The Protective Role of Family Meals for Youth Obesity: 10-Year Longitudinal Associations

Published:September 27, 2014DOI:https://doi.org/10.1016/j.jpeds.2014.08.030

      Objective

      To examine whether having family meals as an adolescent protects against becoming overweight or obese 10 years later as a young adult.

      Study design

      Data from Project Eating and Activity in Teens -III, a longitudinal cohort study with emerging young adults, were used. At baseline (1998-1999), adolescents completed surveys in middle or high schools, and at 10-year follow-up (2008-2009) surveys were completed online or via mailed surveys. Young adult participants (n = 2117) were racially/ethnically and socioeconomically diverse (52% minority; 38% low income) between the ages of 19 and 31 years (mean age = 25.3; 55% female). Logistic regression was used to associate weight status at follow-up with family meal frequency 10 years earlier during adolescence, controlling and testing for interactions with demographic characteristics.

      Results

      All levels of baseline family meal frequency (ie, 1-2, 3-4, ≥5 family meals/wk) during adolescence were significantly associated with reduced odds of overweight or obesity 10 years later in young adulthood compared with never having family meals as an adolescent. Interactions by race indicated that family meals had a stronger protective effect for obesity in black vs white young adults.

      Conclusions

      Family meals during adolescence were protective against the development of overweight and obesity in young adulthood. Professionals who work with adolescents and parents may want to strategize with them how to successfully carry out at least 1 to 2 family meals per week in order to protect adolescents from overweight or obesity in young adulthood.
      BMI (Body mass index), EAT (Eating and Activity in Teens), SES (Socioeconomic status)
      See editorial, p 220
      Family meals have been suggested as one potential factor that may be protective against obesity.
      • McIntosh W.A.
      • Dean W.
      • Torres C.C.
      • Anding J.
      • KKubena K.S.
      • Nyga R.
      The American family meal.
      • Chan J.C.
      • Sobal J.
      Family meals and body weight. Analysis of multiple family members and family units.
      Higher frequency of family meals are associated with more consumption of fruit and vegetables,
      • Robertson W.
      • Friede T.
      • Blissett J.
      • Rudolf M.C.J.
      • Wailis M.
      • Stewart-Brown S.
      Pilot of “Families for Health”: community-based family intervention for obesity.
      • Lissau I.
      Prevention of overweight in the school arena.
      • Neumark-Sztainer D.
      • Hannan P.J.
      • Story M.
      • Croll J.
      • Perry C.
      Family meal patterns: associations with sociodemographic characteristics and improved dietary intake among adolescents.
      calcium and whole grains,
      • Neumark-Sztainer D.
      • Story M.
      • Ackard D.
      • Moe J.
      • Perry C.
      Family meals among adolescents: findings from a pilot study.
      • Neumark-Sztainer D.
      • Eisenberg M.E.
      • Fulkerson J.A.
      • Story M.
      • Larson N.I.
      Family meals and disordered eating in adolescents: longitudinal findings from Project EAT.
      • Larson N.I.
      • Neumark-Sztainer D.
      • Hannan P.J.
      • Story M.
      Family meals during adolescence are associated with higher diet quality and healthful meal patterns during young adulthood.
      lower levels of extreme weight control behaviors and binge eating,
      • Neumark-Sztainer D.
      • Eisenberg M.E.
      • Fulkerson J.A.
      • Story M.
      • Larson N.I.
      Family meals and disordered eating in adolescents: longitudinal findings from Project EAT.
      • Haines J.
      • Gillman M.W.
      • Rifas-Shiman S.
      • Field A.E.
      • Austin A.B.
      Family dinner and disordered eating behaviors in a large cohort of adolescents.
      and better psychosocial health in adolescents.
      • Eisenberg M.E.
      • Olson R.E.
      • Neumark-Sztainer D.
      • Story M.
      • Bearinger L.H.
      Correlations between family meals and psychosocial well-being among adolescents.
      These significant associations have held across diverse ethnic/racial backgrounds and sex. However, research examining the association between family meal frequency and adolescent body mass index (BMI) or weight status (ie, overweight, obese) has shown inconsistent findings.
      • Gable S.
      • Chang Y.
      • Krull J.L.
      Television watching and frequency of family meals are predictive of overweight onset and persistence in a national sample of school-aged children.
      • Gillman M.W.
      • Rifas-Shiman S.L.
      • Frazier A.L.
      • Rockett H.R.
      • Camargo Jr., C.A.
      • Field A.E.
      • et al.
      Family dinner and diet quality among older children and adolescents.
      • Sen B.
      Frequency of family dinner and adolescent body weight status: evidence from the National Longitudinal Survey of Youth, 1997.
      • Taveras E.
      • Rifas-Shiman S.
      • Berkey C.
      • Rockett H.
      • Field A.
      • Frazier A.
      • et al.
      Family dinner and adolescent overweight.
      • Yuasa K.
      • Sei M.
      • Takeda E.
      • Ewis A.
      • Munakata H.
      • Onishi C.
      • et al.
      Effects of lifestyle habits and eating meals together with the family on the prevalence of obesity among school children in Tokushima, Japan: a cross-sectional questionnaire-based survey.
      • Anderson S.E.
      • Whitaker R.C.
      Household routines and obesity in US preschool-aged children.
      • Rollins B.Y.
      • Belue R.Z.
      • Francis L.A.
      The beneficial effect of family meals on obesity differs by race, sex, and household education: the national survey of children's health, 2003-2004.
      • Mamun A.A.
      • Lawlor D.A.
      • O'Callaghan M.J.
      • Williams G.M.
      • Najman J.M.
      Positive maternal attitude to the family eating together decreases the risk of adolescent overweight.
      • Utter J.
      • Scragg R.
      • Schaaf D.
      • Mhurchu C.N.
      Relationships between frequency of family meals, BMI and nutritional aspects of the home food environment among New Zealand adolescents.
      • Woodruff S.J.
      • Hanning R.M.
      • McGoldrick K.
      • Brown K.S.
      Healthy eating index-C is positively associated with family dinner frequency among students in grades 6-8 from Southern Ontario, Canada.
      • Fulkerson J.A.
      • Pasch K.E.
      • Stigler M.H.
      • Farbakhsh K.
      • Perry C.L.
      • Komro K.A.
      Longitudinal associations between family dinner and adolescent perceptions of parent-child communication among racially diverse urban youth.
      To date, the majority of research looking at the association between family meals and adolescent overweight/obesity has primarily been cross-sectional and found mixed results. More specifically, some cross-sectional studies have shown an inverse relationship between the frequency of family meals and adolescent BMI,
      • Gable S.
      • Chang Y.
      • Krull J.L.
      Television watching and frequency of family meals are predictive of overweight onset and persistence in a national sample of school-aged children.
      • Gillman M.W.
      • Rifas-Shiman S.L.
      • Frazier A.L.
      • Rockett H.R.
      • Camargo Jr., C.A.
      • Field A.E.
      • et al.
      Family dinner and diet quality among older children and adolescents.
      • Sen B.
      Frequency of family dinner and adolescent body weight status: evidence from the National Longitudinal Survey of Youth, 1997.
      • Taveras E.
      • Rifas-Shiman S.
      • Berkey C.
      • Rockett H.
      • Field A.
      • Frazier A.
      • et al.
      Family dinner and adolescent overweight.
      • Yuasa K.
      • Sei M.
      • Takeda E.
      • Ewis A.
      • Munakata H.
      • Onishi C.
      • et al.
      Effects of lifestyle habits and eating meals together with the family on the prevalence of obesity among school children in Tokushima, Japan: a cross-sectional questionnaire-based survey.
      • Anderson S.E.
      • Whitaker R.C.
      Household routines and obesity in US preschool-aged children.
      and others have shown no association.
      • Mamun A.A.
      • Lawlor D.A.
      • O'Callaghan M.J.
      • Williams G.M.
      • Najman J.M.
      Positive maternal attitude to the family eating together decreases the risk of adolescent overweight.
      • Utter J.
      • Scragg R.
      • Schaaf D.
      • Mhurchu C.N.
      Relationships between frequency of family meals, BMI and nutritional aspects of the home food environment among New Zealand adolescents.
      • Woodruff S.J.
      • Hanning R.M.
      • McGoldrick K.
      • Brown K.S.
      Healthy eating index-C is positively associated with family dinner frequency among students in grades 6-8 from Southern Ontario, Canada.
      • Wurback A.
      • Zellner K.
      • Kromeyer-Hauschild K.
      Meals patterns among children and adolescents and their associations with weight status and parental characteristics.
      Furthermore, these cross-section findings have often been specific to 1 sex
      • Fulkerson J.A.
      • Pasch K.E.
      • Stigler M.H.
      • Farbakhsh K.
      • Perry C.L.
      • Komro K.A.
      Longitudinal associations between family dinner and adolescent perceptions of parent-child communication among racially diverse urban youth.
      or age group,
      • Rollins B.Y.
      • Belue R.Z.
      • Francis L.A.
      The beneficial effect of family meals on obesity differs by race, sex, and household education: the national survey of children's health, 2003-2004.
      • Fulkerson J.A.
      • Pasch K.E.
      • Stigler M.H.
      • Farbakhsh K.
      • Perry C.L.
      • Komro K.A.
      Longitudinal associations between family dinner and adolescent perceptions of parent-child communication among racially diverse urban youth.
      race/ethnicity,
      • Rollins B.Y.
      • Belue R.Z.
      • Francis L.A.
      The beneficial effect of family meals on obesity differs by race, sex, and household education: the national survey of children's health, 2003-2004.
      or social class.
      • Rollins B.Y.
      • Belue R.Z.
      • Francis L.A.
      The beneficial effect of family meals on obesity differs by race, sex, and household education: the national survey of children's health, 2003-2004.
      • Fulkerson J.A.
      • Pasch K.E.
      • Stigler M.H.
      • Farbakhsh K.
      • Perry C.L.
      • Komro K.A.
      Longitudinal associations between family dinner and adolescent perceptions of parent-child communication among racially diverse urban youth.
      A meta-analysis also noted the limitations of studies described above and further identified that research has mostly been conducted with children (ages 6-12 years) rather than adolescents (ages 13-18 years) or young adults (ages 18 years and older).
      • Valdes J.R.-A.F.
      • Aguilar L.
      • Jaen-Casquero M.B.
      • Royo-Bordonada M.A.
      Frequency of family meals and childhood overweight: a systematic review.
      Given the high prevalence of adolescent obesity in the US,
      • Ogden C.L.
      • Carroll M.D.
      • Curtin L.R.
      • McDowell M.A.
      • Tabak C.J.
      • Flegal K.M.
      Prevalence of overweight and obesity in the United States, 1999-2004.
      • Ogden C.L.
      • Carroll M.D.
      • Kit B.K.
      • Flegal K.M.
      Prevalence of obesity and trends in body mass index among US children and adolescents, 1999-2010.
      and the likelihood of obesity tracking into adulthood,
      • Pi-Sunyer F.X.
      The obesity epidemic: pathophysiology and consequences of obesity.
      it is important to identify modifiable factors in the home environment that can protect against overweight/obesity through the transition to adulthood. Identifying whether race/ethnicity modifies the association between family meal frequency and young adult overweight/obesity is important in order to identify whether certain groups benefit more from having family meals. The current study addresses these gaps and, furthermore, examines an important developmental time frame, transitioning from adolescence to young adulthood, which is a common time to gain weight and to have fewer family meals.
      • Nonnemaker J.M.
      • Morgan-Lopez A.A.
      • Pais J.M.
      • Finkelstein E.A.
      Youth BMI trajectories: evidence from the NLSY97.
      The main aim of this study is to use longitudinal data from a 10-year population-based study (Project Eating and Activity in Teens [EAT] III) with racially/ethnically and socioeconomically diverse adolescents to investigate whether having frequent family meals as an adolescent is protective for overweight and obesity in young adulthood. In addition, interactions were examined to identify whether associations between family meal frequency in adolescence and overweight/obesity status in young adulthood differ by race/ethnicity. The specific hypotheses for the study are that higher family meal frequency during adolescence will predict lower odds of overweight and obesity longitudinally into young adulthood and race/ethnicity will modify this significant association.

      Methods

      Data were drawn from the first and third waves of a 10-year longitudinal cohort study designed to examine weight-related variables (eg, dietary intake, physical activity, weight control behaviors, weight status) among adolescents; Project EAT I and Project EAT III.
      • Neumark-Sztainer D.
      • Story M.
      • Hannan P.J.
      • Croll J.
      Overweight status and eating patterns among adolescents: where do youth stand in comparison to the Health People 2010 Objectives?.
      The EAT III sample includes participants (n = 2287) who responded to the 10-year follow-up survey. At baseline (ie, Project EAT I), middle school and high school students in 31 public schools in the Minneapolis/St. Paul metropolitan area of Minnesota completed anthropometric measures and surveys in class during the 1998-1999 academic school year. The 10-year follow-up (ie, EAT III) study was designed to follow-up on the original participants in 2008-2009 as they progressed from adolescence into emerging adulthood. Data collection for EAT III ran from November 2008 to October 2009 and was conducted by the Health Survey Research Center in the School of Public Health at the University of Minnesota, Minneapolis. The Institutional Review Board Human Subjects Committee at the University of Minnesota approved all protocols used in Project EAT at each time point.
      Of the original Project EAT I participants (n = 4746), 27.5% (n = 1304) were lost to follow-up primarily for missing contact information at time 1 (n = 411) and no address found at follow-up (n = 712).
      • Eisenberg M.E.
      • Berge J.M.
      • Fulkerson J.A.
      • Neumark-Sztainer D.
      Weight comments by family and significant others in young adulthood.
      For Project EAT III, survey invitation letters were sent to participants with a web address and a unique password for completing the online version of the Project EAT III survey. Participants were able to select an option to be sent a paper version of the survey. Nonresponders were sent 3 reminder letters. All reminder letters included a postage-paid card for requesting paper copies. The second reminder letter additionally included paper copies of the survey. In addition, reminder postcards were sent to participants who did not complete the survey after logging into the online version or who requested paper copies at some point in the process. Internet tracking services were employed to identify correct addresses when mailings were returned due to incorrect addresses.
      A total of 1257 females and 1030 males completed Project EAT III surveys that were determined to be valid and adequately complete for inclusion in analyses. This represented 66.4% of participants who could be contacted (48.2% of the original school-based sample). The majority (86.5%) of survey respondents completed the online survey. One-third of participants (29.9%) were in middle school at baseline (mean age = 12.8 ± 0.7 years) and were in emerging young adulthood (mean age = 23.2 ± 1.0 years) at follow-up. Two-thirds of participants (70.1%) were in high school at baseline (mean age = 15.9 ± 0.8 years) and were in young adulthood (mean age = 26.2 ± 0.9 years) at follow-up.
      • Larson N.
      • Neumark-Sztainer D.
      • Harwood E.M.
      • Eisenberg M.E.
      • Wall M.
      • Hannan P.J.
      Do young adults participate in surveys that 'go green'? Response rates to a web and mailed survey of weight-related behaviors.
      The Project EAT survey was developed at baseline and revised for use at 10-year follow-up specifically to assess items of relevance to young people as they transitioned into young adulthood and developed more independent lifestyles. The survey is a 221-item survey assessing a range of socioenvironmental, personal, behavioral, and familial factors of potential relevance to nutritional health and obesity among adolescents. The survey development is described elsewhere.
      • Neumark-Sztainer D.
      • Story M.
      • Hannan P.J.
      • Croll J.
      Overweight status and eating patterns among adolescents: where do youth stand in comparison to the Health People 2010 Objectives?.
      Two-week survey test-retest reliability data were collected from 167 adolescents at baseline, and validity of the survey was established and details are previously published.
      • Neumark-Sztainer D.
      • Story M.
      • Hannan P.J.
      • Croll J.
      Overweight status and eating patterns among adolescents: where do youth stand in comparison to the Health People 2010 Objectives?.
      To assess family meal frequency, the exposure variable, adolescents were asked, “During the past seven days, how many times did all, or most, of your family living in your house eat a meal together?” Response options included never, 1-2 times, 3-4 times, 5-6 times, 7 times, and more than 7 times (test-retest r = .83).
      • Fulkerson J.A.
      • Neumark-Sztainer D.
      • Story M.
      Adolescent and parent views of family meals.
      Although this measure is a single item, it has been used in the majority of prior studies looking at family meal frequency,
      • Larson N.I.
      • Neumark-Sztainer D.
      • Hannan P.J.
      • Story M.
      Family meals during adolescence are associated with higher diet quality and healthful meal patterns during young adulthood.
      • Goldfield G.S.
      • Murray M.A.
      • Buchholz A.
      • Henderson K.
      • Obeid N.
      • Kukaswadia A.
      • et al.
      Family meals and body mass index among adolescents: effects of gender.
      • Utter J.
      • Denny S.
      • Robinson E.
      • Fleming T.
      • Ameratunga S.
      • Grant S.
      Family meals and the well-being of adolescents.
      thus, increasing the ability to compare results across studies. It has also been shown to be reliable and valid with diverse populations.
      • Neumark-Sztainer D.
      • Story M.
      • Ackard D.
      • Moe J.
      • Perry C.
      Family meals among adolescents: findings from a pilot study.
      • Fulkerson J.A.
      • Neumark-Sztainer D.
      • Story M.
      Adolescent and parent views of family meals.
      The 3 highest categories were collapsed to allow for meaningful comparisons between parents who had infrequent or occasional family meals with families who had more regular family meals.
      • Berge J.M.
      • MacLehose R.
      • Loth K.A.
      • Eisenberg M.E.
      • Fulkerson J.A.
      • Neumark-Sztainer D.
      Family meals: associations with weight and eating behaviors among mothers and fathers.
      To assess overweight and obesity status, the outcome variable, self-reported height and weight values were used to calculate BMI (weight [kg]/height [m2]) at baseline and 10-year follow-up. At baseline, high correlations were found between self-reported and measured BMI in male (r = 0.88) and female (r = 0.85) adolescents.
      • Himes J.
      • Hannan P.
      • Wall M.
      • Neumark-Sztainer D.
      Factors associated with errors in self-reports of stature, weight, and body mass index in Minnesota adolescents.
      In addition, at EAT III very high correlations between self-reported and measured BMI were found in a validation subsample of 63 male and 62 female EAT III study participants (r = 0.95 for males and r = 0.98 for females). At baseline, overweight status was determined based on a BMI at or above the 85th percentile for sex and age using reference data from the Centers for Disease Control and Prevention.
      • Barlow S.
      Expert committee recommendations regarding the prevention, assessment, and treatment of child and adolescent overweight and obesity: summary report.
      Weight status at 10-year follow-up was defined according to current BMI guidelines for adults (overweight: BMI ≥25 and <30 kg/m2; obese: BMI ≥30 kg/m2).
      Executive summary of the clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults.
      Control variables included sex, age, race/ethnicity, and socioeconomic status (SES). These variables were assessed by self-report at EAT I. Race/ethnicity was assessed with one survey item: “Do you think of yourself as: (1) white; (2) black or African American; (3) Hispanic or Latino; (4) Asian American; (5) Hawaiian or Pacific Islander; or (6) American Indian or Native American” and respondents were asked to check all that apply. Those reporting more than 1 response were assigned to the category “mixed/other” background. As few participants identified their background as “Hawaiian or Pacific Islander” or “Native American,” these youth were also included in the category “mixed/other.” Classification tree methodology
      • Breiman L.
      • Friedman J.
      • Olshen R.
      • Stone C.
      Classification and regression trees.
      was used to generate 5 categories of SES (low SES, low-mid SES, middle SES, mid-high SES, and high SES).
      • Neumark-Sztainer D.
      • Wall M.
      • Larson N.I.
      • Eisenberg M.E.
      • Loth K.
      Dieting and disordered eating behaviors from adolescence to young adulthood: Findings from a 10-year longitudinal study.
      The prime determinant of SES was the higher education level of either parent. Subsidiary variables were family eligibility for free/reduced-price school lunch, family receipt of public assistance, and parent employment status.
      The analytic sample (n = 2117) for the current study excluded women who were pregnant at follow-up (n = 90) and individuals who did not have self-reported BMI measurement at baseline or 10-year follow-up (n = 63) or did not respond to the family meal frequency question at baseline (n = 17). For each level of baseline family meal frequency (ie, never, 1-2, 3-4, 5, or more), the proportion of individuals who were overweight or obese at 10-year follow-up was calculated. Logistic regressions of overweight and obese status at 10-year follow-up predicted by baseline family meal frequency controlling for sex, age, race, SES, and baseline overweight or obese status were conducted to generate ORs of the effect of family meal frequency on overweight and obsety status 10 years later. Tests for interaction were performed between family meal frequency and each of the following: sex, age cohort (younger/older), race/ethnicity, and baseline overweight status. Only the interaction with race was significant, and thus, results are additionally presented stratified by race/ethnicity. All analyses were performed in 2012 using SAS (v 9.2, 2011, SAS Institute, Cary, North Carolina) and incorporating nonresponse sampling weights. Tests were considered statistically significant when P values were less than .05. The 95% CIs are given for all ORs.
      Because attrition from the baseline sample did not occur at random, in all analyses, the data were weighted using the response propensity method.
      • Little R.
      Survey nonresponse adjustments for estimates of means.
      Response propensities (ie, the probability of responding to the Project EAT III survey) were estimated using a logistic regression of response at 10-year follow-up on a large number of predictor variables from the Project EAT I survey. Weights were additionally calibrated so that the weighted total sample sizes used in analyses for each sex cohort accurately reflect the actual observed sample sizes in those groups. The weighting method resulted in estimates representative of the demographic make-up of the original school-based sample, thereby allowing results to be more fully generalizable to the population of young people in the Minneapolis/St. Paul metropolitan area. Specifically, the weighted sample was 48.4% white, 18.6% African American, 19.6% Asian (primarily Hmong), 5.9% Hispanic, 3.3% Native American, and 4.2% mixed or other race/ethnicity. The sample was well-distributed across the 5 categories of SES: 18.0% low, 19.0% low-middle, 26.2% middle, 23.3% upper-middle, and 13.5% high. Respondents with 10-year follow-up data included in analyses were more likely to be female, higher SES, white, and from the older cohort than those not included, but these differences were eliminated through the use of nonresponse weighting. After weighting there were no differences in baseline weight status between responders and nonresponders. However, even after weighting for nonresponse and controlling for demographics, respondents had on average lower family meals at baseline than nonrespondents (3.5 meals per week vs 3.9, P value <.001).

      Results

      Table I shows the proportion of young adults who were overweight at 10-year follow-up by each baseline level of family meal frequency (ie, never, 1-2, 3-4, 5 or more). At 10-year follow-up, 51% of the sample was overweight and 22% were obese. Among adolescents who reported that they never ate family meals together (15% of the sample) at baseline, 60% of them were overweight at 10-year follow-up and 29% were obese. In comparison, among adolescents who reported that they ate 1-2, 3-4, or 5 or more family meals together at baseline, 47%-51% of them were overweight at 10-year follow-up and 19%-22% were obese.
      Table IPercent overweight or obesity at 10-year follow-up by frequency of family meals at baseline
      % (n) family meal frequency at baseline% (n) overweight (BMI ≥25 and <30 kg/m2) at 10-y follow-up by family meals at baseline% (n) obese (BMI ≥30 kg/m2) at 10-y follow-up by family meals at baseline
      Overall sample100% (n = 2117)51% (n = 1070)22% (n = 458)
      Family meals frequency
       Never15 (310)60 (187)29 (91)
       1-219 (413)47 (196)22 (89)
       3-422 (475)51 (242)19 (91)
       5 or more43 (920)49 (446)20 (187)
      Table II shows ORs (adjusted for baseline demographic characteristics and overweight or obese status) and 95% CIs for 10-year follow-up overweight and obese status by baseline family meal frequency. Overall, all levels of baseline family meal frequency (ie, 1-2, 3-4, 5 or more per week) during adolescence were significantly associated with reduced odds of overweight 10-years later in young adulthood compared with never having family meals during adolescence. For example, the OR of 0.55 indicates that adolescents who had 1-2 family meals at baseline had 45% decreased odds of being overweight 10-years later as young adults compared with young adults who never had family meals as adolescents (baseline) (Table II). Additional post-hoc tests did not find any significant differences between the different levels of family meal frequency (ie, 5 or more meals did not show any improvement compared with 3-4 meals).
      Table IIORs
      Bold text is P < .05. ORs and 95% CIs from logistic regression controlling for baseline sex, age, race, SES, and overweight or obese status.
      of young adult overweight status or obesity at 10-year follow-up
      10-y follow-up overweight (BMI ≥25 and <30 kg/m2) OR95% CI10-y follow-up obese (BMI ≥30 kg/m2)

      OR
      95% CI
      Family meal frequency at baseline
       NeverReferenceReference
       1-2 vs never0.550.38-0.790.670.44-1.02
       3-4 vs never0.600.42-0.850.500.33-0.76
       5 or more vs never0.630.46-0.870.680.47-0.99
      Bold text is P < .05. ORs and 95% CIs from logistic regression controlling for baseline sex, age, race, SES, and overweight or obese status.
      A significant interaction between family meal frequency and race was found when predicting odds of obesity at 10-year follow-up (likelihood ratio test for interaction = 28.5, d.f. = 5, P < .0001). The interaction indicated that the strongest protective effect of increasing family meal frequency during adolescence on obesity was among black young adults. Specifically, black adolescents who had 1-2 family meals vs never (OR = 0.60; CI = 0.49-0.73), 3-4 family meals vs never (OR = 0.30; CI = 0.19-0.47), or 5 or more family meals vs never (OR = 0.15; CI = 0.07-0.31) had reduced odds of obesity as young adults. No significant associations were found between family meals and obesity in any of the other races in stratified analyses, and no interaction between race and family meals was found when considering overweight status as the outcome rather than obesity.

      Discussion

      Overall, results from the current study suggest that having family meals as an adolescent is protective for young adult overweight and obesity 10 years later. All levels of family meal frequency (ie, 1-2, 3-4, 5 or more) were protective for young adult overweight, compared with never having family meals. This study lends support to previous limited research on family meals that has shown family meal frequency is inversely associated with adolescent overweight/obesity.
      • Gable S.
      • Chang Y.
      • Krull J.L.
      Television watching and frequency of family meals are predictive of overweight onset and persistence in a national sample of school-aged children.
      • Gillman M.W.
      • Rifas-Shiman S.L.
      • Frazier A.L.
      • Rockett H.R.
      • Camargo Jr., C.A.
      • Field A.E.
      • et al.
      Family dinner and diet quality among older children and adolescents.
      • Sen B.
      Frequency of family dinner and adolescent body weight status: evidence from the National Longitudinal Survey of Youth, 1997.
      • Taveras E.
      • Rifas-Shiman S.
      • Berkey C.
      • Rockett H.
      • Field A.
      • Frazier A.
      • et al.
      Family dinner and adolescent overweight.
      • Yuasa K.
      • Sei M.
      • Takeda E.
      • Ewis A.
      • Munakata H.
      • Onishi C.
      • et al.
      Effects of lifestyle habits and eating meals together with the family on the prevalence of obesity among school children in Tokushima, Japan: a cross-sectional questionnaire-based survey.
      • Anderson S.E.
      • Whitaker R.C.
      Household routines and obesity in US preschool-aged children.
      Furthermore, results from the current study extend prior research by showing that having family meals in adolescence is protective for overweight/obesity longitudinally into young adulthood.
      Our results also indicated a stronger protective effect of family meal frequency on obesity among black young adults compared with white young adults. This finding supports previous results showing differences in the benefits (eg, healthful dietary intake, weight status) of family meals by race/ethnicity (not protective for Hispanic boys).
      • Mamun A.A.
      • Lawlor D.A.
      • O'Callaghan M.J.
      • Williams G.M.
      • Najman J.M.
      Positive maternal attitude to the family eating together decreases the risk of adolescent overweight.
      However, the limited overall significant interactions by race/ethnicity, sex, SES, and age in the current study suggests the protective nature of family meals for adolescents across all race/ethnicities, sex, age, and SES status.
      There are several potential mechanisms that may explain the protective association between family meal frequency and young adult overweight/obesity. First, it may be that family meals are healthier than other meals because it is common to serve fruits and vegetables during the meals.
      • Neumark-Sztainer D.
      • Maclehose R.
      • Loth K.
      • Fulkerson J.A.
      • Eisenberg M.E.
      • Berge J.
      What's for dinner? Types of food served at family dinner differ across parent and family characteristics.
      In addition, family meals may provide opportunities for emotional connection among family members, creating a supportive environment for emotion regulation and a sense of security that gives children the ability to regulate their own eating behaviors in their day-to-day lives.
      • Videon T.M.
      • Manning C.K.
      Influences on adolescent eating patterns: the importance of family meals.
      Third, parental modeling of healthful eating behaviors and recognizing satiety cues may also be occurring at family meals.
      • Berge J.
      • Wall M.
      • Bauer K.
      • Neumark-Sztainer D.
      Parenting characteristics in the home environment and adolescent overweight.
      Thus, it would be important for future research to examine further the potential mechanisms behind the results found in the current study such as looking at the emotional atmosphere at the meal, the quality and quantity of food served, and modeling behavior that occurs during family meals. Furthermore, although it is unclear why having only 1-2 family meals per week was protective for adolescents, it may be the case that the routines/patterns and associated benefits of family meal patterns (ie, supportive environment, connection) established in childhood carry forward into adolescence and young adulthood. For example, if families have frequent families meals with their young children (eg, 4-5 per week), by the time their child reaches adolescence—and they have to reduce meal frequency because of activities or work schedules—the benefits and protective nature of family meals may be sustainable because of previously exerted efforts.
      Informing parents that even having 1 to 2 family meals per week may protect their child from overweight or obesity in young adulthood would be important.
      • Breaugh J.
      • Frye N.
      Work and family conflict: The importance of family-friendly employment practices and family-supportive supervisors.
      Results from the current study may also help bolster the evidence-base for pushing efforts forward to create workplace policies and procedures that allow for parents to have more flexible hours, benefits for less than full time employment, and providing employees greater control over where and when their work is done in order for families to have more frequent family meals.
      • Breaugh J.
      • Frye N.
      Work and family conflict: The importance of family-friendly employment practices and family-supportive supervisors.
      The findings from the current study may also be useful for informing future research on family meals. For example, it would be important to understand more about the potential mechanism(s) contributing to the association between family meal frequency and reduced risk of overweight/obesity in young adults (eg, the context and emotional atmosphere of the family meal, the content of conversations at the meal, who is present at the meals, distractions such as television, texting or electronic tablets/games that may moderate the protective influence of family meals).
      • Berge J.M.
      • Jin S.W.
      • Hannan P.
      • Neumark-Sztainer D.
      Structural and Interpersonal characteristics of family meals: associations with Adolescent Body Mass Index and Dietary Patterns.
      One limitation of the study includes self-reported BMI. Although a validity substudy was conducted with these adolescents and self-reported BMI and objectively measured BMI were highly correlated, it is still preferable to have objectively measured weight and heights. In addition, the generalizability of study findings may be limited as all the baseline data were collected in the Minneapolis/St. Paul area of Minnesota. Furthermore, frequency of family meals was differentially reported by responders vs non-responders to the EAT survey (3.7 vs 3.9 meals per week respectively). However, because responders to the project EAT survey were more likely to eat fewer family meals on average than nonresponders, current analyses focus on families who may have adolescents at higher risk of obesity. Thus, results are able to speak to protective factors for this at-risk population. Finally, this study was not able to address why having family meals was protective for adolescence. Examining the quality (ie, emotional atmosphere, interpersonal interactions) of family meals would be an important next step to potentially provide more details regarding why family meals are protective and how they function on a day-to-day basis so more families can take advantage of the protective nature of family meals.

      References

        • McIntosh W.A.
        • Dean W.
        • Torres C.C.
        • Anding J.
        • KKubena K.S.
        • Nyga R.
        The American family meal.
        in: Meiselman H.L. CRC Press, New York2009
        • Chan J.C.
        • Sobal J.
        Family meals and body weight. Analysis of multiple family members and family units.
        Appetite. 2011; 57: 517-524
        • Robertson W.
        • Friede T.
        • Blissett J.
        • Rudolf M.C.J.
        • Wailis M.
        • Stewart-Brown S.
        Pilot of “Families for Health”: community-based family intervention for obesity.
        Arch Dis Child. 2008; 93: 921-926
        • Lissau I.
        Prevention of overweight in the school arena.
        Acta Paediatr Suppl. 2007; 96: 12-18
        • Neumark-Sztainer D.
        • Hannan P.J.
        • Story M.
        • Croll J.
        • Perry C.
        Family meal patterns: associations with sociodemographic characteristics and improved dietary intake among adolescents.
        J Am Diet Assoc. 2003; 103: 317-322
        • Neumark-Sztainer D.
        • Story M.
        • Ackard D.
        • Moe J.
        • Perry C.
        Family meals among adolescents: findings from a pilot study.
        J Nutr Educ. 2000; 32: 335-340
        • Neumark-Sztainer D.
        • Eisenberg M.E.
        • Fulkerson J.A.
        • Story M.
        • Larson N.I.
        Family meals and disordered eating in adolescents: longitudinal findings from Project EAT.
        Arch Pediatr Adolesc Med. 2008; 162: 17-22
        • Larson N.I.
        • Neumark-Sztainer D.
        • Hannan P.J.
        • Story M.
        Family meals during adolescence are associated with higher diet quality and healthful meal patterns during young adulthood.
        J Am Diet Assoc. 2007; 107: 1502-1510
        • Haines J.
        • Gillman M.W.
        • Rifas-Shiman S.
        • Field A.E.
        • Austin A.B.
        Family dinner and disordered eating behaviors in a large cohort of adolescents.
        Eating Disord. 2010; 18: 10-24
        • Eisenberg M.E.
        • Olson R.E.
        • Neumark-Sztainer D.
        • Story M.
        • Bearinger L.H.
        Correlations between family meals and psychosocial well-being among adolescents.
        Arch Pediatr Adolesc Med. 2004; 158: 792-796
        • Gable S.
        • Chang Y.
        • Krull J.L.
        Television watching and frequency of family meals are predictive of overweight onset and persistence in a national sample of school-aged children.
        J Am Diet Assoc. 2007; 107: 53-61
        • Gillman M.W.
        • Rifas-Shiman S.L.
        • Frazier A.L.
        • Rockett H.R.
        • Camargo Jr., C.A.
        • Field A.E.
        • et al.
        Family dinner and diet quality among older children and adolescents.
        Arch Fam Med. 2000; 9: 235-240
        • Sen B.
        Frequency of family dinner and adolescent body weight status: evidence from the National Longitudinal Survey of Youth, 1997.
        Obesity. 2006; 14: 2266-2276
        • Taveras E.
        • Rifas-Shiman S.
        • Berkey C.
        • Rockett H.
        • Field A.
        • Frazier A.
        • et al.
        Family dinner and adolescent overweight.
        Obes Res. 2005; 13: 900-906
        • Yuasa K.
        • Sei M.
        • Takeda E.
        • Ewis A.
        • Munakata H.
        • Onishi C.
        • et al.
        Effects of lifestyle habits and eating meals together with the family on the prevalence of obesity among school children in Tokushima, Japan: a cross-sectional questionnaire-based survey.
        J Med Invest. 2008; 55: 71-77
        • Anderson S.E.
        • Whitaker R.C.
        Household routines and obesity in US preschool-aged children.
        Pediatrics. 2010; 125: 420-428
        • Rollins B.Y.
        • Belue R.Z.
        • Francis L.A.
        The beneficial effect of family meals on obesity differs by race, sex, and household education: the national survey of children's health, 2003-2004.
        J Am Diet Assoc. 2010; 110: 1335-1339
        • Mamun A.A.
        • Lawlor D.A.
        • O'Callaghan M.J.
        • Williams G.M.
        • Najman J.M.
        Positive maternal attitude to the family eating together decreases the risk of adolescent overweight.
        Obes Res. 2005; 13: 1422-1430
        • Utter J.
        • Scragg R.
        • Schaaf D.
        • Mhurchu C.N.
        Relationships between frequency of family meals, BMI and nutritional aspects of the home food environment among New Zealand adolescents.
        Int J Behav Nutr Phys Activity. 2008; 5: 50
        • Woodruff S.J.
        • Hanning R.M.
        • McGoldrick K.
        • Brown K.S.
        Healthy eating index-C is positively associated with family dinner frequency among students in grades 6-8 from Southern Ontario, Canada.
        Eur J Clin Nutr. 2010; 64: 454-460
        • Fulkerson J.A.
        • Pasch K.E.
        • Stigler M.H.
        • Farbakhsh K.
        • Perry C.L.
        • Komro K.A.
        Longitudinal associations between family dinner and adolescent perceptions of parent-child communication among racially diverse urban youth.
        J Fam Psychol. 2010; 24: 261-270
        • Wurback A.
        • Zellner K.
        • Kromeyer-Hauschild K.
        Meals patterns among children and adolescents and their associations with weight status and parental characteristics.
        Public Health Nutr. 2009; 12: 71-77
        • Valdes J.R.-A.F.
        • Aguilar L.
        • Jaen-Casquero M.B.
        • Royo-Bordonada M.A.
        Frequency of family meals and childhood overweight: a systematic review.
        Pediatr Obes. 2013; 8: e1-e13
        • Ogden C.L.
        • Carroll M.D.
        • Curtin L.R.
        • McDowell M.A.
        • Tabak C.J.
        • Flegal K.M.
        Prevalence of overweight and obesity in the United States, 1999-2004.
        JAMA. 2006; 295: 1549-1555
        • Ogden C.L.
        • Carroll M.D.
        • Kit B.K.
        • Flegal K.M.
        Prevalence of obesity and trends in body mass index among US children and adolescents, 1999-2010.
        JAMA. 2012; 307: 483-490
        • Pi-Sunyer F.X.
        The obesity epidemic: pathophysiology and consequences of obesity.
        Obes Res. 2002; 10: 97S-104S
        • Nonnemaker J.M.
        • Morgan-Lopez A.A.
        • Pais J.M.
        • Finkelstein E.A.
        Youth BMI trajectories: evidence from the NLSY97.
        Obesity (Silver Spring, Md). 2009; 17: 1274-1280
        • Neumark-Sztainer D.
        • Story M.
        • Hannan P.J.
        • Croll J.
        Overweight status and eating patterns among adolescents: where do youth stand in comparison to the Health People 2010 Objectives?.
        Am J Public Health. 2002; 92: 844-851
        • Eisenberg M.E.
        • Berge J.M.
        • Fulkerson J.A.
        • Neumark-Sztainer D.
        Weight comments by family and significant others in young adulthood.
        Body Image. 2011; 8: 12-19
        • Larson N.
        • Neumark-Sztainer D.
        • Harwood E.M.
        • Eisenberg M.E.
        • Wall M.
        • Hannan P.J.
        Do young adults participate in surveys that 'go green'? Response rates to a web and mailed survey of weight-related behaviors.
        Int J Child Health Hum Develop. 2011; 4: 225-231
        • Fulkerson J.A.
        • Neumark-Sztainer D.
        • Story M.
        Adolescent and parent views of family meals.
        J Am Diet Assoc. 2006; 106: 526-532
        • Goldfield G.S.
        • Murray M.A.
        • Buchholz A.
        • Henderson K.
        • Obeid N.
        • Kukaswadia A.
        • et al.
        Family meals and body mass index among adolescents: effects of gender.
        Appl Physiol Nutr Metab. 2011; 36: 539-546
        • Utter J.
        • Denny S.
        • Robinson E.
        • Fleming T.
        • Ameratunga S.
        • Grant S.
        Family meals and the well-being of adolescents.
        J Paediatr Child Health. 2013; 49: 906-911
        • Berge J.M.
        • MacLehose R.
        • Loth K.A.
        • Eisenberg M.E.
        • Fulkerson J.A.
        • Neumark-Sztainer D.
        Family meals: associations with weight and eating behaviors among mothers and fathers.
        Appetite. 2012; 58: 1128-1135
        • Himes J.
        • Hannan P.
        • Wall M.
        • Neumark-Sztainer D.
        Factors associated with errors in self-reports of stature, weight, and body mass index in Minnesota adolescents.
        Ann Epidemiol. 2005; 4: 272-278
        • Barlow S.
        Expert committee recommendations regarding the prevention, assessment, and treatment of child and adolescent overweight and obesity: summary report.
        Pediatrics. 2007; 120: S164-S192
      1. Executive summary of the clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults.
        Arch Intern Med. 1998; 158: 1855-1867
        • Breiman L.
        • Friedman J.
        • Olshen R.
        • Stone C.
        Classification and regression trees.
        Wadsworth International Group, Belmont, CA1984
        • Neumark-Sztainer D.
        • Wall M.
        • Larson N.I.
        • Eisenberg M.E.
        • Loth K.
        Dieting and disordered eating behaviors from adolescence to young adulthood: Findings from a 10-year longitudinal study.
        J Am Diet Assoc. 2011; 111: 1004-1011
        • Little R.
        Survey nonresponse adjustments for estimates of means.
        Int Stat Rev. 1986; 54: 139-157
        • Neumark-Sztainer D.
        • Maclehose R.
        • Loth K.
        • Fulkerson J.A.
        • Eisenberg M.E.
        • Berge J.
        What's for dinner? Types of food served at family dinner differ across parent and family characteristics.
        Public Health Nutr. 2012; 17: 145-155
        • Videon T.M.
        • Manning C.K.
        Influences on adolescent eating patterns: the importance of family meals.
        J Adolesc Health. 2003; 32: 365-373
        • Berge J.
        • Wall M.
        • Bauer K.
        • Neumark-Sztainer D.
        Parenting characteristics in the home environment and adolescent overweight.
        Obesity. 2009; 17: 1-8
        • Breaugh J.
        • Frye N.
        Work and family conflict: The importance of family-friendly employment practices and family-supportive supervisors.
        J Business Psychol. 2008; 22: 345-353
        • Berge J.M.
        • Jin S.W.
        • Hannan P.
        • Neumark-Sztainer D.
        Structural and Interpersonal characteristics of family meals: associations with Adolescent Body Mass Index and Dietary Patterns.
        J Acad Nutr Diet. 2013; 113: 816-822