Association of Positive Family Relationships With Mental Health Trajectories From Adolescence to Midlife | Adolescent Medicine | JAMA Pediatrics | JAMA Network
[Skip to Navigation]
Sign In
Figure 1.  Center for Epidemiologic Studies–Depression Scale (CES-D) Growth Curve by Levels of Family Cohesion Across Ages 12 to 40 Years
Center for Epidemiologic Studies–Depression Scale (CES-D) Growth Curve by Levels of Family Cohesion Across Ages 12 to 40 Years

The CES-D scores were calculated from propensity score weighting conditional growth curve models.

aThe CES-D scores by 2 levels of family cohesion were not significantly different at α = .05.

Figure 2.  Center for Epidemiologic Studies–Depression Scale (CES-D) Growth Curve by Levels of Parent-Child Conflict Across Ages 12 to 40 Years
Center for Epidemiologic Studies–Depression Scale (CES-D) Growth Curve by Levels of Parent-Child Conflict Across Ages 12 to 40 Years

The CES-D scores were calculated from propensity score weighting conditional growth curve models.

aThe CES-D scores by 2 levels of family conflict were not significantly different at α = .05.

Table 1.  Sample Demographics and Study Characteristicsa
Sample Demographics and Study Characteristicsa
Table 2.  Age-Specific CES-D Scores From Growth Curve Models for Overall Sample and by Levels of Family Cohesion in Add Health (1995-2017)
Age-Specific CES-D Scores From Growth Curve Models for Overall Sample and by Levels of Family Cohesion in Add Health (1995-2017)
Table 3.  Age-Specific CES-D Growth Curve Scores by Levels of Parent-Child Conflict in Add Health (1995-2017)
Age-Specific CES-D Growth Curve Scores by Levels of Parent-Child Conflict in Add Health (1995-2017)
1.
World Health Organization. Depression. 2017. https://www.who.int/mediacentre/factsheets/fs369/en/. Accessed December 15, 2018.
2.
Whiteford  HA, Degenhardt  L, Rehm  J,  et al.  Global burden of disease attributable to mental and substance use disorders: findings from the Global Burden of Disease Study 2010.  Lancet. 2013;382(9904):1575-1586. doi:10.1016/S0140-6736(13)61611-6PubMedGoogle ScholarCrossref
3.
Kessler  RC, Avenevoli  S, Ries Merikangas  K.  Mood disorders in children and adolescents: an epidemiologic perspective.  Biol Psychiatry. 2001;49(12):1002-1014. doi:10.1016/S0006-3223(01)01129-5PubMedGoogle ScholarCrossref
4.
Lewinsohn  PM, Rohde  P, Seeley  JR, Klein  DN, Gotlib  IH.  Psychosocial functioning of young adults who have experienced and recovered from major depressive disorder during adolescence.  J Abnorm Psychol. 2003;112(3):353-363. doi:10.1037/0021-843X.112.3.353PubMedGoogle ScholarCrossref
5.
Kessler  RC, Angermeyer  M, Anthony  JC,  et al.  Lifetime prevalence and age-of-onset distributions of mental disorders in the World Health Organization’s World Mental Health Survey Initiative.  World Psychiatry. 2007;6(3):168-176.PubMedGoogle Scholar
6.
Case  A, Deaton  A.  Rising morbidity and mortality in midlife among white non-Hispanic Americans in the 21st century.  Proc Natl Acad Sci U S A. 2015;112(49):15078-15083. doi:10.1073/pnas.1518393112PubMedGoogle ScholarCrossref
7.
Meara  E, Skinner  J.  Losing ground at midlife in America.  Proc Natl Acad Sci U S A. 2015;112(49):15006-15007. doi:10.1073/pnas.1519763112PubMedGoogle ScholarCrossref
8.
Kochanek  KD, Arias  E, Bastian  BA.  The effect of changes in selected age-specific causes of death on non-Hispanic white life expectancy between 2000 and 2014.  NCHS Data Brief. 2016;(250):1-8.PubMedGoogle Scholar
9.
Stein  EM, Gennuso  KP, Ugboaja  DC, Remington  PL.  The epidemic of despair among white Americans: trends in the leading causes of premature death, 1999-2015.  Am J Public Health. 2017;107(10):1541-1547. doi:10.2105/AJPH.2017.303941PubMedGoogle ScholarCrossref
10.
Masters  RK, Tilstra  AM, Simon  DH.  Explaining recent mortality trends among younger and middle-aged white Americans.  Int J Epidemiol. 2018;47(1):81-88. doi:10.1093/ije/dyx127PubMedGoogle ScholarCrossref
11.
Andrews  G, Issakidis  C, Sanderson  K, Corry  J, Lapsley  H.  Utilising survey data to inform public policy: comparison of the cost-effectiveness of treatment of ten mental disorders.  Br J Psychiatry. 2004;184:526-533. doi:10.1192/bjp.184.6.526PubMedGoogle ScholarCrossref
12.
Patel  V, Flisher  AJ, Hetrick  S, McGorry  P.  Mental health of young people: a global public-health challenge.  Lancet. 2007;369(9569):1302-1313. doi:10.1016/S0140-6736(07)60368-7PubMedGoogle ScholarCrossref
13.
Hogan  P, Hogan  BK. The family treatment of depression. In: Flach  FF, Draghi  SC, eds.  The Nature and Treatment of Depression. New York, NY: Wiley; 1978:197-228.
14.
Young  JF, Mufson  L. Interpersonal psychotherapy for treatment and prevention of adolescent depression. In: Abela  JR, Hankin  BL, eds.  Handbook of Depression in Children and Adolescents. New York, NY: Guilford Press; 2008:288-306.
15.
Allen  JP, Porter  M, McFarland  C, McElhaney  KB, Marsh  P.  The relation of attachment security to adolescents’ paternal and peer relationships, depression, and externalizing behavior.  Child Dev. 2007;78(4):1222-1239. doi:10.1111/j.1467-8624.2007.01062.xPubMedGoogle ScholarCrossref
16.
Eberhart  NK, Shih  JH, Hammen  CL, Brennan  PA.  Understanding the sex difference in vulnerability to adolescent depression: an examination of child and parent characteristics.  J Abnorm Child Psychol. 2006;34(4):495-508. doi:10.1007/s10802-006-9020-4PubMedGoogle ScholarCrossref
17.
Meadows  SO, Brown  JS, Elder  GH.  Depressive symptoms, stress, and support: gendered trajectories from adolescence to young adulthood.  J Youth Adolesc. 2006;35(1):89-99. doi:10.1007/s10964-005-9021-6Google ScholarCrossref
18.
Field  T, Diego  M, Sanders  C.  Adolescent depression and risk factors.  Adolescence. 2001;36(143):491-498.PubMedGoogle Scholar
19.
MacPhee  AR, Andrews  JJW.  Risk factors for depression in early adolescence.  Adolescence. 2006;41(163):435-466.PubMedGoogle Scholar
20.
Makri-Botsari  E.  Risk/protective effects on adolescent depression: role of individual, family and peer factors.  Psychol Stud (Mysore). 2005;50(1):50-61.Google Scholar
21.
Kessler  RC.  Epidemiology of women and depression.  J Affect Disord. 2003;74(1):5-13. doi:10.1016/S0165-0327(02)00426-3PubMedGoogle ScholarCrossref
22.
Hyde  JS, Mezulis  AH, Abramson  LY.  The ABCs of depression: integrating affective, biological, and cognitive models to explain the emergence of the gender difference in depression.  Psychol Rev. 2008;115(2):291-313. doi:10.1037/0033-295X.115.2.291PubMedGoogle ScholarCrossref
23.
Landman-Peeters  KMC, Hartman  CA, van der Pompe  G, den Boer  JA, Minderaa  RB, Ormel  J.  Gender differences in the relation between social support, problems in parent-offspring communication, and depression and anxiety.  Soc Sci Med. 2005;60(11):2549-2559. doi:10.1016/j.socscimed.2004.10.024PubMedGoogle ScholarCrossref
24.
Berkman  LF, Glass  T, Brissette  I, Seeman  TE.  From social integration to health: Durkheim in the new millennium.  Soc Sci Med. 2000;51(6):843-857. doi:10.1016/S0277-9536(00)00065-4PubMedGoogle ScholarCrossref
25.
Thoits  PA.  Mechanisms linking social ties and support to physical and mental health.  J Health Soc Behav. 2011;52(2):145-161. doi:10.1177/0022146510395592PubMedGoogle ScholarCrossref
26.
Umberson  D, Montez  JK.  Social relationships and health: a flashpoint for health policy.  J Health Soc Behav. 2010;51(suppl):S54-S66. doi:10.1177/0022146510383501PubMedGoogle ScholarCrossref
27.
Gariépy  G, Honkaniemi  H, Quesnel-Vallée  A.  Social support and protection from depression: systematic review of current findings in Western countries.  Br J Psychiatry. 2016;209(4):284-293. doi:10.1192/bjp.bp.115.169094PubMedGoogle ScholarCrossref
28.
Mirowsky  J, Ross  CE.  Social Causes of Psychological Distress. 2nd ed. New York, NY: Aldine de Gruyter; 2003.
29.
Sutin  AR, Terracciano  A, Milaneschi  Y, An  Y, Ferrucci  L, Zonderman  AB.  The trajectory of depressive symptoms across the adult life span.  JAMA Psychiatry. 2013;70(8):803-811. doi:10.1001/jamapsychiatry.2013.193PubMedGoogle ScholarCrossref
30.
Cyranowski  JM, Frank  E, Young  E, Shear  MK.  Adolescent onset of the gender difference in lifetime rates of major depression: a theoretical model.  Arch Gen Psychiatry. 2000;57(1):21-27. doi:10.1001/archpsyc.57.1.21PubMedGoogle ScholarCrossref
31.
Harris  KM, Halpern  CT, Whitsel  EA,  et al. Cohort profile: the National Longitudinal Study of Adolescent to Adult Health (Add Health) [published online June 29, 2019]. Int J Epidemiol. doi:10.1093/ije/dyz115
32.
Harris  KM, Halpern  CT, Hussey  J,  et al.  Social, behavioral, and genetic linkages from adolescence into adulthood.  Am J Public Health. 2013;103(suppl 1):S25-S32. doi:10.2105/AJPH.2012.301181PubMedGoogle ScholarCrossref
33.
Radloff  LS.  The CES-D scale: a self-report depression scale for research in the general population.  Appl Psychol Meas. 1977;1(3):385-401. doi:10.1177/014662167700100306Google ScholarCrossref
34.
Perreira  KM, Deeb-Sossa  N, Harris  KM, Bollen  K.  What are we measuring? an evaluation of the CES-D across race/ethnicity and immigrant generation.  Soc Forces. 2005;83(4):1567-1601. doi:10.1353/sof.2005.0077Google ScholarCrossref
35.
Radloff  LS.  The use of the Center for Epidemiologic Studies Depression Scale in adolescents and young adults.  J Youth Adolesc. 1991;20(2):149-166. doi:10.1007/BF01537606PubMedGoogle ScholarCrossref
36.
Jensen  TM, Harris  KM.  Stepfamily relationship quality and stepchildren’s depression in adolescence and adulthood.  Emerg Adulthood. 2017;5(3):191-203. doi:10.1177/2167696816669901PubMedGoogle ScholarCrossref
37.
Harris  KM, Gerken  K, Chen  P, Yang  YC. Social integration and health from adolescence into adulthood. Presented at the: DeJong Lecture in Social Demography; November 12, 2016; Pennsylvania State University, University Park, Pennsylvania.
38.
Chow  GC.  Tests of equality between sets of coefficients in two linear regressions.  Econometrica. 1960;28(3):591. doi:10.2307/1910133Google ScholarCrossref
39.
Gould  W. Can You Explain Chow Tests? https://www.stata.com/support/faqs/statistics/chow-tests/. Accessed March 23, 2019.
40.
Imbens  GW.  Nonparametric estimation of average treatment effects under exogeneity: a review.  Rev Econ Stat. 2004;86(1):4-29. doi:10.1162/003465304323023651Google ScholarCrossref
41.
Robins  JM.  Association, causation, and marginal structural models.  Synthese. 1999;121(1-2):151-179. doi:10.1023/A:1005285815569Google ScholarCrossref
42.
Sato  T, Matsuyama  Y.  Marginal structural models as a tool for standardization.  Epidemiology. 2003;14(6):680-686. doi:10.1097/01.EDE.0000081989.82616.7dPubMedGoogle ScholarCrossref
43.
Harris  KM, Chen  P. The acculturation of parent-child relations in immigrant families. Paper presented at: Annual Meeting of the Population Association of America; April 2, 2004; Boston, Massachusetts.
44.
Chen  P, Chantala  K.  Guidelines for Analyzing Add Health Data. Chapel Hill: Carolina Population Center, University of North Carolina at Chapel Hill; 2014.
45.
Singer  JD, John  B.  Willett. Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence. London, England: Oxford University Press; 2003. doi:10.1093/acprof:oso/9780195152968.001.0001
46.
Raudenbush  SW, Bryk  AS.  Hierarchical Linear Models: Applications and Data Analysis Methods. 2nd ed. Thousand Oaks, California: Sage Publications, Inc; 2002:512.
47.
Rabe-Hesketh  S, Skrondal  A.  Multilevel and Longitudinal Modeling Using Stata. 3rd ed. College Station, TX: Stata Press Publication; 2012.
48.
Stice  E, Ragan  J, Randall  P.  Prospective relations between social support and depression: differential direction of effects for parent and peer support?  J Abnorm Psychol. 2004;113(1):155-159. doi:10.1037/0021-843X.113.1.155PubMedGoogle ScholarCrossref
49.
Armsden  GC, Greenberg  MT.  The inventory of parent and peer attachment: individual differences and their relationship to psychological well-being in adolescence.  J Youth Adolesc. 1987;16(5):427-454. doi:10.1007/BF02202939PubMedGoogle ScholarCrossref
50.
McNeely  C, Blanchard  J.  The Teen Years Explained: A Guide to Healthy Adolescent Development. Baltimore, MD: Johns Hopkins Bloomberg School of Public Health: Center for Adolescent Health; 2009.
51.
Bonnie  RJ, Stroud  C, Breiner  H; Committee on Improving the Health, Safety, and Well-Being of Young Adults, Board on Children, Youth, and Families, Institute of Medicine.  National Research Council. Investing in the Health and Well-Being of Young Adults. Washington, DC: National Academies Press; 2015.
52.
Takeuchi  Y, Matsushita  H, Sakai  H, Kawano  H, Ochi  M.  Developmental changes in cerebrospinal fluid concentrations of monoamine-related substances in patients with dentatorubral-pallidoluysian atrophy.  J Child Neurol. 2001;16(2):79-82. doi:10.1177/088307380101600202PubMedGoogle ScholarCrossref
53.
Sawyer  SM, Afifi  RA, Bearinger  LH,  et al.  Adolescence: a foundation for future health.  Lancet. 2012;379(9826):1630-1640. doi:10.1016/S0140-6736(12)60072-5PubMedGoogle ScholarCrossref
54.
Spear  LP.  The adolescent brain and age-related behavioral manifestations.  Neurosci Biobehav Rev. 2000;24(4):417-463. doi:10.1016/S0149-7634(00)00014-2PubMedGoogle ScholarCrossref
55.
Case  R.  Intellectual Development: Birth to Adulthood. Orlando, FL: Academic Press; 1985.
56.
Chalmers  D, Lawrence  JA.  Investigating the effects of planning aids on adults’ and adolescents’ organisation of a complex task.  Int J Behav Dev. 1993;16(2):191-214. doi:10.1177/016502549301600206Google ScholarCrossref
57.
Angold  A, Costello  EJ.  Depressive comorbidity in children and adolescents: empirical, theoretical, and methodological issues.  Am J Psychiatry. 1993;150(12):1779-1791. doi:10.1176/ajp.150.12.1779PubMedGoogle ScholarCrossref
58.
Abela  JR, Hankin  BL. Depression in children and adolescents: causes, treatment, and prevention. In: Abela  JR, Hankin  BL, eds.  Handbook of Depression in Children and Adolescents. New York, NY: The Guilford Press; 2008:3-5.
59.
Yap  MBH, Pilkington  PD, Ryan  SM, Jorm  AF.  Parental factors associated with depression and anxiety in young people: a systematic review and meta-analysis.  J Affect Disord. 2014;156:8-23. doi:10.1016/j.jad.2013.11.007PubMedGoogle ScholarCrossref
60.
Crook  T, Raskin  A, Eliot  J.  Parent-child relationships and adult depression.  Child Dev. 1981;52(3):950-957. doi:10.2307/1129099PubMedGoogle ScholarCrossref
61.
Enns  MW, Cox  BJ, Clara  I.  Parental bonding and adult psychopathology: results from the US National Comorbidity Survey.  Psychol Med. 2002;32(6):997-1008. doi:10.1017/S0033291702005937PubMedGoogle ScholarCrossref
62.
Russek  LG, Schwartz  GE.  Perceptions of parental caring predict health status in midlife: a 35-year follow-up of the Harvard Mastery of Stress Study.  Psychosom Med. 1997;59(2):144-149. doi:10.1097/00006842-199703000-00005PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Original Investigation
    October 7, 2019

    Association of Positive Family Relationships With Mental Health Trajectories From Adolescence to Midlife

    Author Affiliations
    • 1Carolina Population Center, University of North Carolina at Chapel Hill, Chapel Hill
    • 2Department of Sociology, University of North Carolina at Chapel Hill, Chapel Hill
    JAMA Pediatr. 2019;173(12):e193336. doi:10.1001/jamapediatrics.2019.3336
    Key Points

    Question  How are adolescent family relationships associated with trajectories of depressive symptoms from adolescence into midlife for women and men?

    Findings  In this cohort study of 18 185 individuals (9233 females and 8952 males), those who experienced positive adolescent family relationships had significantly lower levels of depressive symptoms from early adolescence to midlife (late 30s to early 40s) than did those who experienced less-positive family relationships.

    Meaning  The findings suggest an association of early intervention in family relationships during adolescence with better mental health into adulthood and midlife.

    Abstract

    Importance  National longitudinal studies that examine the linkages between early family experiences and sex-specific development of depression across the life course are lacking despite the urgent need for interventions in family settings to prevent adult depression.

    Objective  To examine whether positive adolescent family relationships are associated with reduced depressive symptoms among women and men as they enter midlife.

    Design, Setting, and Participants  This study analyzed data from the National Longitudinal Study of Adolescent to Adult Health, which used a multistage, stratified school-based design to select a prospective cohort of 20 745 adolescents in grades 7 to 12 from January 3, 1994, to December 26, 1995 (wave 1). Respondents were followed up during 4 additional waves from April 14 to September 9, 1996 (wave 2); April 2, 2001, to May 9, 2002 (wave 3); April 3, 2007, to February 1, 2009 (wave 4); and March 3, 2016, to May 8, 2017 (sample 1, wave 5), when the cohort was aged 32 to 42 years. The study sample of 8952 male adolescents and 9233 female adolescents that were analyzed was a US national representation of all population subgroups by sex, race/ethnicity, socioeconomic status, and geography.

    Exposures  Adolescent family cohesion and low parent-child conflict.

    Main Outcomes and Measures  Levels of depressive symptoms (Center for Epidemiologic Studies–Depression Scale [CES-D]) from ages 12 to 42 years were used to estimate propensity score–weighted growth curve models to assess sex differences in trajectories of depression by levels of positive adolescent family relationships.

    Results  A total of 18 185 individuals (mean [SD] age at wave 1, 15.42 [0.12] years; 9233 [50.8%] female) participated in the study. Females and males who experienced positive adolescent family relationships had significantly lower levels of depressive symptoms from early adolescence to midlife than did those who experienced less positive adolescent family relationships. For example, depressive symptoms were lower among those with high levels of family cohesion compared with those with low cohesion between 12 (1.26 lower CES-D score; 95% CI, 1.10-1.42) and 40 (0.78 lower CES-D score; 95% CI, 0.50-1.06) years of age among females and between 12 (0.72 lower CES-D score; 95% CI, 0.57-0.86) and 37 (0.21 lower CES-D score; 95% CI, 0.00-0.41) years of age among males. The reduction in depressive symptoms associated with positive adolescent family relationships was greater for females than males during the adolescent and early adulthood years (ie, early 20s) (eg, low-high cohesion difference in mean CES-D score, −1.26 [95% CI, −1.42 to −1.10] for females and −0.72 [95% CI, −0.86 to −0.57] for males at 12 years of age; low-high cohesion difference in mean CES-D score, −0.61 [95% CI, −0.69 to −0.53] for females and −0.40 [95% CI, −0.48 to −0.31] for males at 20 years of age), after which females and males benefited equally from positive adolescent relationships throughout young adulthood to midlife.

    Conclusions and Relevance  The findings suggest that positive adolescent family relationships are associated with better mental health among females and males from early adolescence to midlife. Interventions in early family life to foster healthy mental development throughout the life course appear to be important.

    Introduction

    Depression is a prevalent mental condition worldwide and a significant contributor to the global burden of disease.1,2 Depression often initially occurs during adolescence,3 may continue or recur in adulthood,4 and tends to become a lifetime chronic mental disorder.5 Poor mental health and depressive symptoms may be associated with the recent increase in midlife premature deaths due to suicide, alcohol, and drugs.6-10

    Although treatment methods and intervention efforts continue to advance, a large proportion of depressive conditions remain irreversible.11 The push for prevention and early, affordable, and feasible interventions has been stronger than ever, especially for young people.12 The family context, in particular, has received considerable attention from mental health care professionals and researchers for early intervention efforts.13,14 Prior research has identified risk (eg, neglect, physical and sexual abuse, financial insecurity, residential mobility, and sexual harassment) and protective (eg, family attachment, parental support, parent-child communication, and financial stability) factors for youth depression in the family setting.15-23 Most research focuses on risk factors, but prevention efforts may be more effective by focusing on protective factors. For example, warm and cohesive family relationships provide social support and resources that help buffer youths from stresses of adolescent life.15,17,20,24-26 In addition, close parent-child relationships facilitate communication about personal problems and coping strategies.15,23

    Most of this research, however, comes from small cross-sectional studies15-20,23 with clinical or community samples and suggests only a short-term role of positive family factors in lessening depressive symptoms during childhood and adolescence. Whether positive family relationships promote better mental health beyond adolescence and through the early years of adulthood is unknown. There is a dearth of national longitudinal studies that track individuals over time to understand the interconnections between early family life and the development of life-course depression.27 We implemented a longitudinal, developmental, and life-course approach using nationally representative data and estimating propensity score–weighted growth curve models to examine the long-term association of adolescent family relationships with the trajectories of depressive symptoms from early adolescence to midlife.

    The literature on sex differences in depression is well established,28,29 especially during adolescence, when rates of depression in females first increase compared with those among males.16,21,22,30 Research also suggests that females benefit more from social support in lowering their risks of depression and anxiety compared with males.17,23 We therefore examined the differences among males and females in the trajectories of depression from early adolescence to midlife and the differential benefits of positive family relationships associated with depressive symptoms over time.

    Methods

    This cohort study used data from the National Longitudinal Study of Adolescent to Adult Health (Add Health), a nationally representative cohort study of 20 745 adolescents in grades 7 to 12 January 3, 1994, to December 26, 1995 (wave 1; ages 12-21 years), in the United States.31,32 Respondents were followed up during 4 additional waves from April 14 to September 9, 1996 (wave 2); April 2, 2001, to May 9, 2002 (wave 3); April 3, 2007, to February 1, 2009 (wave 4); and March 3, 2016, to May 8, 2017 (sample 1, wave 5), when the cohort was aged 32 to 42 years. Add Health used a multistage, stratified, school-based, cluster sampling design to select a probability sample of more than 20 000 adolescents from school rosters. Written informed consent was obtained from all participants. Survey procedures were approved by the institutional review board at the University of North Carolina. All data were deidentified.

    Add Health was designed to study the association of social contexts of adolescent life with the health and behavior of adolescents and their outcomes in adulthood (more design details are given elsewhere31,32). The innovative design provided us with ideal data to achieve our study goals. The analytical sample included 18 185 respondents (8952 males and 9233 females) who had sampling weights and data on family relationships and 1 to 5 measures of depressive symptoms over time. The composition of the analytical sample was similar to that of those excluded because of missing data (eTable 2 in the Supplement).

    Measures

    Depressive symptoms were measured using a modified version of the Center for Epidemiologic Studies–Depression Scale (CES-D).33,34 The scale included the frequency of experiencing 3 depressive symptoms: (1) could not shake off the blues, even with help from family and friends; (2) felt depressed during the past 7 days; and (3) felt sad (responses include 0, never or rarely; 1, sometimes; 2, a lot of the time; and 3, most or all of the time) during the past 7 days. This composite measure is reliable and valid for assessing depressive symptoms and is invariant across age, race/ethnicity, and immigrant status.34,35 We summed the responses on the 3 CES-D items (range, 0-9, with 0 indicating never having any depressive symptoms and 9 indicating having symptoms most or all of the time) at each wave, with a higher score representing greater depressive symptoms. We explored a different version of the depressive symptoms measure by standardizing the sum of all CES-D items available during each wave, which yielded similar statistical results (eFigures 1-4 in the Supplement). We used the 3-item CES-D for the measure of depressive symptoms because it is invariant across the diverse racial/ethnic Add Health sample and is more easily interpretable than the standardization approach.

    We measured family relationships in 2 domains: (1) family cohesion and (2) absence of parent-child conflict (described below). For each domain, we created a binary indicator of positive family relationships (coded as 1) based on face validity by categorizing positive at approximately the top quartile of the sample distribution on each measure.

    Family cohesion was measured by mean responses (1, not at all; 2, very little; 3, somewhat; 4, quite a bit; and 5, very much) of adolescent reports on feelings about how much people in their family understood them, they and their family had fun together, and their family paid attention to them (mean [SD] index, 3.7 [0.84]; α = .79).36 The binary version was coded 1 for scores higher than 4 on this composite measure.

    No parent-child conflict was measured by the mean response (range, 0-1, with 0 indicating highest level of conflict and 1 indicating no conflict at all between the parent and child) of adolescent reports on whether they had a serious argument about their behavior with their mother and/or their father in the past 4 weeks (separate question for each parent; mean [SD], 0.31 [0.42]). Those who scored a mean of 0 were coded as 1, indicating low conflict between the parent(s) and child.

    Statistical Analysis

    We estimated sex-specific growth curve models of depressive symptoms (eTables 4-7 in the Supplement). Centered age (in years by subtracting 12) was modeled in linear, quadratic, and cubic forms so that the start (time = 0) of the growth curve represented 12 years of age. Model 1 estimated an unconditional growth curve that included only linear, quadratic, and cubic age variables. Model 2 added exposure variables (family cohesion, low parent-child conflict) and their interaction with age variables to examine whether the initial level and slope of depressive symptoms varied by the 2 levels of family relationships. Model 3 added a number of key risk and protective factors associated with trajectories of depression, including (1) sociodemographic factors (race, ethnicity, and parental educational level), (2) family context (family structure and physical and sexual abuse before 12 years of age), (3) sleep problems as a measure of adolescent health, and (4) life course nonfamily social support (school engagement, friendships, religious involvement, and romantic relationships). Model 3 estimated whether differences in CES-D trajectories by exposure variables remained significant when risk and protective factors were held constant.37 Race/ethnicity was self-reported during the survey. Chow tests were conducted to examine whether a fully interactive model with sex (ie, all variables interacted with sex) resulted in a significantly improved fit to the data than a model without the full set of interactions, requiring stratified models by sex.38,39 The Chow test also enabled us to test whether the exposure variables differed significantly by sex38,39 (eMethods in the Supplement).

    Growth curve models incorporated propensity score weighting,40-42 which generated a pseudopopulation in which exposure to positive family relationships was randomized. Adolescents who experience positive family relationships may be different from adolescents who do not on various dimensions that affect their likelihood to have good relationships within their families and experience less depression; thus, propensity score weighting was used to adjust for this potential selection bias and included observed confounders (eg, self-esteem, moodiness, family structure, physical abuse before 12 years of age, parental educational level, and parents’ feelings of happiness)40-43 (eMethods in the Supplement).

    All statistical analyses applied survey weights to account for the unequal probabilities of sample selection to produce population estimates.31,44 Descriptive analyses and inverse probability treatment models also adjusted variance estimates for school clustering and stratification by region.44 Our mixed and multilevel growth curve modeling approach corrected SEs for correlation within respondents on CES-D.45-47 We analyzed interactions of each exposure with all age variables but only retained those that were significant at α = .05. All analyses were performed in Stata, version 14.2 (StataCorp). A 2-tailed P < .05 was considered to be statistically significant.

    Results

    A total of 18 185 individuals (mean [SD] age at wave 1, 15.42 [0.12] years; 9233 [50.8%] female) participated in the study. Table 1 presents the weighted sample statistics for all analytic variables (eTable 1 in the Supplement shows unweighted distributions). Although 8772 adolescents (49.4%) experienced high levels of family cohesion and 11 164 (62.1%) experienced no parent-child conflict, males (4448 [50.5%] for family cohesion and 5770 [65.5%] for no conflict) had slightly more positive relationships than females (4324 [48.0%] for family cohesion and 5394 [58.6%] for no conflict).

    Estimated levels of depressive symptoms among females and males in unconditional growth curve models are shown in Table 2. Females experienced high levels of depressive symptoms during the adolescent years (ages 13-18 years), after which levels of depression declined until the early 30s, when they began to increase again to levels equal to those in adolescence. The pattern for males, however, was flatter, with stable levels from adolescence into the 30s, increasing to the highest levels in the late 30s.

    Results from the conditional growth curve models (Table 2 and Table 3) suggest that positive family relationships were associated with lower depressive symptoms across the life course for both females and males. For example, depressive symptoms at 12 years of age were significantly lower among females in highly cohesive families (difference in mean CES-D scores, −1.26; 95% CI, −1.42 to −1.10) (Table 2) and among females without parent-child conflict (difference in mean CES-D scores, −0.80; 95% CI, −0.92 to −0.69) (Table 3) compared with females with less positive family relationships. Depressive symptoms were significantly lower at 12 years of age among males in highly cohesive families (difference in mean CES-D scores, −0.72; 95% CI, −0.86 to 0.57) (Table 2) and males without parent-child conflict (difference in mean CES-D scores, −0.50; 95% CI, −0.61 to −0.39) (Table 3) compared with males with less positive relationships.

    The association between positive family relationships and the trajectory of depressive symptoms was best seen by the growth curves across ages 12 to 40 years by family cohesion (Figure 1) and parent-child conflict (Figure 2) for females and males. The small cell sizes at ages 41 to 42 years preclude showing these estimates in the tables and figures. Females who experienced high levels of family cohesion had significantly fewer depressive symptoms across all observed ages from 12 to 40 years compared with females with low family cohesion (Figure 1). The difference in CES-D scores by levels of family cohesion was greatest in adolescence, beginning at 12 years of age (difference in mean CES-D score, 1.26; 95% CI, 1.10-1.42) (Table 2), decreased during the late 20s (eg, at 30 years of age; difference in mean CES-D score, 0.38; 95% CI, 0.28-0.48), and increased again during the 30s until 40 years of age (difference in mean CES-D score, 0.78; 95% CI, 0.50-1.06), when the difference reached a level similar to the one at the age of 18 years (difference in mean CES-D score, 0.74; 95% CI, 0.65-0.82).

    Among males, high levels of family cohesion were also associated with lower levels of depressive symptoms between 12 and 37 years of age compared with low levels of cohesion (Figure 1). The difference in CES-D scores among males by levels of family cohesion was greatest in adolescence, beginning at 12 years of age (difference in mean CES-D score, 0.72; 95% CI, 0.57-0.86) (Table 2), gradually decreased in early adulthood, and remained stable in the late 20s and 30s.

    Females with little parent-child conflict in adolescence had significantly lower levels of depressive symptoms from 12 to 34 years of age, whereas the significant association of low parent-child conflict with lower levels of depression lasted longer for males, from 12 to 39 years of age (Figure 2). In contrast, adolescents who had low-quality family relationships (low cohesion and high parent-child conflict) were more likely to experience early onset of depressive symptoms, with the highest CES-D scores during the ages of 12 to 15 years, especially among females. Adolescents with poor family relationships continued to experience high levels of depressive symptoms during late adolescence (ages 15-18 years) and the transition to adulthood (ages 19-22 years) and had high risk of depression in their late 30s as they approached midlife.

    Females’ mental health tended to benefit more from positive family relationships than males’ mental health (Table 2 and Table 3 and eTables 4-7 in the Supplement). For example, at 12 years of age (ie, the intercept), the absolute values of the coefficients were significantly higher among females than males (2.13 [95% CI, 1.95-2.31] for females in the family-cohesion model 3 of eTable 4 in the Supplement and 1.21 [95% CI, 1.05-1.38] for males in model 3 of eTable 5 in the Supplement; Wald test for female-male intercept difference, P < .001; 2.03 [95% CI, 1.87-2.20] for females in the conflict model 3 of eTable 6 in the Supplement and 1.16 [95% CI, 1.01-1.32] for males in model 3 of eTable 7 in the Supplement; Wald test, P < .001), suggesting that exposure to positive family relationships had a stronger association with lowering the initial level of depressive symptoms among females than among males.

    Discussion

    Few studies have used a longitudinal approach to examine the association between early adolescent family relationships and long-term development of depression into midlife. Our study was, to our knowledge, the first to examine how family relationships during the sensitive period of adolescence are associated with mental health trajectories through adulthood, providing a new contribution to the research on early family experiences and lifetime depression. We highlight several key findings from our longitudinal, developmental, and life-course study based on results from growth curve models that account for potential selection effects of family relationships. First, levels of depressive symptoms changed during a 30-year life course from early adolescence to midlife among both females and males. Second, females experienced significantly higher levels of depressive symptoms than males between early adolescence and the fourth decade of life through the beginning of their 40s. Third, females’ highest levels of depression occurred during the middle to late adolescent years, whereas males experienced a shorter period of high depressive symptoms during late adolescence but increasing depression levels during the 30s to the start of midlife. Fourth, family cohesion and the absence of parent-child conflict were associated with lower risk of depression, not only in the short term during adolescence but also throughout young adulthood and into midlife. Fifth, affective emotions and positive feelings that derive from being a part of a cohesive family were more important for females than for males and were more strongly associated with improved mental health through the transition to adulthood in females. Similarly, parental support and approval reflected by low parent-child conflict was more strongly associated with improved mental health among females than among males during adolescence and into the early 20s. Once females and males were in their middle 20s, the associations of cohesive and low-conflict adolescent family relationships with improved mental health were equal into their 30s, with the association lasting longer for males with low parent-child conflict.

    Although high-quality family relationships in adolescence buffer the stresses of teenage life,48,49 our findings suggest that positive adolescent family relationships continue to be associated with benefits for mental health throughout the transition to adulthood and into midlife as individuals face various challenges that involve peer networking, continuing education, managing romantic relationships, building careers, climbing the social ladder, and starting new families.50,51 These sources of social and emotional support in early family life likely encourage the development of skills for coping with changing and cumulative stressors,24-26 promoting mental health throughout the life course from early adolescence to midlife and helping to prevent negative outcomes and premature deaths due to suicide, alcohol, or drugs in middle age.

    Our findings emphasize the need for early preventive interventions of depression in adolescent family life. Given the profound transformations in neurologic, biological, cognitive, and social development during adolescence when the onset of depression occurs and its rates increase,52-58 mental health and other health education professionals should target this life stage for interventions in family settings. Public health initiatives that teach and encourage parents and family members to nurture positive family relationships with their adolescents will be most effective in fostering healthy mental development.12-14,59 Our research shows that this approach may not only promote mental health during the sensitive and vulnerable period of adolescence but also may be associated with lower risk of adult mental illness and potentially midlife premature deaths due to alcohol addiction, drug abuse, or suicide. Research is needed to better understand the mechanisms by which family cohesion promotes mental health outcomes for youth and young adults. Although theory suggests key mechanisms involve social support, help in working out problems, coping skills, and knowing there is someone to rely on in times of emotional need, research designs are needed to measure these family processes and test for their effectiveness.15,17,20,25 The skills and coping strategies that youth learn to cope with emotional problems may last across the life course and continue to promote mental health well into adulthood.

    Limitations

    There are some limitations to this study. First, our research focused on family relationships during adolescence and the development of depression into midlife. As such, specific measures of preadolescent family relationships were not examined (or available in our data source), which might also be associated with depression. However, we controlled for physical and sexual abuse before 12 years of age as a proxy to capture an overall picture of family interactions and risks before adolescence. Second, we did not use mother- or father-specific relational measures. Compared with measures from previous research based on retrospective reports with recollection errors,60-62 we believe that we improved the operationalization of adolescent family relationships by using prospective reports and comprehensive measures to capture several dimensions, such as emotional affection, bonding, and parental approval. Third, the measure of depressive symptoms in our study was self-reported and does not represent a clinical diagnosis. Future research should attempt to replicate our findings using a clinical measure in a longitudinal study.

    Conclusions

    In this study, females experienced higher levels of depressive symptoms than males throughout the life course from early adolescence to 40 years of age. During this life stage, females had the highest risks of depression during middle to late adolescence, whereas males had the highest risks of depression later, during their 30s and 40s. Positive adolescent family relationships were associated with better mental health among both females and males from early adolescence to midlife, with the association being stronger for females than males during adolescence and the transition to adulthood. Our findings appear to provide new understanding of the long-term association between early family relationships and lifetime development of depression.

    Back to top
    Article Information

    Accepted for Publication: July 10, 2019.

    Corresponding Author: Ping Chen, PhD, Carolina Population Center, University of North Carolina at Chapel Hill, Campus Box 8120, 123 Franklin St, Chapel Hill, NC 27516 (pc@unc.edu).

    Published Online: October 7, 2019. doi:10.1001/jamapediatrics.2019.3336

    Author Contributions: Drs Harris and Chen had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: All authors.

    Acquisition, analysis, or interpretation of data: All authors.

    Drafting of the manuscript: Chen.

    Critical revision of the manuscript for important intellectual content: All authors.

    Statistical analysis: All authors.

    Obtained funding: Harris.

    Administrative, technical, or material support: Harris.

    Supervision: Harris.

    Conflict of Interest Disclosures: Dr Harris reported receiving grants from the National Institutes of Health during the conduct of the study. No other disclosures were reported.

    Funding/Support: This study was supported by the Carolina Population Center and grant P2C HD050924 from the National Institutes of Health for general support.

    Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the present study; management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    Meeting Presentation: Portions of these data were presented at the Annual Meeting of the Population Association of America; April 26, 2018; Denver, Colorado.

    Additional Information: This research used data from Add Health, a program project directed by Kathleen Mullan Harris and designed by J. Richard Udry, Peter S. Bearman, and Kathleen Mullan Harris at the University of North Carolina at Chapel Hill, and funded by grant P01-HD31921 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, with cooperative funding from 23 other federal agencies and foundations. Information on how to obtain the Add Health data files is available on the Add Health website (http://www.cpc.unc.edu/addhealth).

    References
    1.
    World Health Organization. Depression. 2017. https://www.who.int/mediacentre/factsheets/fs369/en/. Accessed December 15, 2018.
    2.
    Whiteford  HA, Degenhardt  L, Rehm  J,  et al.  Global burden of disease attributable to mental and substance use disorders: findings from the Global Burden of Disease Study 2010.  Lancet. 2013;382(9904):1575-1586. doi:10.1016/S0140-6736(13)61611-6PubMedGoogle ScholarCrossref
    3.
    Kessler  RC, Avenevoli  S, Ries Merikangas  K.  Mood disorders in children and adolescents: an epidemiologic perspective.  Biol Psychiatry. 2001;49(12):1002-1014. doi:10.1016/S0006-3223(01)01129-5PubMedGoogle ScholarCrossref
    4.
    Lewinsohn  PM, Rohde  P, Seeley  JR, Klein  DN, Gotlib  IH.  Psychosocial functioning of young adults who have experienced and recovered from major depressive disorder during adolescence.  J Abnorm Psychol. 2003;112(3):353-363. doi:10.1037/0021-843X.112.3.353PubMedGoogle ScholarCrossref
    5.
    Kessler  RC, Angermeyer  M, Anthony  JC,  et al.  Lifetime prevalence and age-of-onset distributions of mental disorders in the World Health Organization’s World Mental Health Survey Initiative.  World Psychiatry. 2007;6(3):168-176.PubMedGoogle Scholar
    6.
    Case  A, Deaton  A.  Rising morbidity and mortality in midlife among white non-Hispanic Americans in the 21st century.  Proc Natl Acad Sci U S A. 2015;112(49):15078-15083. doi:10.1073/pnas.1518393112PubMedGoogle ScholarCrossref
    7.
    Meara  E, Skinner  J.  Losing ground at midlife in America.  Proc Natl Acad Sci U S A. 2015;112(49):15006-15007. doi:10.1073/pnas.1519763112PubMedGoogle ScholarCrossref
    8.
    Kochanek  KD, Arias  E, Bastian  BA.  The effect of changes in selected age-specific causes of death on non-Hispanic white life expectancy between 2000 and 2014.  NCHS Data Brief. 2016;(250):1-8.PubMedGoogle Scholar
    9.
    Stein  EM, Gennuso  KP, Ugboaja  DC, Remington  PL.  The epidemic of despair among white Americans: trends in the leading causes of premature death, 1999-2015.  Am J Public Health. 2017;107(10):1541-1547. doi:10.2105/AJPH.2017.303941PubMedGoogle ScholarCrossref
    10.
    Masters  RK, Tilstra  AM, Simon  DH.  Explaining recent mortality trends among younger and middle-aged white Americans.  Int J Epidemiol. 2018;47(1):81-88. doi:10.1093/ije/dyx127PubMedGoogle ScholarCrossref
    11.
    Andrews  G, Issakidis  C, Sanderson  K, Corry  J, Lapsley  H.  Utilising survey data to inform public policy: comparison of the cost-effectiveness of treatment of ten mental disorders.  Br J Psychiatry. 2004;184:526-533. doi:10.1192/bjp.184.6.526PubMedGoogle ScholarCrossref
    12.
    Patel  V, Flisher  AJ, Hetrick  S, McGorry  P.  Mental health of young people: a global public-health challenge.  Lancet. 2007;369(9569):1302-1313. doi:10.1016/S0140-6736(07)60368-7PubMedGoogle ScholarCrossref
    13.
    Hogan  P, Hogan  BK. The family treatment of depression. In: Flach  FF, Draghi  SC, eds.  The Nature and Treatment of Depression. New York, NY: Wiley; 1978:197-228.
    14.
    Young  JF, Mufson  L. Interpersonal psychotherapy for treatment and prevention of adolescent depression. In: Abela  JR, Hankin  BL, eds.  Handbook of Depression in Children and Adolescents. New York, NY: Guilford Press; 2008:288-306.
    15.
    Allen  JP, Porter  M, McFarland  C, McElhaney  KB, Marsh  P.  The relation of attachment security to adolescents’ paternal and peer relationships, depression, and externalizing behavior.  Child Dev. 2007;78(4):1222-1239. doi:10.1111/j.1467-8624.2007.01062.xPubMedGoogle ScholarCrossref
    16.
    Eberhart  NK, Shih  JH, Hammen  CL, Brennan  PA.  Understanding the sex difference in vulnerability to adolescent depression: an examination of child and parent characteristics.  J Abnorm Child Psychol. 2006;34(4):495-508. doi:10.1007/s10802-006-9020-4PubMedGoogle ScholarCrossref
    17.
    Meadows  SO, Brown  JS, Elder  GH.  Depressive symptoms, stress, and support: gendered trajectories from adolescence to young adulthood.  J Youth Adolesc. 2006;35(1):89-99. doi:10.1007/s10964-005-9021-6Google ScholarCrossref
    18.
    Field  T, Diego  M, Sanders  C.  Adolescent depression and risk factors.  Adolescence. 2001;36(143):491-498.PubMedGoogle Scholar
    19.
    MacPhee  AR, Andrews  JJW.  Risk factors for depression in early adolescence.  Adolescence. 2006;41(163):435-466.PubMedGoogle Scholar
    20.
    Makri-Botsari  E.  Risk/protective effects on adolescent depression: role of individual, family and peer factors.  Psychol Stud (Mysore). 2005;50(1):50-61.Google Scholar
    21.
    Kessler  RC.  Epidemiology of women and depression.  J Affect Disord. 2003;74(1):5-13. doi:10.1016/S0165-0327(02)00426-3PubMedGoogle ScholarCrossref
    22.
    Hyde  JS, Mezulis  AH, Abramson  LY.  The ABCs of depression: integrating affective, biological, and cognitive models to explain the emergence of the gender difference in depression.  Psychol Rev. 2008;115(2):291-313. doi:10.1037/0033-295X.115.2.291PubMedGoogle ScholarCrossref
    23.
    Landman-Peeters  KMC, Hartman  CA, van der Pompe  G, den Boer  JA, Minderaa  RB, Ormel  J.  Gender differences in the relation between social support, problems in parent-offspring communication, and depression and anxiety.  Soc Sci Med. 2005;60(11):2549-2559. doi:10.1016/j.socscimed.2004.10.024PubMedGoogle ScholarCrossref
    24.
    Berkman  LF, Glass  T, Brissette  I, Seeman  TE.  From social integration to health: Durkheim in the new millennium.  Soc Sci Med. 2000;51(6):843-857. doi:10.1016/S0277-9536(00)00065-4PubMedGoogle ScholarCrossref
    25.
    Thoits  PA.  Mechanisms linking social ties and support to physical and mental health.  J Health Soc Behav. 2011;52(2):145-161. doi:10.1177/0022146510395592PubMedGoogle ScholarCrossref
    26.
    Umberson  D, Montez  JK.  Social relationships and health: a flashpoint for health policy.  J Health Soc Behav. 2010;51(suppl):S54-S66. doi:10.1177/0022146510383501PubMedGoogle ScholarCrossref
    27.
    Gariépy  G, Honkaniemi  H, Quesnel-Vallée  A.  Social support and protection from depression: systematic review of current findings in Western countries.  Br J Psychiatry. 2016;209(4):284-293. doi:10.1192/bjp.bp.115.169094PubMedGoogle ScholarCrossref
    28.
    Mirowsky  J, Ross  CE.  Social Causes of Psychological Distress. 2nd ed. New York, NY: Aldine de Gruyter; 2003.
    29.
    Sutin  AR, Terracciano  A, Milaneschi  Y, An  Y, Ferrucci  L, Zonderman  AB.  The trajectory of depressive symptoms across the adult life span.  JAMA Psychiatry. 2013;70(8):803-811. doi:10.1001/jamapsychiatry.2013.193PubMedGoogle ScholarCrossref
    30.
    Cyranowski  JM, Frank  E, Young  E, Shear  MK.  Adolescent onset of the gender difference in lifetime rates of major depression: a theoretical model.  Arch Gen Psychiatry. 2000;57(1):21-27. doi:10.1001/archpsyc.57.1.21PubMedGoogle ScholarCrossref
    31.
    Harris  KM, Halpern  CT, Whitsel  EA,  et al. Cohort profile: the National Longitudinal Study of Adolescent to Adult Health (Add Health) [published online June 29, 2019]. Int J Epidemiol. doi:10.1093/ije/dyz115
    32.
    Harris  KM, Halpern  CT, Hussey  J,  et al.  Social, behavioral, and genetic linkages from adolescence into adulthood.  Am J Public Health. 2013;103(suppl 1):S25-S32. doi:10.2105/AJPH.2012.301181PubMedGoogle ScholarCrossref
    33.
    Radloff  LS.  The CES-D scale: a self-report depression scale for research in the general population.  Appl Psychol Meas. 1977;1(3):385-401. doi:10.1177/014662167700100306Google ScholarCrossref
    34.
    Perreira  KM, Deeb-Sossa  N, Harris  KM, Bollen  K.  What are we measuring? an evaluation of the CES-D across race/ethnicity and immigrant generation.  Soc Forces. 2005;83(4):1567-1601. doi:10.1353/sof.2005.0077Google ScholarCrossref
    35.
    Radloff  LS.  The use of the Center for Epidemiologic Studies Depression Scale in adolescents and young adults.  J Youth Adolesc. 1991;20(2):149-166. doi:10.1007/BF01537606PubMedGoogle ScholarCrossref
    36.
    Jensen  TM, Harris  KM.  Stepfamily relationship quality and stepchildren’s depression in adolescence and adulthood.  Emerg Adulthood. 2017;5(3):191-203. doi:10.1177/2167696816669901PubMedGoogle ScholarCrossref
    37.
    Harris  KM, Gerken  K, Chen  P, Yang  YC. Social integration and health from adolescence into adulthood. Presented at the: DeJong Lecture in Social Demography; November 12, 2016; Pennsylvania State University, University Park, Pennsylvania.
    38.
    Chow  GC.  Tests of equality between sets of coefficients in two linear regressions.  Econometrica. 1960;28(3):591. doi:10.2307/1910133Google ScholarCrossref
    39.
    Gould  W. Can You Explain Chow Tests? https://www.stata.com/support/faqs/statistics/chow-tests/. Accessed March 23, 2019.
    40.
    Imbens  GW.  Nonparametric estimation of average treatment effects under exogeneity: a review.  Rev Econ Stat. 2004;86(1):4-29. doi:10.1162/003465304323023651Google ScholarCrossref
    41.
    Robins  JM.  Association, causation, and marginal structural models.  Synthese. 1999;121(1-2):151-179. doi:10.1023/A:1005285815569Google ScholarCrossref
    42.
    Sato  T, Matsuyama  Y.  Marginal structural models as a tool for standardization.  Epidemiology. 2003;14(6):680-686. doi:10.1097/01.EDE.0000081989.82616.7dPubMedGoogle ScholarCrossref
    43.
    Harris  KM, Chen  P. The acculturation of parent-child relations in immigrant families. Paper presented at: Annual Meeting of the Population Association of America; April 2, 2004; Boston, Massachusetts.
    44.
    Chen  P, Chantala  K.  Guidelines for Analyzing Add Health Data. Chapel Hill: Carolina Population Center, University of North Carolina at Chapel Hill; 2014.
    45.
    Singer  JD, John  B.  Willett. Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence. London, England: Oxford University Press; 2003. doi:10.1093/acprof:oso/9780195152968.001.0001
    46.
    Raudenbush  SW, Bryk  AS.  Hierarchical Linear Models: Applications and Data Analysis Methods. 2nd ed. Thousand Oaks, California: Sage Publications, Inc; 2002:512.
    47.
    Rabe-Hesketh  S, Skrondal  A.  Multilevel and Longitudinal Modeling Using Stata. 3rd ed. College Station, TX: Stata Press Publication; 2012.
    48.
    Stice  E, Ragan  J, Randall  P.  Prospective relations between social support and depression: differential direction of effects for parent and peer support?  J Abnorm Psychol. 2004;113(1):155-159. doi:10.1037/0021-843X.113.1.155PubMedGoogle ScholarCrossref
    49.
    Armsden  GC, Greenberg  MT.  The inventory of parent and peer attachment: individual differences and their relationship to psychological well-being in adolescence.  J Youth Adolesc. 1987;16(5):427-454. doi:10.1007/BF02202939PubMedGoogle ScholarCrossref
    50.
    McNeely  C, Blanchard  J.  The Teen Years Explained: A Guide to Healthy Adolescent Development. Baltimore, MD: Johns Hopkins Bloomberg School of Public Health: Center for Adolescent Health; 2009.
    51.
    Bonnie  RJ, Stroud  C, Breiner  H; Committee on Improving the Health, Safety, and Well-Being of Young Adults, Board on Children, Youth, and Families, Institute of Medicine.  National Research Council. Investing in the Health and Well-Being of Young Adults. Washington, DC: National Academies Press; 2015.
    52.
    Takeuchi  Y, Matsushita  H, Sakai  H, Kawano  H, Ochi  M.  Developmental changes in cerebrospinal fluid concentrations of monoamine-related substances in patients with dentatorubral-pallidoluysian atrophy.  J Child Neurol. 2001;16(2):79-82. doi:10.1177/088307380101600202PubMedGoogle ScholarCrossref
    53.
    Sawyer  SM, Afifi  RA, Bearinger  LH,  et al.  Adolescence: a foundation for future health.  Lancet. 2012;379(9826):1630-1640. doi:10.1016/S0140-6736(12)60072-5PubMedGoogle ScholarCrossref
    54.
    Spear  LP.  The adolescent brain and age-related behavioral manifestations.  Neurosci Biobehav Rev. 2000;24(4):417-463. doi:10.1016/S0149-7634(00)00014-2PubMedGoogle ScholarCrossref
    55.
    Case  R.  Intellectual Development: Birth to Adulthood. Orlando, FL: Academic Press; 1985.
    56.
    Chalmers  D, Lawrence  JA.  Investigating the effects of planning aids on adults’ and adolescents’ organisation of a complex task.  Int J Behav Dev. 1993;16(2):191-214. doi:10.1177/016502549301600206Google ScholarCrossref
    57.
    Angold  A, Costello  EJ.  Depressive comorbidity in children and adolescents: empirical, theoretical, and methodological issues.  Am J Psychiatry. 1993;150(12):1779-1791. doi:10.1176/ajp.150.12.1779PubMedGoogle ScholarCrossref
    58.
    Abela  JR, Hankin  BL. Depression in children and adolescents: causes, treatment, and prevention. In: Abela  JR, Hankin  BL, eds.  Handbook of Depression in Children and Adolescents. New York, NY: The Guilford Press; 2008:3-5.
    59.
    Yap  MBH, Pilkington  PD, Ryan  SM, Jorm  AF.  Parental factors associated with depression and anxiety in young people: a systematic review and meta-analysis.  J Affect Disord. 2014;156:8-23. doi:10.1016/j.jad.2013.11.007PubMedGoogle ScholarCrossref
    60.
    Crook  T, Raskin  A, Eliot  J.  Parent-child relationships and adult depression.  Child Dev. 1981;52(3):950-957. doi:10.2307/1129099PubMedGoogle ScholarCrossref
    61.
    Enns  MW, Cox  BJ, Clara  I.  Parental bonding and adult psychopathology: results from the US National Comorbidity Survey.  Psychol Med. 2002;32(6):997-1008. doi:10.1017/S0033291702005937PubMedGoogle ScholarCrossref
    62.
    Russek  LG, Schwartz  GE.  Perceptions of parental caring predict health status in midlife: a 35-year follow-up of the Harvard Mastery of Stress Study.  Psychosom Med. 1997;59(2):144-149. doi:10.1097/00006842-199703000-00005PubMedGoogle ScholarCrossref
    ×