Table 3.

Study Measures and Outcomes.

Study Authors	SES of Cohort	Observed Snacking Behaviors	Weight Outcomes	Assessments
Grenard et al. [40]	Low SES	High consumption of calorie-dense snacks in low SES sample (16.5% salty; 36.1% sweet).	Not assessed.	SES was self-reported via a questionnaire; snacking was measured via one-to-one interviews at baseline and through self-initiated and random eating event reports during the 7 day monitoring period.
Hardy et al. [36]	Various SES groups	Male adolescents with lower SES have high rates of snacking (p = 0.045); female adolescents with higher SES have high rates of snacking (p = 0.034).	Smaller portion of male adolescents with overweight (OW)/obesity (OB) reported higher snacking than male adolescents without OW/OB (p < 0.001); no significant weight-related outcomes associated with snacking in female adolescents (p = 0.235).	Parents completed questionnaire with mother’s educational attainment and household income; dietary information was collected using questions based on the NSW Population Health Surveys’ food frequency questionnaire (FFQ). BMI was calculated for each participant using height and weight and OW/OB categorization was conducted via the International Obesity Task Force standards.
Larson et al. [37]	Various SES groups	High–medium and high SES categories had low consumption of snacks.	Not assessed.	SES was self-reported via survey; Youth and Adolescent FFQ was used to assess adolescent snack consumption.
Maruapula et al. [38]	Low and high SES groups	Higher SES adolescents consumed a larger quantity of snacks (p < 0.01).	High snack food diet was found to increase risk for OW/OB (p = 0.028); however, SES was not a significant determinant of OW/OB.	SES determined by attendance at tuition-free public or tuition-requiring private school and by the number of household assets. A single portable precision electronic scale and stadiometer were used to obtain anthropometric measurements; participants self-recorded their recall of food intake for the previous day. BMI was calculated for each participant using height and weight and was evaluated using the World Health Organization’s reference data for age and gender.
Pérez et al. [41]	Low, middle, and high SES groups	In males, the habit of snacking decreases as SES increases, except in quartile 2—in which, snacking increases. In females, the habit of snacking increases, except from the first to second quartile—in which, snacking decreases. Quartile 1 has the highest snacking habits of any of the other SES quartiles.	Not assessed.	SES determined using the family socioeconomic level (FSEL); snack consumption measured using 24 h dietary recall and food frequency questionnaire adaptation of the short questionnaire on frequency of dietary intake.
Schumacher et al. [42]	Low SES	Evidence of high snack consumption. In total, 6.8% of daily energy from snacks. Average of 1.5 snacks per day.	Not a significant association (p = 0.47).	The Australian Child and Adolescent Eating Survey (ACAES, version 1.2), an FFQ estimated dietary intake data over the previous 6 months; measured height and weight. BMI was calculated for each participant and then ranked into underweight, healthy, overweight, or obese categories.
Verstraeten et al. [39]	“Poor” SES and “better off” SES groups	No significant association between low SES and unhealthy snack consumption.	Not assessed.	SES was measured using the Integrated Social Indicator System for Ecuador. Then, participants classified as “poor” or “better off”; snack consumption measured using 24 h dietary recall interviews.