OBJECTIVES

First, to leverage 15 years of longitudinal data, from child ages 2 to 17, to examine whether maternal depressive symptoms in early and middle childhood and in adolescence predict their child’s unhealthy behaviors during adolescence. Second, to examine whether the timing of maternal depressive symptoms or specific unhealthy behaviors matter and whether child depressive symptoms and body mass index explain these associations.

METHODS

Data came from a prospective-longitudinal community sample with multi-informant data (N = 213) from child ages 2 to17. A cumulative adolescent unhealthy behavior index was calculated, summing the presence of poor sleep, poor diet, physical inactivity, sedentary behavior, and smoking. Regression analyses examined associations of maternal depressive symptoms in early childhood (ages 2 to 5), middle childhood (ages 7 to 10), and adolescence (age 15) with adolescents’ unhealthy behaviors (ages 16 to17). Indirect effects of child depressive symptoms and body mass index were tested using a path model.

RESULTS

Adolescents’ unhealthy behaviors were common (eg, 2 out of 3 engaged in at least 1 unhealthy behavior). Higher levels of maternal depressive symptoms in middle childhood and adolescence were associated with adolescent engagement in more unhealthy behaviors at ages 16 to 17. Maternal depressive symptoms in early childhood were associated with adolescent unhealthy behaviors through indirect effects involving children’s depressive symptoms and continuity of maternal depressive symptoms.

CONCLUSIONS

Maternal depressive symptoms are associated with the number of adolescent unhealthy behaviors, both directly and indirectly. Promoting mothers’ mental health can be crucial for promoting children’s health behaviors and health.

What’s Known on This Subject

Cross-sectional studies document an association of maternal depressive symptoms with children’s select unhealthy behaviors. These associations are rarely studied longitudinally, and the roles of the developmental timing of maternal depression and potential pathways in these associations are not understood.

What This Study Adds

This study finds longitudinal associations of maternal depressive symptoms in middle childhood and adolescence with adolescents’ later cumulative unhealthy behaviors, and indirect pathways from early childhood maternal depressive symptoms through continuity of maternal depression and child depressive symptoms.

Unhealthy behaviors – including poor sleep, unhealthy diet, physical inactivity, sedentary behavior, and smoking – constitute key risk factors for later cardiovascular disease.1,2  During adolescence, new opportunities for engaging in unhealthy behaviors arise. Despite the many public health efforts made to reduce the rates of adolescents’ unhealthy behaviors, they have been high in recent decades (except for smoking).1,3  To prevent these behaviors, further research is required to identify malleable targets. Maternal depressive symptoms constitute 1 such potential target. About 1 in 5 United States children is exposed to their mothers’ clinical depression at some point and many more are exposed to subthreshold depressive symptoms.4,5  Core depressive symptoms include lack of energy and motivation. Such anhedonia constitutes a significant barrier to maternal engagement in health-promoting activities. For example, mothers with depression might lack the energy to be active with their child(ren) and model unhealthy habits, such as maladaptive sleep patterns or unhealthy diet.68 

Cross-sectional research has documented associations of maternal depression with children’s poor sleep,9,10  unhealthy diet,11  physical inactivity, increased sedentary time,12  and smoking.13  However, the above research does not extend to adolescents’ unhealthy behaviors14  and typically limits its focus to 1 unhealthy behavior at-a-time, rather than the real-world phenomenon of accumulated unhealthy behaviors.15  It is unlikely that any single unhealthy behavior will result in chronic disease risk early in the life course, but their accumulation might set the stage for chronic disease development. This is supported by research suggesting that with an increasing number of unhealthy behaviors, young people’s cardiometabolic risk profile also increases.15  Similarly, cumulative risk research for socioemotional development has shown that the accumulation of risk factors is more strongly associated with adjustment than individual risk factors.16  This cumulative approach aims to capture the pervasive and heterogeneous associations of maternal depressive symptoms with adolescents’ unhealthy behaviors and to represent the reality of clustered unhealthy behaviors.15,17 

Owing to its 1-time assessments, the above cross-sectional research is not informative about the role of developmental timing of maternal depressive symptoms in adolescents’ unhealthy behaviors. Nevertheless, some researchers have viewed adolescence as a sensitive period during which adolescents’ susceptibility to (negative) external influences might increase.14  In addition, more recent exposure compared with exposures that occurred a longer time ago (eg, during childhood) might be more strongly linked to outcomes.18 

Furthermore, not much is known about the mechanisms underlying these associations. Maternal depression can impact maternal behavior (eg, modeling unhealthy behaviors, lack of encouragement, and scaffolding) and, hence, a child’s unhealthy behavior. Maternal depression may also influence child unhealthy behaviors by influencing children’s adjustment, that then goes on to influence child unhealthy behaviors. This could include the intergenerational transmission of depression,4,19  or child overweight or obesity.8,20,21  Maternal depression is associated with child depression,4  which in turn is a risk factor for adolescents’ unhealthy behaviors.19  In addition, maternal depressive symptoms have also been linked to a higher body mass index (BMI) in children,8  which clusters with unhealthy behaviors.21  But these links have only been shown individually, without testing longitudinal mediation models.

The current study addresses these limitations. It leverages data from a prospective-longitudinal community study (N = 213) spanning 15 years (child ages 2–17 years) to examine 4 research questions. (1) Are maternal depressive symptoms in early childhood (child ages 2 and 5), middle childhood (child ages 7 and 10), and adolescence (child age 15) linked to the accumulation of unhealthy behaviors (poor sleep, poor diet, physical inactivity, sedentary behavior, and smoking) at ages 16 and 17? (2) Does the timing of maternal depressive symptoms matter? (3) Do child depressive symptoms and BMI constitute pathways from maternal depressive symptoms to adolescent unhealthy behaviors? (4) Are there sex differences in these associations? The results could point to new avenues for reducing or preventing the public health burden stemming from adolescents’ unhealthy behaviors.

The data came from the RIGHT Track project, a prospective-longitudinal multi-informant community study conducted in North Carolina. This study followed children (N = 447) and their primary caregivers from age 2 to young adulthood. The participating families were recruited through child day-care centers, the County Health Department, and the local Women, Infants and Children program. The study was approved by the University Institutional Review Board at the University of North Carolina Greensboro, North Carolina.

Here, we focus on a subsample of participants who participated in the RIGHT Track health project22  at least once during adolescence (average age at the 2 adolescent assessments: 16.62, standard deviation [SD] = 0.39, and 17.65, SD = 0.34). More information about the original sample and attrition analyses is provided in the Supplemental Material. In short, those with lower socioeconomic status (SES) at age 2 were less likely to participate at age 7; males were less likely to participate than females at ages 15, 16, and 17.

To reduce bias caused by attrition and to maximize statistical power, multiple imputation (m = 20) using chained equations was implemented.23  All variables in the analyses were included in the imputation model. The results were pooled across datasets following Rubin’s rules.24 

The analytic sample comprised 213 participants, of which 42% were male. In addition, 65% were White, 29% were Black, and 6% were of mixed or other race and ethnicity, consistent with the community from which the sample was drawn. The families belonged to all social strata (Hollingshead index25  range: 14–66; mean = 39.88, SD = 10.42; this composite 4-factor index considers education, occupation, and sex, with scores 40–54 representing middle class).

Below, we briefly summarize the measures adopted in this study. Detailed information about these measures is presented in Table 1. The overall procedures for the health data collection are detailed elsewhere.22 

TABLE 1

Description of Measures Used in This Study

ConstructMeasures DescriptionCronbach’s αCut-Off Based On:
Maternal depressive symptoms Maternal depressive symptoms were self-reported on the depression subscale of the Symptom Checklist 90 Revised (SCL-90R).39  The SCL-90R is a well-validated broad measure of psychopathology in community and clinical samples. Its subscales show high and selective convergence with other scales assessing depressive symptoms, such as the General Health Questionnaire.4042  Mothers rated whether 13 symptoms (eg, “feeling low in energy or slowed down,” “feelings of worthlessness”) had caused them distress over the previous 7 d on a 5-point scale from 0 = “not at all” to 4 = “extremely.” The items were averaged. 0.88 (age 2) T-scores above 63 were used for reporting the likelihood of clinically significant symptoms in the results section. 
0.90 (age 5) 
0.88 (age 7) 
0.92 (age 10) 
0.92 (age 15) 
Adolescent unhealthy behaviors 
Poor sleep quality Poor sleep quality was assessed with the total score of the Pittsburgh Sleep Quality Index (PSQI).43  Seven components (ie, subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbance, use of sleep medication, and daytime dysfunction) were derived from a 19-item questionnaire. Items were rated on a 4-point scale from 0 = “not during the past month,” “very good,” or ”no problem at all” to 3 = “3 or more times a week,” “very bad,” or “a very big problem,” depending on the question. For each subscale, a score ranging from 0 to 3 was derived, as described in Buysse et al43  For the total score, the components were summed (range: 0–21). — PSQI cut-off (values >5 on the total score indicates poor sleep). A score greater than 5 indicates poor sleep and yields a sensitivity index of 89.6% and a specificity index of 86.5% when distinguishing healthy controls from patients with sleep disorders. 
Unhealthy diet quality Unhealthy dietary quality was assessed using the Food Habits Checklist (FHC).44  The FHC consists of 23 true or false statements asking about respondents’ intake of fruits, vegetables, and energy-dense foods (eg, “I usually avoid eating fried foods,” “I usually eat a dessert if there is one available”). Responses were first averaged and then reversed; thus, higher scores indicated unhealthier dietary habits. 0.84 (age 16)
0.86 (age 17) 
The quartile with the unhealthiest scores was coded to indicate poor diet on a binary variable. 
Physical activity Physical inactivity was assessed with the Godin-Shepard Leisure-Time Physical Activity Questionnaire, which has shown acceptable validity and reliability in various settings.45,46  The participants reported how many hours a week they engaged in moderate (eg, fast walking, baseball, tennis, easy bicycling, volleyball, badminton, easy swimming, dancing, and skiing) and strenuous (eg, running, jogging, hockey, football, soccer, squash, basketball, martial arts, roller skating, vigorous swimming, vigorous long distance bicycling, and cross-country skiing) physical activity. Outlying cases (>7) were winsorized to avoid biased parameter estimates (number of winsorized cases: 7 at age 16 and 2 at age 17 for moderate physical activity and 5 at age 16 and 9 at age 17 for strenuous physical activity). Moderate and strenuous physical activities were weighted to reflect metabolic equivalents using the Godin formula: (5x moderate activity) + (9x strenuous activity). — A binary physical inactivity variable was created, with a score of <24 indicating physical inactivity, as recommended.45  
Sedentary Behavior Sedentary behavior was assessed with an adapted version of the Sedentary Behavior Questionnaire.47  Ten items, 5 concerning weekday and 5 concerning weekend behaviors, assessed how much time per day the participants spent on sedentary activities, including watching TV, doing homework, or time on mobile devices, on a scale from 0 = “less than 1 h” to 6 = “more than 5 h.” Weekday and weekend sedentary behaviors were weighted and summed. 0.79 (age 16) A binary sedentary behavior variable was created to reflect that >2 h of sedentary behavior a day (>14 h per week) constitutes a health risk. 
 0.69 (age 17) 
Smoking Smoking in the past 30 d was assessed as part of the Youth Risk Behavior Survey.48  — Adolescents who reported smoking at least once in the previous month in at least 1 assessment received a code of 1 on the binary smoking variable. 
Mediators 
Child depressive symptoms Child depressive symptoms were self-reported on the Children’s Depression Inventory.49  For each of the 27 items, the participants chose among 3 sentences to best describe their feelings in the past 2 weeks (eg, “I am sad once in a while,” “I am sad much of the time,” and “I am sad all the time”). 0.82 (age 7) — 
0.90 (age 10) 
0.88 (age 15) 
Child BMI To calculate raw BMI, height was measured in centimeters and wt was measured in kilograms using a digital scale. — — 
ConstructMeasures DescriptionCronbach’s αCut-Off Based On:
Maternal depressive symptoms Maternal depressive symptoms were self-reported on the depression subscale of the Symptom Checklist 90 Revised (SCL-90R).39  The SCL-90R is a well-validated broad measure of psychopathology in community and clinical samples. Its subscales show high and selective convergence with other scales assessing depressive symptoms, such as the General Health Questionnaire.4042  Mothers rated whether 13 symptoms (eg, “feeling low in energy or slowed down,” “feelings of worthlessness”) had caused them distress over the previous 7 d on a 5-point scale from 0 = “not at all” to 4 = “extremely.” The items were averaged. 0.88 (age 2) T-scores above 63 were used for reporting the likelihood of clinically significant symptoms in the results section. 
0.90 (age 5) 
0.88 (age 7) 
0.92 (age 10) 
0.92 (age 15) 
Adolescent unhealthy behaviors 
Poor sleep quality Poor sleep quality was assessed with the total score of the Pittsburgh Sleep Quality Index (PSQI).43  Seven components (ie, subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbance, use of sleep medication, and daytime dysfunction) were derived from a 19-item questionnaire. Items were rated on a 4-point scale from 0 = “not during the past month,” “very good,” or ”no problem at all” to 3 = “3 or more times a week,” “very bad,” or “a very big problem,” depending on the question. For each subscale, a score ranging from 0 to 3 was derived, as described in Buysse et al43  For the total score, the components were summed (range: 0–21). — PSQI cut-off (values >5 on the total score indicates poor sleep). A score greater than 5 indicates poor sleep and yields a sensitivity index of 89.6% and a specificity index of 86.5% when distinguishing healthy controls from patients with sleep disorders. 
Unhealthy diet quality Unhealthy dietary quality was assessed using the Food Habits Checklist (FHC).44  The FHC consists of 23 true or false statements asking about respondents’ intake of fruits, vegetables, and energy-dense foods (eg, “I usually avoid eating fried foods,” “I usually eat a dessert if there is one available”). Responses were first averaged and then reversed; thus, higher scores indicated unhealthier dietary habits. 0.84 (age 16)
0.86 (age 17) 
The quartile with the unhealthiest scores was coded to indicate poor diet on a binary variable. 
Physical activity Physical inactivity was assessed with the Godin-Shepard Leisure-Time Physical Activity Questionnaire, which has shown acceptable validity and reliability in various settings.45,46  The participants reported how many hours a week they engaged in moderate (eg, fast walking, baseball, tennis, easy bicycling, volleyball, badminton, easy swimming, dancing, and skiing) and strenuous (eg, running, jogging, hockey, football, soccer, squash, basketball, martial arts, roller skating, vigorous swimming, vigorous long distance bicycling, and cross-country skiing) physical activity. Outlying cases (>7) were winsorized to avoid biased parameter estimates (number of winsorized cases: 7 at age 16 and 2 at age 17 for moderate physical activity and 5 at age 16 and 9 at age 17 for strenuous physical activity). Moderate and strenuous physical activities were weighted to reflect metabolic equivalents using the Godin formula: (5x moderate activity) + (9x strenuous activity). — A binary physical inactivity variable was created, with a score of <24 indicating physical inactivity, as recommended.45  
Sedentary Behavior Sedentary behavior was assessed with an adapted version of the Sedentary Behavior Questionnaire.47  Ten items, 5 concerning weekday and 5 concerning weekend behaviors, assessed how much time per day the participants spent on sedentary activities, including watching TV, doing homework, or time on mobile devices, on a scale from 0 = “less than 1 h” to 6 = “more than 5 h.” Weekday and weekend sedentary behaviors were weighted and summed. 0.79 (age 16) A binary sedentary behavior variable was created to reflect that >2 h of sedentary behavior a day (>14 h per week) constitutes a health risk. 
 0.69 (age 17) 
Smoking Smoking in the past 30 d was assessed as part of the Youth Risk Behavior Survey.48  — Adolescents who reported smoking at least once in the previous month in at least 1 assessment received a code of 1 on the binary smoking variable. 
Mediators 
Child depressive symptoms Child depressive symptoms were self-reported on the Children’s Depression Inventory.49  For each of the 27 items, the participants chose among 3 sentences to best describe their feelings in the past 2 weeks (eg, “I am sad once in a while,” “I am sad much of the time,” and “I am sad all the time”). 0.82 (age 7) — 
0.90 (age 10) 
0.88 (age 15) 
Child BMI To calculate raw BMI, height was measured in centimeters and wt was measured in kilograms using a digital scale. — — 

—, not applicable.

Maternal depressive symptoms observed in the previous 7 days were self-reported at child ages 2, 5, 7, 10, and 15. Maternal depressive symptoms at ages 2 and 5 and those at ages 7 and 10 were averaged, respectively, to form indicators of early- and middle-childhood maternal depressive symptoms.

Adolescents’ unhealthy behaviors included poor sleep, poor dietary quality, physical inactivity, sedentary behavior, and smoking. These were self-reported at ages 16 and 17 and continuous measures were averaged. For dichotomous variables, an either/or rule was applied to derive 1 observation from the 2 assessments. For each unhealthy behavior, a binary variable was created based on either theoretical or statistical cut-offs, which are described in Table 1. These binary variables were then summed to form an index of the accumulation of unhealthy behaviors, with 5 being the highest possible score. This approach is consistent with previous studies that have examined multiple unhealthy behaviors with summed binary indicators.26 

Child depressive symptoms at ages 7, 10, and 15 were self-reported (no child-rated measures of depressive symptoms were available for early childhood). Scores at ages 7 and 10 were averaged to represent depressive symptoms in middle childhood, and those at age 15 represented adolescent depressive symptoms. Height and weight were assessed during laboratory visits at ages 5, 7, 10, and 15 and raw BMI was calculated. BMI at ages 5 and 15 represented BMI in early childhood and adolescence, respectively, and those at ages 7 and 10 were averaged to represent BMI in middle childhood.

Sex was coded as 0 = female and 1 = male. Race and ethnicity was coded as 0 = White (65%) and 1 = other (35%). Family SES was measured using the Hollingshead index.25  For 2-parent households, an index was calculated for each parent and averaged. Single motherhood was mother-reported as 0 = (re)married and 1 = single, separated, or divorced.

First, we specified linear regressions to examine the associations of maternal depressive symptoms in early childhood, middle childhood, and adolescence with the number of adolescent unhealthy behaviors. The regression models included the variable and demographic covariates of maternal depressive symptoms. For ease of interpretation, 3 regression models were estimated, 1 for each developmental period.

Second, by using a path model, we examined whether maternal depressive symptoms were associated with adolescents’ unhealthy behaviors through children’s depressive symptoms and BMI (see Supplemental Fig 3 for a graphical representation of the path model). The model included stability paths of maternal and child depressive symptoms and child BMI. Maternal depressive symptoms at 1 time point were specified as a predictor of children’s depressive symptoms and BMI at the subsequent time point (eg, from maternal depressive symptoms in early childhood to child depressive symptoms in middle childhood). Maternal and children’s depressive symptoms and BMI in adolescence were specified as predictors of adolescents’ unhealthy behaviors. Variables assessed at the same time were allowed to covary. All demographic covariates were included in the original model, and only covariates with P <.10 were retained in the final model. The retained covariates are listed in Supplemental Table 5.

Good model fit was evaluated based on a nonsignificant χ2 value, comparative fit index ≥0.90, root mean square error approximation ≤0.08, and square root mean residual ≤0.08.27,28  The significance of indirect paths was evaluated using the product of coefficients method. Robust confidence intervals were estimated using the Monte Carlo method (20 000 draws), a recommended alternative to bias-corrected bootstrap confidence intervals when the latter are not feasible (eg, when working with multiple imputed datasets).29 

Third, we explored sex differences in the associations between maternal depressive symptoms and adolescents’ unhealthy behaviors by including the interaction term of sex and maternal depressive symptoms in the regression models. We did not examine sex differences in the path model because of the low sample size when splitting the sample.

Follow-up analyses were conducted to examine whether the association of maternal depressive symptoms and adolescents’ unhealthy behaviors was driven by a specific unhealthy behavior. We specified regression models with every unhealthy behavior (binary variables) separately. A follow-up analysis of the described path model also included cross-paths between child depressive symptoms and child BMI and from child depressive symptoms and child BMI to maternal depressive symptoms (Supplemental Tables 5 and 6 and Supplemental Figs 3 and 4).

All analyses were performed in R using the mice, lavaan, and semTool packages. This allowed us to use maximum likelihood estimation with robust standard errors, which provides reliable estimates in the presence of non-normality.28 

Table 2 presents the descriptive statistics for the overall sample and for males and females. The raw scores of the average levels of maternal depressive symptoms were low. Yet, 9.1%, 5.7%, and 13.4% of mothers met criteria for potential clinically significant symptoms in early childhood, middle childhood, and adolescence, respectively, based on the SCL-90 T-score cut-offs. This is comparable to mothers in other community samples.30  Adolescents’ unhealthy behaviors were prevalent, with two-thirds reporting ≥1 unhealthy behavior and one-fifth reporting ≥3 unhealthy behaviors (Fig 1). ∼1 in 4 adolescents reported poor sleep, 1 in 4 reported an unhealthy diet, 2 in 5 reported physical inactivity, 1 in 7 were often sedentary, and 1 in 7 smoked at some time point. No sex differences emerged in the overall number of unhealthy behaviors; however, males were more likely to report a poor diet, whereas females were more likely to report physical inactivity and sedentary behavior.

FIGURE 1

Number of unhealthy behaviors in adolescence.

FIGURE 1

Number of unhealthy behaviors in adolescence.

Close modal
TABLE 2

Descriptive Statistics Based on Observed Data

OverallMalesFemalesSex Diff.
Na%MeanSDNa%MeanSDNa%MeanSDP
MDS early childhood (ages 2 and 5) 164 — 0.56 0.47 69 — 0.58 0.50 95 — 0.54 0.45 .538 
MDS middle childhood (ages 7 and 10) 175 — 0.41 0.47 73 — 0.39 0.45 102 — 0.43 0.48 .550 
MDS age 15 179 — 0.52 0.67 75 — 0.57 0.78 104 — 0.49 0.57 .393 
Unhealthy behavior ages 16–17b 206 — 1.19 1.08 87 — 1.14 1.01 119 — 1.24 1.13 .525 
Poor sleep cut-off 207 28 — — 88 21 — — 119 33 — — .071 
Unhealthy eating cut-off 213 24 — — 90 33 — — 123 18 — — .015 
Physical inactivity cut-off 213 39 — — 90 28 — — 123 47 — — .006 
Sedentary behavior cut-off 212 15 — — 89 12 — — 123 17 — — .011 
Smoking 212 15 — — 90 19 — — 122 12 — — .189 
Middle childhood depression (ages 7 and 10) 151 — 0.25 0.20 61 — 0.23 0.18 90 — 0.26 0.21 .351 
Adolescent depression age 15 176 — 0.27 0.24 71 — 0.24 0.21 105 — 0.28 0.25 .207 
Early childhood BMI (age 5) 170 — 16.40 2.27 73 — 16.47 2.29 97 — 16.35 2.26 .719 
Middle childhood BMI (ages 7 and 10) 127 — 19.29 4.08 51 — 18.9 4.4 76 — 19.55 3.86 .383 
Adolescent BMI age 15 146 — 23.76 4.55 59 — 23.49 4.73 87 — 23.94 4.44 .557 
OverallMalesFemalesSex Diff.
Na%MeanSDNa%MeanSDNa%MeanSDP
MDS early childhood (ages 2 and 5) 164 — 0.56 0.47 69 — 0.58 0.50 95 — 0.54 0.45 .538 
MDS middle childhood (ages 7 and 10) 175 — 0.41 0.47 73 — 0.39 0.45 102 — 0.43 0.48 .550 
MDS age 15 179 — 0.52 0.67 75 — 0.57 0.78 104 — 0.49 0.57 .393 
Unhealthy behavior ages 16–17b 206 — 1.19 1.08 87 — 1.14 1.01 119 — 1.24 1.13 .525 
Poor sleep cut-off 207 28 — — 88 21 — — 119 33 — — .071 
Unhealthy eating cut-off 213 24 — — 90 33 — — 123 18 — — .015 
Physical inactivity cut-off 213 39 — — 90 28 — — 123 47 — — .006 
Sedentary behavior cut-off 212 15 — — 89 12 — — 123 17 — — .011 
Smoking 212 15 — — 90 19 — — 122 12 — — .189 
Middle childhood depression (ages 7 and 10) 151 — 0.25 0.20 61 — 0.23 0.18 90 — 0.26 0.21 .351 
Adolescent depression age 15 176 — 0.27 0.24 71 — 0.24 0.21 105 — 0.28 0.25 .207 
Early childhood BMI (age 5) 170 — 16.40 2.27 73 — 16.47 2.29 97 — 16.35 2.26 .719 
Middle childhood BMI (ages 7 and 10) 127 — 19.29 4.08 51 — 18.9 4.4 76 — 19.55 3.86 .383 
Adolescent BMI age 15 146 — 23.76 4.55 59 — 23.49 4.73 87 — 23.94 4.44 .557 

—, not applicable; MDS, maternal depressive symptoms.

a

Indicates number of participants with valid data points.

b

For the descriptive statistics and correlations, the unhealthy behavior composite was created from the mean of unhealthy behaviors at ages 16 and 17, where possible. If the participants only participated in 1 adolescent’s health assessment, unhealthy behavior data from that assessment were taken.

The bivariate correlations presented in Table 3 suggest that more maternal depressive symptoms in middle childhood and adolescence were associated with a higher number of unhealthy behaviors. Maternal depressive symptoms in adolescence were associated with each unhealthy behavior. Bivariate correlations by sex suggested that maternal depressive symptoms in middle childhood and adolescence were associated with more unhealthy behaviors in females compared with males.

TABLE 3

Correlations for Overall Sample and by Sex Based on Observed Data

1234567891011121314
Overall sample               
 (1) MDS early childhood (ages 2 and 5) — — — — — — — — — — — — — — 
 (2) MDS middle childhood (ages 7 and 10) 0.71*** — — — — — — — — — — — — — 
 (3) MDS age 15 0.65*** 0.68*** — — — — — — — — — — — — 
 (4) Unhealthy behavior age 16–17a 0.05 0.22** 0.26** — — — — — — — — — — — 
 (5) Poor sleep 0.12 0.25** 0.19** 0.53*** — — — — — — — — — — 
 (6) Unhealthy eating 0.10 0.08 0.15* 0.45*** 0.05 — — — — — — — — — 
 (7) Physical inactivity 0.04 0.13+ 0.15* 0.52*** 0.17* 0.19** — — — — — — — — 
 (8) Sedentary behavior −0.04 −0.11 −0.16* 0.30*** 0.03 0.05 0.15* — — — — — — — 
 (9) Smoking 0.01 0.07 0.28*** 0.47*** 0.23** 0.10 0.11 −0.03 — — — — — — 
 (10) CDS middle childhood (ages 7 and 10) 0.22* 0.26** 0.28** 0.25** 0.23** 0.16+ 0.12 −0.01 0.16+ — — — — — 
 (11) CDS age 15 0.20* 0.16* 0.33*** 0.31*** 0.24** 0.21** 0.19* −0.03 0.19** 0.24** — — — — 
 (12) Child BMI early childhood (age 5) 0.04 −0.08 0.03 0.04 0.13+ −0.05 0.09 −0.05 0.05 0.23** 0.05 — — — 
 (13) Child BMI middle childhood (ages 7 and 10) 0.01 −0.16+ 0.06 0.05 0.17+ −0.11 0.13 0.06 0.01 0.22* 0.04 0.84** — — 
 (14) Adolescent BMI age 15 −0.05 −0.07 −0.03 0.10 0.06 −0.01 0.17* 0.13 0.12 0.24** -0.00 0.70*** 0.76*** — 
By sex               
 (1) MDS early childhood (ages 2 and 5) — 0.72*** 0.61*** 0.12 0.14 0.19 0.10 0.02 −0.08 0.17 0.17 −0.02 −0.15 0.03 
 (2) MDS middle childhood (ages 7 and 10) 0.69*** — 0.71*** 0.33** 0.42*** 0.13 0.25* −0.09 0.08 0.34** 0.17 0.02 −0.17 −0.02 
 (3) MDS age 15 0.69*** 0.68*** — 0.35*** 0.28** 0.27** 0.29** −0.12 0.25** 0.29** 0.36*** −0.10 −0.22+ −0.01 
 (4) Unhealthy behavior age 16–17a −0.03 0.03 0.19 — 0.58*** 0.50*** 0.61*** 0.38*** 0.47*** 0.31* 0.29** −0.02 −0.02 0.10 
 (5) Poor sleep 0.11 −0.03 0.15 0.42*** — 0.08 0.24** 0.04 0.29** 0.27* 0.15 0.02 0.06 0.14 
 (6) Unhealthy eating −0.05 0.03 −0.03 0.44*** 0.11 — 0.44*** 0.12 0.19* 0.27** 0.38*** −0.04 −0.11 0.01 
 (7) Physical inactivity −0.00 −0.04 0.07 0.41*** −0.02 0.07 — 0.16+ 0.20* 0.20+ 0.26** 0.01 −0.03 0.10 
 (8) Sedentary behavior −0.08 −0.15 −0.19 0.19+ −0.07 0.07 0.06 — 0.03 −0.06 −0.06 −0.06 0.07 0.15 
 (9) Smoking 0.07 0.08 0.30* 0.49*** 0.23* −0.07 0.08 −0.05 — 0.15 0.07 0.07 0.06 0.26* 
 (10) CDS middle childhood (ages 7 and 10) 0.31* 0.07 0.27* 0.13 0.16 −0.02 −0.03 0.02 0.20 — 0.16 0.32** 0.21+ 0.35** 
 (11) CDS age 15 0.26+ 0.14 0.34** 0.35** 0.39** −0.04 0.03 −0.05 0.41*** 0.39** — −0.00 −0.09 −0.08 
 (12) Child BMI early childhood (age 5) 0.10 −0.22+ 0.14 0.14 0.33** −0.11 0.21+ −0.02 0.02 0.10 0.13 — 0.81*** 0.70*** 
 (13) Child BMI middle childhood (ages 7 and 10) 0.21 −0.14 0.35* 0.16 0.30* −0.07 0.27+ 0.00 −0.02 0.22 0.25 0.88*** — 0.74*** 
 (14) Adolescent BMI age 15 −0.15 −0.16 −0.06 0.08 −0.08 0.00 0.26* 0.09 −0.01 0.03 0.11 0.71*** 0.81*** — 
1234567891011121314
Overall sample               
 (1) MDS early childhood (ages 2 and 5) — — — — — — — — — — — — — — 
 (2) MDS middle childhood (ages 7 and 10) 0.71*** — — — — — — — — — — — — — 
 (3) MDS age 15 0.65*** 0.68*** — — — — — — — — — — — — 
 (4) Unhealthy behavior age 16–17a 0.05 0.22** 0.26** — — — — — — — — — — — 
 (5) Poor sleep 0.12 0.25** 0.19** 0.53*** — — — — — — — — — — 
 (6) Unhealthy eating 0.10 0.08 0.15* 0.45*** 0.05 — — — — — — — — — 
 (7) Physical inactivity 0.04 0.13+ 0.15* 0.52*** 0.17* 0.19** — — — — — — — — 
 (8) Sedentary behavior −0.04 −0.11 −0.16* 0.30*** 0.03 0.05 0.15* — — — — — — — 
 (9) Smoking 0.01 0.07 0.28*** 0.47*** 0.23** 0.10 0.11 −0.03 — — — — — — 
 (10) CDS middle childhood (ages 7 and 10) 0.22* 0.26** 0.28** 0.25** 0.23** 0.16+ 0.12 −0.01 0.16+ — — — — — 
 (11) CDS age 15 0.20* 0.16* 0.33*** 0.31*** 0.24** 0.21** 0.19* −0.03 0.19** 0.24** — — — — 
 (12) Child BMI early childhood (age 5) 0.04 −0.08 0.03 0.04 0.13+ −0.05 0.09 −0.05 0.05 0.23** 0.05 — — — 
 (13) Child BMI middle childhood (ages 7 and 10) 0.01 −0.16+ 0.06 0.05 0.17+ −0.11 0.13 0.06 0.01 0.22* 0.04 0.84** — — 
 (14) Adolescent BMI age 15 −0.05 −0.07 −0.03 0.10 0.06 −0.01 0.17* 0.13 0.12 0.24** -0.00 0.70*** 0.76*** — 
By sex               
 (1) MDS early childhood (ages 2 and 5) — 0.72*** 0.61*** 0.12 0.14 0.19 0.10 0.02 −0.08 0.17 0.17 −0.02 −0.15 0.03 
 (2) MDS middle childhood (ages 7 and 10) 0.69*** — 0.71*** 0.33** 0.42*** 0.13 0.25* −0.09 0.08 0.34** 0.17 0.02 −0.17 −0.02 
 (3) MDS age 15 0.69*** 0.68*** — 0.35*** 0.28** 0.27** 0.29** −0.12 0.25** 0.29** 0.36*** −0.10 −0.22+ −0.01 
 (4) Unhealthy behavior age 16–17a −0.03 0.03 0.19 — 0.58*** 0.50*** 0.61*** 0.38*** 0.47*** 0.31* 0.29** −0.02 −0.02 0.10 
 (5) Poor sleep 0.11 −0.03 0.15 0.42*** — 0.08 0.24** 0.04 0.29** 0.27* 0.15 0.02 0.06 0.14 
 (6) Unhealthy eating −0.05 0.03 −0.03 0.44*** 0.11 — 0.44*** 0.12 0.19* 0.27** 0.38*** −0.04 −0.11 0.01 
 (7) Physical inactivity −0.00 −0.04 0.07 0.41*** −0.02 0.07 — 0.16+ 0.20* 0.20+ 0.26** 0.01 −0.03 0.10 
 (8) Sedentary behavior −0.08 −0.15 −0.19 0.19+ −0.07 0.07 0.06 — 0.03 −0.06 −0.06 −0.06 0.07 0.15 
 (9) Smoking 0.07 0.08 0.30* 0.49*** 0.23* −0.07 0.08 −0.05 — 0.15 0.07 0.07 0.06 0.26* 
 (10) CDS middle childhood (ages 7 and 10) 0.31* 0.07 0.27* 0.13 0.16 −0.02 −0.03 0.02 0.20 — 0.16 0.32** 0.21+ 0.35** 
 (11) CDS age 15 0.26+ 0.14 0.34** 0.35** 0.39** −0.04 0.03 −0.05 0.41*** 0.39** — −0.00 −0.09 −0.08 
 (12) Child BMI early childhood (age 5) 0.10 −0.22+ 0.14 0.14 0.33** −0.11 0.21+ −0.02 0.02 0.10 0.13 — 0.81*** 0.70*** 
 (13) Child BMI middle childhood (ages 7 and 10) 0.21 −0.14 0.35* 0.16 0.30* −0.07 0.27+ 0.00 −0.02 0.22 0.25 0.88*** — 0.74*** 
 (14) Adolescent BMI age 15 −0.15 −0.16 −0.06 0.08 −0.08 0.00 0.26* 0.09 −0.01 0.03 0.11 0.71*** 0.81*** — 

For individual unhealthy behaviors, the continuous mean score was used. Correlations for males are shown below the diagonal, and those for females are shown above the diagonal. The range of the number of observations is n = 113 to 213 for the overall sample, n = 47 to 90 for males, and n = 66 to 123 for females. CDS, child depressive symptoms; MDS, maternal depressive symptoms; —, correlation of 1.00 or duplicate information.

a

For the descriptive statistics and correlations, the unhealthy behavior composite was created from the mean of unhealthy behaviors at ages 16 and 17, where possible. If the participants only participated in 1 adolescent health assessment, unhealthy behavior data were used from that assessment only.

+

P < .01;

*

P < .05;

**

P < .01;

***

P < .001

Regression analyses showed that maternal depressive symptoms in middle childhood and adolescence were significantly associated with adolescents’ unhealthy behaviors after adjusting for sociodemographic characteristics (Table 4). No significant interactions emerged between sex and maternal depressive symptoms.

TABLE 4

Linear Regression Models of Composite Health Behavior for the Overall Sample

Sex Interaction
Timing of Maternal Depressive Symptomsβpp
Early childhood 0.08 .338 .553 
Middle childhood 0.18 .015 .055 
Adolescence 0.26 .001 .151 
Sex Interaction
Timing of Maternal Depressive Symptomsβpp
Early childhood 0.08 .338 .553 
Middle childhood 0.18 .015 .055 
Adolescence 0.26 .001 .151 

Predictions of maternal depressive symptoms were examined in separate regression models.

Figure 2 depicts the standardized coefficients from the path analysis. These suggest that the total indirect effect from maternal depressive symptoms to children’s depressive symptoms (ie, the sum of the path from maternal depressive symptoms in early childhood to child depressive symptoms in middle childhood to child depressive symptoms in adolescence, and the path from maternal depressive symptoms in early childhood to maternal depressive symptoms in middle childhood to child depressive symptoms in adolescence) to unhealthy behaviors was significant (B = 0.08, 95% confidence interval [CI] = 0.01 to 0.18). The indirect effect observed through the continuity of maternal depressive symptoms was also significant (B = 0.19, 95% CI = 0.01 to 0.39). No indirect paths emerged through child BMI.

FIGURE 2

Standardized coefficients from the path model.

Model fit χ2 = 52.44 (29), comparative fit indexr = .005, root mean square error approximation r = .065, square root mean residual = .067. The analysis was adjusted for sex and race and ethnicity at age 2 (Supplemental Table 5) +P < .01; *P < .05; **P < .01; ***P < .001

FIGURE 2

Standardized coefficients from the path model.

Model fit χ2 = 52.44 (29), comparative fit indexr = .005, root mean square error approximation r = .065, square root mean residual = .067. The analysis was adjusted for sex and race and ethnicity at age 2 (Supplemental Table 5) +P < .01; *P < .05; **P < .01; ***P < .001

Close modal

The follow-up analyses are presented in Supplemental Table 5 and Supplemental Fig 4. The analyses of maternal depressive symptoms and specific unhealthy behaviors suggested that maternal depressive symptoms in adolescence were associated with each unhealthy behaviors included in our cumulative unhealthy behavior index, except sedentary behavior. The follow-up path model, which included cross-paths between all constructs, confirmed the findings described above.

This study leveraged 15 years of data from a prospective-longitudinal community study to examine whether exposure to maternal depressive symptoms in 3 developmental periods – early childhood, middle childhood, and adolescence – are associated with the accumulation of unhealthy behaviors in adolescence.

First, higher levels of maternal depressive symptoms in middle childhood and adolescence were directly associated with the accumulation of adolescent unhealthy behaviors. The association between adolescent maternal depressive symptoms and unhealthy behaviors was consistent with previous findings suggesting that adolescence is a sensitive period for shaping behavior and that recency of risk is relevant.14,18  Because maternal depressive symptoms measured in adolescence were also the most recent to adolescent unhealthy behaviors, we could not disentangle recency from developmental timing.

Second, although maternal depressive symptoms in early childhood and adolescents’ unhealthy behaviors >10 years later were not directly associated, indirect paths from maternal depressive symptoms in early childhood to unhealthy behaviors emerged through child depressive symptoms and the continuity of maternal depressive symptoms. These findings are consistent with a large body of literature on the intergenerational transmission of depressive symptoms, which shows that children of mothers with high levels of depression tend to experience more depressive symptoms themselves.4,5  These results are further consistent with findings indicating that in the context of subclinically elevated maternal depressive symptoms; it is their chronicity that is associated with a higher risk for child adjustment difficulties.31,32  Going beyond the previous research, we showed that this is also the case for children’s unhealthy behaviors. Maternal depressive symptoms in adolescence remained a significant predictor of adolescents’ unhealthy behaviors when accounting for child depressive symptoms, suggesting that other mediating factors are also at play, such as maternal (unhealthy) behaviors.

Third, no indirect associations through children’s BMI emerged. In fact, BMI was uncorrelated with maternal depressive symptoms or cumulative unhealthy behaviors in this sample. Interindividual differences in BMI were highly stable across the 10 year study period. Other early-life factors, such as gestational weight gain, birth weight, and self-regulation, might have been more strongly associated with early childhood BMI than maternal depressive symptoms.33,34 

Fourth, no sex differences emerged in the association between maternal depressive symptoms and adolescents’ unhealthy behavior. This might be because of the importance of family functioning for unhealthy behavior for all children. Nevertheless, future research should continue to explore potential sex differences in risk factors for unhealthy behaviors, including the underlying pathways, as the bivariate correlations suggested that maternal depressive symptoms are more strongly associated with females’ than with males’ unhealthy behaviors.

Notably, all associations are small in size, which highlights that maternal depressive symptoms are not inevitably associated with unhealthy behaviors in adolescents. This suggests complex developmental processes to adolescent unhealthy behaviors that need to be examined in future research.

This analysis of long-term associations between maternal depressive symptoms and adolescents’ unhealthy behaviors has many strengths, including its community-representative prospective-longitudinal sample with multi-informant data and repeated measures of maternal depressive symptoms during 3 developmental periods.

Our findings should also be considered in light of some limitations. First, unhealthy behaviors were measured in adolescence only. This limited our ability to draw conclusions about the directionality of associations, as we were unable to examine whether unhealthy behaviors preceded maternal and child depressive symptoms. Our analysis did, however, include child BMI, which could be considered a proxy for previous unhealthy behaviors, but child BMI was uncorrelated with unhealthy behaviors in adolescence. Second, this study focused on the accumulation of unhealthy behaviors in adolescence, which could mask associations of maternal depression with specific unhealthy behaviors. Yet, our sensitivity analysis showed that maternal depressive symptoms in adolescence were associated with most adolescent unhealthy behaviors, suggesting that no single behavior drives the association, but that maternal depression is associated with the child’s unhealthy lifestyle as a whole.

Third, our sample was moderate in size and might have been underpowered to detect small associations and sex differences.35  Fourth, unhealthy behaviors were self-reported by participants using established questionnaires, which might have differed from objectively measured behavior.36  Fifth, although we applied longitudinal modeling, our design was correlational in nature and not conclusive regarding causality. Sixth, mothers’ psychiatric diagnoses and services use were not measured. Our findings are in line with those obtained from other community studies5 ; nevertheless, we cannot rule out that findings could have been (in part) driven by mothers with psychiatric diagnoses. Seventh, unmeasured confounders might exist. Most notably, the association between maternal and child depressive symptoms is partly attributable to biological risk.4  Such biological confounding can also underlie the associations between maternal depression and adolescent unhealthy behaviors.

In families from the community, maternal depressive symptoms in childhood and adolescence are associated with an increased risk of children’s accumulated unhealthy behaviors in adolescence. For maternal depressive symptoms in early childhood, this association is explained by the continuity of maternal depressive symptoms across time and its association with children’s depressive symptoms. Promoting mothers’ well-being and mental health is not only crucial for mothers themselves but can contribute to addressing 1 of the most pressing public health problems by promoting children’s healthy behaviors. Screening for parental and child mental health in primary care settings, such as during well-child visits, could be 1 way to identify affected families early and refer them to relevant resources and services, as well as an opportunity to initiate a conversation about unhealthy behaviors in adults and children.37,38 

We thank the parents and children who have repeatedly given their time and effort to participate in this research and are grateful to the entire RIGHT Track staff for their help in collecting, entering, and coding data.

Ms Bechtiger conceptualized and designed the study, conducted the data analysis, drafted the initial manuscript, designed the data analysis strategy, and critically reviewed the manuscript for important intellectual content; Dr Steinhoff conceptualized and designed the study, drafted the initial manuscript, designed the data analysis strategy, and critically reviewed the manuscript for important intellectual content; Dr Dollar conceptualized and designed the study and critically reviewed the manuscript for important intellectual content; Drs Keane, Shriver, and Wideman conceptualized and designed the study, designed the data collection instruments, and critically reviewed the manuscript for important intellectual content; Dr Calkins conceptualized and designed the study and the data collection instruments; Dr Shanahan conceptualized and designed the study, drafted the initial manuscript, designed the data collection instruments, designed the data analysis strategy, and critically reviewed the manuscript for important intellectual content; and all authors reviewed and revised the manuscript, approved the final manuscript as submitted, and agree to be accountable for all aspects of the work.

FUNDING: This research was supported by the National Institutes of Health (grant numbers MH 55625, MH 55584, MH 58144, HD 078346). The NIH had no role in the design and conduct of the study. Funded by the National Institutes of Health (NIH).

CONFLICT OF INTEREST DISCLOSURES: The authors have indicated they have no conflicts of interest to disclose.

BMI

body mass index

SES

socioeconomic status

1
Shay
CM
,
Ning
H
,
Daniels
SR
,
Rooks
CR
,
Gidding
SS
,
Lloyd-Jones
DM
.
Status of cardiovascular health in US adolescents: prevalence estimates from the National Health and Nutrition Examination Surveys (NHANES) 2005-2010
.
Circulation
.
2013
;
127
(
13
):
1369
1376
2
Mellendick
K
,
Shanahan
L
,
Wideman
L
,
Calkins
S
,
Keane
S
,
Lovelady
C
.
Diets rich in fruits and vegetables are associated with lower cardiovascular disease risk in adolescents
.
Nutrients
.
2018
;
10
(
2
):
136
3
Miech
R
,
Keyes
KM
,
O’Malley
PM
, %
Johnston
LD
.
The great decline in adolescent cigarette smoking since 2000: consequences for drug use among US adolescents
.
Tob Control
.
2020
;
29
(
6
):
638
643
4
Goodman
SH
,
Rouse
MH
,
Connell
AM
,
Broth
MR
,
Hall
CM
,
Heyward
D
.
Maternal depression and child psychopathology: a meta-analytic review
.
Clin Child Fam Psychol Rev
.
2011
;
14
(
1
):
1
27
5
Campbell
SB
,
Morgan-Lopez
AA
,
Cox
MJ
,
McLoyd
VC
.
A latent class analysis of maternal depressive symptoms over 12 years and offspring adjustment in adolescence
.
J Abnorm Psychol
.
2009
;
118
(
3
):
479
493
6
Goeglein
SK
,
Yatchmink
YE
.
Maternal depression is a public health crisis: the time to act is now
.
Pediatrics
.
2020
;
146
(
3
):
e2020010413
7
Shonkoff
JP
,
Boyce
WT
,
Levitt
P
,
Martinez
FD
,
McEwen
B
.
Leveraging the biology of adversity and resilience to transform pediatric practice
.
Pediatrics
.
2021
;
147
(
2
):
e20193845
8
Topham
GL
,
Page
MC
,
Hubbs-Tait
L
, et al
.
Maternal depression and socio-economic status moderate the parenting style/child obesity association
.
Public Health Nutr
.
2010
;
13
(
8
):
1237
1244
9
Teti
DM
,
Crosby
B
.
Maternal depressive symptoms, dysfunctional cognitions, and infant night waking: the role of maternal nighttime behavior
.
Child Dev
.
2012
;
83
(
3
):
939
953
10
El-Sheikh
M
,
Kelly
RJ
,
Bagley
EJ
,
Wetter
EK
.
Parental depressive symptoms and children’s sleep: the role of family conflict
.
J Child Psychol Psychiatry
.
2012
;
53
(
7
):
806
814
11
Thompson
SM
,
Jiang
L
,
Hammen
C
,
Whaley
SE
.
Association of maternal depressive symptoms and offspring physical health in low-income families
.
Matern Child Health J
.
2018
;
22
(
6
):
874
882
12
Conners
NA
,
Tripathi
SP
,
Clubb
R
, %
Bradley
RH
.
Maternal characteristics associated with television viewing habits of low-income preschool children
.
J Child Fam Stud
.
2007
;
16
(
3
):
415
425
13
Wickham
ME
,
Senthilselvan
A
,
Wild
TC
,
Hoglund
WLG
,
Colman
I
.
Maternal depressive symptoms during childhood and risky adolescent health behaviors
.
Pediatrics
.
2015
;
135
(
1
):
59
67
14
Dorn
LD
,
Hostinar
CE
,
Susman
EJ
, %
Pervanidou
P
.
Conceptualizing puberty as a window of opportunity for impacting health and well-being across the life Span
.
J Res Adolesc
.
2019
;
29
(
1
):
155
176
15
Kracht
CL
,
Champagne
CM
,
Hsia
DS
, et al
.
Association between meeting physical activity, sleep, and dietary guidelines and cardiometabolic risk factors and adiposity in adolescents
.
J Adolesc Health
.
2020
;
66
(
6
):
733
739
16
Evans
GW
,
Li
D
,
Whipple
SS
.
Cumulative risk and child development
.
Psychol Bull
.
2013
;
139
(
6
):
1342
1396
17
Dollar
JM
,
Calkins
SD
,
Shriver
L
,
Keane
SP
,
Shanahan
L
,
Wideman
L
.
Emerging self-regulatory skills in childhood predict cardiometabolic risk in adolescence
.
Compr Psychoneuroendocrinol
.
2021
;
7
:
100070
18
Shanahan
L
,
Copeland
WE
,
Costello
EJ
,
Angold
A
.
Child-, adolescent- and young adult-onset depressions: differential risk factors in development?
Psychol Med
.
2011
;
41
(
11
):
2265
2274
19
Asarnow
JR
,
Zeledon
LR
,
D’Amico
E
, et al
.
Depression and health risk behaviors: towards optimizing primary care service strategies for addressing risk
.
Prim Health Care
.
2014
;
4
(
1
):
152
20
Benton
PM
,
Skouteris
H
,
Hayden
M
.
Does maternal psychopathology increase the risk of pre-schooler obesity? a systematic review
.
Appetite
.
2015
;
87
:
259
282
21
Daw
J
,
Margolis
R
,
Wright
L
.
Emerging adulthood, emergent health lifestyles: sociodemographic determinants of trajectories of smoking, binge drinking, obesity, and sedentary behavior
.
J Health Soc Behav
.
2017
;
58
(
2
):
181
197
22
Wideman
L
,
Calkins
SD
,
Janssen
JA
, et al
.
Rationale, design and methods for the RIGHT Track Health Study: pathways from childhood self-regulation to cardiovascular risk in adolescence
.
BMC Public Health
.
2016
;
16
(
1
):
459
23
Schafer
JL
,
Graham
JW
.
Missing data: our view of the state of the art
.
Psychol Methods
.
2002
;
7
(
2
):
147
177
24
Rubin
DB
.
Multiple Imputation for Nonresponse in Surveys
.
Vol. 81
.
Hoboken, NJ
:
John Wiley & Sons
;
2004
25
Hollingshead
AB
.
Four-Factor Index of Social Status
.
New Haven, CT
:
Yale University
;
1975
26
McAloney
K
,
Graham
H
,
Law
C
,
Platt
L
.
A scoping review of statistical approaches to the analysis of multiple health-related behaviours
.
Prev Med
.
2013
;
56
(
6
):
365
371
27
Hu
L
,
Bentler
PM
.
Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives
.
Struct Equ Modeling
.
1999
;
6
(
1
):
1
55
28
Little
T
.
Longitudinal Structural Equation Modeling
.
New York, London
:
The Guilford Press
;
2013
29
Preacher
KJ
,
Selig
JP
.
Advantages of monte carlo confidence intervals for indirect effects
.
Commun Methods Meas
.
2012
;
6
(
2
):
77
98
30
Paul
E
,
Pearson
RM
.
Depressive symptoms measured using the Edinburgh Postnatal Depression Scale in mothers and partners in the ALSPAC study: a data note
.
Wellcome Open Res
.
2020
;
5
:
108
31
Jaffee
SR
,
Sligo
JL
,
McAnally
HM
,
Bolton
AE
,
Baxter
JM
,
Hancox
RJ
.
Early-onset and recurrent depression in parents increases risk of intergenerational transmission to adolescent offspring
.
J Child Psychol Psychiatry
.
2021
;
62
(
8
):
979
988
32
Hammen
C
,
Brennan
PA
.
Severity, chronicity, and timing of maternal depression and risk for adolescent offspring diagnoses in a community sample
.
Arch Gen Psychiatry
.
2003
;
60
(
3
):
253
258
33
Woo Baidal
JA
,
Locks
LM
,
Cheng
ER
,
Blake-Lamb
TL
,
Perkins
ME
,
Taveras
EM
.
Risk factors for childhood obesity in the first 1,000 days: a systematic review
.
Am J Prev Med
.
2016
;
50
(
6
):
761
779
34
Graziano
PA
,
Calkins
SD
,
Keane
SP
.
Toddler self-regulation skills predict risk for pediatric obesity
.
Int J Obes
.
2010
;
34
(
4
):
633
641
35
Götz
FM
,
Gosling
SD
,
Rentfrow
PJ
.
Small effects: the indispensable foundation for a cumulative psychological science
.
Perspect Psychol Sci
.
2021
;
17
(
1
):
205
215
36
Slootmaker
SM
,
Schuit
AJ
,
Chinapaw
MJ
,
Seidell
JC
,
van Mechelen
W
.
Disagreement in physical activity assessed by accelerometer and self-report in subgroups of age, gender, education and weight status
.
Int J Behav Nutr Phys Act
.
2009
;
6
:
17
37
Pascoe
JM
,
Stolfi
A
.
Maternal depression and the pediatrician
.
Pediatrics
.
2004
;
113
(
2
):
424
38
Olson
AL
,
Dietrich
AJ
,
Prazar
G
,
Hurley
J
.
Brief maternal depression screening at well-child visits
.
Pediatrics
.
2006
;
118
(
1
):
207
216
39
Derogatis
LR
.
Manual for the Symptom Checklist 90 Revised (SCL-90R)
.
Baltimore, MD
;
1994
40
Derogatis
LR
,
Unger
R
.
Symptom checklist‐90‐revised
.
The Corsini Encyclopedia of Psychology
.
2010
:
1
2
41
Kennedy
BL
,
Morris
RL
,
Pedley
LL
,
Schwab
JJ
.
The ability of the Symptom Checklist SCL-90 to differentiate various anxiety and depressive disorders
.
Psychiatr Q
.
2001
;
72
(
3
):
277
288
42
Schmitz
N
,
Hartkamp
N
,
Kiuse
J
,
Franke
GH
,
Reister
G
,
Tress
W
.
The Symptom Check-List-90-R (SCL-90-R): a German validation study
.
Qual Life Res
.
2000
;
9
(
2
):
185
193
43
Buysse
DJ
,
Reynolds
CF
III
,
Monk
TH
,
Berman
SR
,
Kupfer
DJ
.
The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research
.
Psychiatry Res
.
1989
;
28
(
2
):
193
213
44
Johnson
F
,
Wardle
J
,
Griffith
J
.
The Adolescent Food Habits Checklist: reliability and validity of a measure of healthy eating behaviour in adolescents
.
Eur J Clin Nutr
.
2002
;
56
(
7
):
644
649
45
Godin
G
.
The Godin-Shephard Leisure-Time Physical Activity Questionnaire
.
Health Fit J Can
.
2011
;
4
(
1
):
18
22
46
Jacobs
DR
Jr
,
Ainsworth
BE
,
Hartman
TJ
,
Leon
AS
.
A simultaneous evaluation of 10 commonly used physical activity questionnaires
.
Med Sci Sports Exerc
.
1993
;
25
(
1
):
81
91
47
Rosenberg
DE
,
Norman
GJ
,
Wagner
N
,
Patrick
K
,
Calfas
KJ
,
Sallis
JF
.
Reliability and validity of the Sedentary Behavior Questionnaire (SBQ) for adults
.
J Phys Act Health
.
2010
;
7
(
6
):
697
705
48
Brener
ND
,
Kann
L
,
McManus
T
,
Kinchen
SA
,
Sundberg
EC
,
Ross
JG
.
Reliability of the 1999 youth risk behavior survey questionnaire
.
J Adolesc Health
.
2002
;
31
(
4
):
336
342
49
Kovacs
M
.
Children’s Depression Inventory
.
North Tonawanda, NY
:
Multi-Health Systems, Inc.
;
1992

Supplementary data