scores for this pattern.
Piehowski et al. (54)	Premenopausal overweight women with BMIs (in kg/m2) of 25–43	Randomized controlled trial, 18 wk	Reduced-calorie diet including either a daily dark-chocolate snack or a nonchocolate snack	Nonchocolate snack group	Women in both the dark-chocolate snack and nonchocolate snack groups experienced decreases in body weight (−5.1 vs. −5.1 kg), hip circumference (−5.8 vs. −5.4 cm), waist circumference (−5.7 vs. −3.5 cm), fat mass (−3.9 vs. −3.6 kg), and body fat percentage (−3.4% vs. −3.1%), respectively, with no change in lean mass.
Phillips et al. (62)	Nonobese premenarchal girls 8–12 y old were enrolled between 1990 and 1993 and followed until 4 y after menarche	Prospective cohort study	Relation between EDS foods and relative weight change	No control group	There was no relation between BMIz or %BF and total EDS food consumption. Soda was the only EDS food that was significantly related to BMIz over the 10-y study period.
Viskaal-van Dongen et al. (58)	16 men and 66 women (mean age: 21.9 y), mean BMI (in kg/m2): 20.7	Randomized controlled trial, 8 wk	Randomly assigned to 1 of 4 parallel groups in a 2 × 2 design: snacks consumed with or between meals and snacks having a low (<4 kJ/g) or high (>12 kJ/g) energy density. For 8 wk, subjects consumed mandatory snacks that provided 25% of energy requirements on each day.	High energy–density snack food	No differences in changes in body weight between the 4 groups. Similarly, there were no differences in changes in body composition, PAL, and energy intake between the 4 groups.

¹AA, African American; BMIz, standardized BMI z score; EA, Euro-American; EDS, energy-dense snack; OREB, obesity-related eating behavior; PAL, physical activity level; %BF, percentage of body fat.