short literature review

P E D

B

M

R

C

I

I r p i a i c s l

o n l

F S S

P 5

A ©

arent–Child Connectedness and Behavioral and motional Health Among Adolescents

iann M. Ackard, PhD, Dianne Neumark-Sztainer, PhD, Mary Story, PhD, Cheryl Perry, PhD

ackground: This study sought to examine teen perceptions of mother– child and father– child connectedness, with focus on valuing parental opinions and perception of parental communication and caring, and associations with behavioral and emotional health.

ethods: A population-based sample of 4746 students in public schools completed the 2001 Project EAT (Eating Among Teens) survey.

esults: Overall, the majority of girls and boys reported valuing their parents’ opinion when making serious decisions and believing that their parents cared about them. Yet, one fourth of girls and boys felt unable to talk to their mother about problems, and over half of girls and one third of boys felt unable to talk to their father. Valuing friends’ opinions over parents’ opinions, and perceiving low parental communication and caring were associated with unhealthy weight control, substance use, suicide attempts, body dissatisfaction, depression, and low self-esteem. Of significant concern, compared to their peers who reported feeling that their mother cared quite a bit or very much, youths who reported feeling as though their mother cared very little or not at all about them reported particularly high prevalence rates of unhealthy weight control behaviors (63.49% girls, 25.45% boys); suicide attempts (33.51% girls, 21.28% boys); low self-esteem (47.15% girls, 24.56% boys); and depression (63.52% girls, 33.35% boys).

onclusions: Adolescents’ perceptions of low parental caring, difficulty talking to their parents about problems, and valuing their friends’ opinions for serious decisions were significantly associated with compromised behavioral and emotional health. Interventions aimed at improving the parent– child relationship may provide an avenue toward preventing health risk behaviors in youth. (Am J Prev Med 2006;30(1):59 – 66) © 2006 American Journal of Preventive Medicine

c i d t t C r a

i i t a a f a h y m a h d

ntroduction

n one of the most significant works investigating the relationship between several types of influential environments (e.g., family and school) and health

isk behaviors among adolescents, Resnick et al.1 re- orted that family connectedness was significantly and

nversely associated with emotional distress, suicidality, lcohol use, marijuana use, and early age of sexual ntercourse. Others have found significant direct asso- iations between pathologic family environments and ubstance use,2 depression,3–5 disordered eating,6

ower self-esteem,5 and suicidality.7–10

Although separation from parents is a normal devel- pmental task for adolescents, it does not always culmi- ate in a connected parent– child relationship. A 1-year

ongitudinal project evaluated the quality of the parent–

rom the private practice (Ackard), and Division of Epidemiology, chool of Public Health, University of Minnesota (Neumark-Sztainer, tory, Perry), Minneapolis, Minnesota Address correspondence and reprint requests to: Diann M. Ackard,

a hD, 5101 Olson Memorial Highway, Suite 4001, Golden Valley MN 5422. E-mail: [email protected].

m J Prev Med 2006;30(1) 2006 American Journal of Preventive Medicine • Published by

hild relationship during the adolescent transition to ncreased individuation.5 Parent–teen relationships eemed as connected were associated with fewer symp- oms of depression and anxiety, and greater self-worth han relationships categorized as detached. Similarly, alifornia teens were surveyed across 2 years, and

esults show an association between family connection nd psychological and behavioral health.11

For youth, feeling connected to their families is an mportant anchor, and many do turn to parents for nformation and guidance. In a nationally representa- ive study of the use of healthcare resources among dolescents, mothers were identified by 41.7% of boys nd 58.4% of girls as the first person they would consult or healthcare concerns.12 More broadly, 60.3% of boys nd 71.7% of girls identified parents as one source of ealthcare information. However, of concern is that outh who are at greatest need for adult intervention ay not seek it. For example, results from a study of 879

dolescents indicated that only about half of youth who ad attempted suicide had approached an adult to iscuss their problems.10 Furthermore, those who had

ttempted suicide reported that they were less likely to

590749-3797/06/$–see front matter Elsevier Inc. doi:10.1016/j.amepre.2005.09.013

d f s

b e g m p c n g h o t h v w a w t w c a

M S

P s g s p w

E a s m M C p r

i c 7 ( (

M

T d n E n t w

w i

P

O h c P

P t m y ( a

B

W h w i f u s a c

m s s e u “ c

S d l c f c m

S o d b c

E

B e m g v q

S c R

6

iscuss their problems with a family member in the uture compared to those who had not ever attempted uicide.10

Previous studies have found significant associations etween family connectedness and the behavioral and motional health of youth, but are limited by investi- ating only a few health risk behaviors or by a smaller, ore homogeneous sample. This study expands on

revious research by exploring parent– child communi- ation and caring in a large, ethnically and socioeco- omically diverse population of youth, and by investi- ating a broader range of behavioral and emotional ealth indicators in order to better inform the devel- pment of effective parent and adolescent interven- ions and to identify populations at greatest risk. It was ypothesized that both girls and boys who indicated aluing their friends’ opinions more than their parents’ ould report higher odds for substance use, suicide ttempts, and unhealthy weight-control behaviors, as ell as higher odds of low self-esteem, body dissatisfac-

ion, and depression. Similar directional associations ere also hypothesized between perceived parental aring and ability to talk to parents about problems, nd behavioral and emotional health outcomes.

ethod tudy Population and Design

articipants in the current study included a total of 4746 tudents enrolled in 31 public middle and high schools in the reater Minneapolis/St. Paul metropolitan area of Minne- ota. Schools with diverse racial/ethnic and socioeconomic rofiles were targeted for recruitment to increase diversity ithin the sample. In 2001, participants completed the confidential Project

AT (Eating Among Teens) survey in school classes and were sked to have their height and weight measured in a private creened area by trained staff using standardized anthropo- etric procedures. The study complied with the University of innesota’s Institutional Review Board and Human Subjects’ ommittee, and with each school district’s research board rocess for obtaining student consent. The student response ate was 81.5%.

The sample comprised 2357 girls and 2377 boys (12 ndividuals had missing data for gender and were not in- luded in analyses). Participants were in the following grades: th (28.2% girls, 27.4% boys); 8th (6.4% girls, 6.7% boys); 9th 0.9% girls, 1.0% boys); 10th (50.3% girls, 52.6% boys); 11th 10.1% girls, 8.6% boys); and 12th (4.1% girls, 3.8% boys).

easures

he Project EAT survey includes 221 self-report questions on emographics, family and personal health attitudes, and utritional and weight-related factors. Although the Project AT survey has not been validated against other question- aires or actual behavior, multidisciplinary professional

eams, youth focus groups, and pilot tests of the questions

ere conducted to provide guidance for the selection and s

0 American Journal of Preventive Medicine, Volume 30, Num

ording of questions.13–15 All questions listed below were ncluded in the Project EAT survey.

arent–Child Connectedness

pinions valued. One question in the survey asked, “If you ad a serious decision to make, like whether or not to ontinue in school, whose opinions would you value most?” ossible responses were “parent” or “friend.”

arent– child communication and caring. Two questions in he survey were asked separately for each parent.16 (1) “How

uch do you feel you can talk to your mother (father) about our problems?” (2) “How much do you feel your mother father) cares about you?” Possible responses follow: “not at ll,” “a little,” “somewhat,” “quite a bit,” or “very much.”

ehavioral Health

eight-control behaviors. “Which of the following things ave you done in order to lose weight or keep from gaining eight during the past year?” Participants were requested to

ndicate “yes” or “no” to the following responses: exercise, asted, ate very little food, took diet pills, made myself vomit, sed laxatives, used diuretics, used food substitute (powder/ pecial drink), skipped meals, ate more fruits and vegetables, te less high-fat foods, ate less sweets, and smoked more igarettes.

Behaviors were grouped as follows: healthy (exercise, ate ore fruits and vegetables, ate less high-fat foods, or ate less

weets); less extreme (fasted, ate very little food, used food ubstitute, skipped meals, or smoked more cigarettes); or xtreme (took diet pills, made myself vomit, used laxatives, or sed diuretics). Participants were grouped by use (“yes” or no”) of any less extreme or extreme unhealthy weight ontrol behaviors in the past year.17

ubstance use. “How often have you used the following uring the past year (12 months)? Liquor (beer, wine, hard

iquor), marijuana, or drugs other than marijuana (acid, ocaine, crack, etc.).”16 Possible responses follow: “never,” “a ew times,” “monthly,” “weekly,” or “daily”. Responses were ollapsed into two categories: never versus a few times or ore.

uicide attempts. “Have you ever tried to kill yourself?” The riginal responses included a temporal component (“Yes, uring the past year,” “Yes, more than a year ago,” or “No”),18

ut responses were dichotomized (“yes” or “no”) for the urrent analyses.

motional Health

ody dissatisfaction. The dissatisfaction that one experi- nces with his or her body appearance was assessed using a odified version of Pingitore’s19 scale. Higher scores indicate

reater dissatisfaction. A binary score was created using the alue separating the highest quartile from the lowest three uartiles.

elf-esteem. The self-esteem instrument asked youth to indi- ate their level of agreement with six sentences from the osenberg Self-Esteem Scale.20 Higher scores reflect lower

elf-esteem. A binary variable was created using the value

ber 1 www.ajpm-online.net

s q

D s s n “ i s w

D

P m “ p

R i A

S e f k h f f m c w p S v

S

F s c i m w c a l g w g a t u f p a q e w s v

R D

P 5 b L g s ( m c 1 m g

D

M e d A t m s 2 c a t m 2 b c 8 b t b l

P H

G t t t r c p a r g p p t v

J

eparating the lowest three quartiles from the highest uartile.

epressive mood. Depressive mood was assessed using a cale by Kandel and Davies21 asking the frequency of six ymptoms of depression (dysthymic mood, tension/nervous- ess, fatigue, worry, sleep disturbance, and hopelessness): not at all,” “somewhat,” or “very much.” Higher values ndicate more severe depressive moods. The summed score eparating the lowest three quartiles from the highest quartile as used as a cut-off to create a binary score.

emographics

arent marital status. Each student was asked to report the arital status of his or her parents as “married,” “divorced,”

separated,” “parents never married,” or “one/both of my arents is dead.”

ace/ethnicity. Students could choose as many of the follow- ng as they wished: white, African American, Hispanic, Asian merican, Native American, and mixed/other.

ocioeconomic status. One or both parents’ highest level of ducation was used to establish socioeconomic status (SES) or most youth. Due to the fact that some students did not now their parent’s educational level (n �1058, 22.3%) or ad missing data for items needed to calculate SES, other

actors (family eligibility for public assistance, eligibility for ree or reduced-cost school meals, and employment status of

other and father) were combined in an algorithm using the lassification and regression trees (CART) method,22 which as found to be predictive of parent education among the articipants who completed all questions needed to calculate ES. By using this cartography, the number of missing SES alues was reduced to 4.1% (n �196).

tatistical Analyses

requencies and percentages were used to describe the ample by sociodemographic variables and by parent– child ommunication and caring. Because the sample came from ntact social clusters in schools, clustered logistic regression

odels, in which school was included as a random effect, ere used to investigate the association between parent– child onnectedness and behavioral and emotional health vari- bles, adjusting for sociodemographic characteristics (grade evel, race/ethnicity, SES, and parental marital status). Cate- oric (grade level, race/ethnicity, and parental marital status, ith “white” and “parents are married” serving as the referent roups) and continuous (SES) sociodemographic “covari- tes” were forced to enter in the first step. In the second step, he parent– child connectedness variable was entered to eval- ate the level of improvement of fit in the model. The ollowing response sets served as the comparison: valuing arents’ opinion, feeling able to talk to mother/father quite bit or very much about, and feeling mother/father cares

uite a bit or very much. Adjusted probabilities, standard rrors, and significance values were generated. The p values ere not adjusted for multiple testing. All analyses were run

eparately in 2005 by gender using SAS/STAT software,

ersion 9.1 (SAS Institute Inc., Cary NC, 2004).23 t

anuary 2006

esults escription of Sample

articipants’ race/ethnicity follows: white (45.6% girls, 1.3% boys); African American (20.1% girls, 17.9% oys); Asian American (20.6% girls, 17.8% boys); atina/Latino (5.2% girls, 6.5% boys); and other (8.6% irls, 6.4% boys). They reported their parents’ marital tatus as married (60.7% girls, 62.6% boys); divorced 18.1% girls, 18.5% boys); or other (separated, never arried, or deceased; 21.2% girls, 18.9% boys). SES was

alculated and reported as follows: low (20.4% girls, 4.5% boys); low-middle (19.1% girls, 18.5% boys); iddle (25.6% girls, 27.6% boys); high-middle (21.5%

irls, 25.3% boys); and high (13.4% girls, 14.1% boys).

escription of Parent–Child Connectedness

ost participants indicated that they valued their par- nts’ opinions over their friends’ opinions for serious ecisions (parents’ opinion: 75.5% girls; 82.2% boys). pproximately half reported that they could talk to

heir mother about their problems “quite a bit” or “very uch” (quite a bit/very much: 52.1% girls, 48.6% boys;

omewhat: 22% girls, 23.1% boys; not at all or a little: 5.9% girls, 28.3% boys). Fewer indicated that they ould talk “quite a bit” or “very much” to their father bout their problems, and in fact, the majority reported hat they could not talk to their father (quite a bit/very

uch: 24.6% girls, 38.2% boys; somewhat: 20.0% girls, 5.2% boys; not at all or a little: 55.4% girls, 36.6% oys). A majority of the girls and boys reported feeling ared about by their mothers (quite a bit/very much: 8.6% girls, 90.8% boys; somewhat: 6.3% girls, 4.5% oys; not at all or a little: 5.1% girls, 4.7% boys) and by heir fathers (quite a bit/very much: 78.6% girls, 81.8% oys; somewhat: 8.5% girls, 7.8% boys; not at all or a

ittle: 12.9% girls, 10.4% boys).

arent–Child Connectedness and Behavioral ealth Indicators

irls who valued their friends’ opinions over those of heir parents, and those who felt that they could not alk (or talk very little) to their mother or father about heir problems reported greater prevalence of health isk behaviors than peers who reported higher parental ommunication and caring (Table 1). Girls who re- orted low paternal caring did not report higher prev- lence of substance use compared to their peers who eported high paternal caring. Of significant interest, irls who reported low maternal caring, compared to eers who reported high maternal caring, reported articularly high prevalence of unhealthy weight con- rol (63.49% vs 18.34%) and suicide attempts (33.51% s 10.17%).

Among boys, valuing friends’ opinions over those of

heir parents, and feeling unable (or only slightly able)

Am J Prev Med 2006;30(1) 61

t c c c o a c u S l h u h

P H

A e ( r c t a l t b F

w d

D

T b r a R p p t c b c f a t l t d p s ( t

T l

W

F

F

F

F

a

b

m * * * A

6

o talk to mother or father about problems was signifi- antly and directly associated with unhealthy weight ontrol behaviors, substance use, and suicide attempts ompared to their peers who reported valuing parents’ pinions and feeling able to talk to mother or father bout problems (Table 2). Perceptions of low paternal aring were not significantly associated with substance se compared to perceptions of high paternal caring. imilar to their female counterparts, boys who reported ow maternal caring, compared to peers who reported igh maternal caring, were much more likely to report nhealthy weight control (25.45% vs 3.63%) and to ave attempted suicide (21.28% vs 3.97%).

arent–Child Connectedness and Emotional ealth Indicators

ssociations between parent– child connectedness and motional health indicators are shown in Table 3 girls) and Table 4 (boys). For all youth, parent– child elationships characterized by valuing friends’ opinion ompared to those of parents, and feeling unable to alk to mother or father about problems were strongly ssociated with scores indicating body dissatisfaction, ow self-esteem, and depression. Among girls, percep- ions of minimal parental caring were associated with ody dissatisfaction, low self-esteem, and depression.

able 1. Girls: parent– child connectedness and behavioral h evelsb

Unhealthy weight control

AP (SE) p value

hose opinion is valued for serious decisions? Parents’ opinion 16.41 (1.34) Friends’ opinion 37.95 (3.33) <0.001***

eel you can talk to mother about problems? Quite a bit or very much 15.14 (1.50) Somewhat 22.10 (2.79) 0.018* Not at all or a little 35.48 (3.25) <0.001***

eel you can talk to father about problems? Quite a bit or very much 15.02 (2.16) Somewhat 17.68 (2.59) 0.422 Not at all or a little 25.72 (1.98) 0.001**

eel your mother cares about you? Quite a bit or very much 18.34 (1.33) Somewhat 31.18 (5.58) 0.001** Not at all or a little 63.49 (7.35) <0.001***

eel your father cares about you? Quite a bit or very much 18.78 (1.39) Somewhat 35.73 (5.36) 0.001** Not at all or a little 27.49 (4.06) 0.027*

Adjustments were made for grade level, socioeconomic status, race/ Significance levels are reported with the following comparison: valuin uch about, and feeling mother/father cares quite a bit or very mu

p � 0.05; *p � 0.01; **p � 0.001 (all bolded). P, adjusted probability; SE, standard error.

or boys, minimal parental caring was associated only p

2 American Journal of Preventive Medicine, Volume 30, Num

ith low self-esteem and depression, but not for body issatisfaction.

iscussion

he aim of the present study was to explore associations etween parent– child connectedness and a broad ange of behavioral and emotional health indicators mong a population-based sample of girls and boys. esults from this large study of youth indicate that arent– child relationships characterized by valuing arent opinions for serious decisions, feeling able to alk to parents about problems, and perceiving parental aring were associated with more healthy indicators of ehavioral and emotional health. These results are onsistent with past research, which indicated that amily connectedness was significantly and inversely ssociated with several health risk behaviors and emo- ional health indicators,1 depressive symptoms and ower self-worth,5 substance use,2 and bulimic symp- oms.6 Yet the current study adds to the literature by emonstrating a significant relationship between arent– child connectedness and a broad range of erious behavioral and emotional health risk behaviors substance use, unhealthy weight control, suicide at- empts, body dissatisfaction, low self-esteem, and de-

: adjusted probabilities,a standard errors, and significance

Substance use Suicide attempts

P (SE) p value AP (SE) p value

3.06 (1.42) 10.44 (0.88) 5.39 (2.52) <0.001*** 18.13 (1.87) <0.001***

3.92 (1.68) 9.39 (0.98) 0.32 (2.56) 0.028* 11.49 (1.59) 0.227 6.81 (2.53 <0.001*** 19.31 (1.91) <0.001***

9.73 (2.30) 9.44 (1.40) 5.73 (2.68) 0.078 10.79 (1.64) 0.521 3.65 (1.77) <0.001*** 14.22 (1.19) 0.012*

7.29 (1.37) 10.17 (0.80) 9.95 (4.96) 0.012* 27.06 (4.13) <0.001*** 7.56 (5.58) 0.065 33.51 (4.95) <0.001***

6.87 (1.45) 10.21 (0.85) 9.74 (4.39) 0.004** 21.23 (3.32) <0.001*** 2.41 (3.67) 0.150 20.87 (2.88) <0.001***

city, and parent marital status, and were clustered for school. ents’ opinion, feeling able to talk to mother/father quite a bit or very

ealth

A

3 5

3 4 4

2 3 4

3 4 4

3 4 4

ethni g par

ch.

ression) in a diverse sample of both boys and girls.

ber 1 www.ajpm-online.net

y s i y t i e p t H w t t a o s t t

o a I f s r a o

t a t a r n

e w s i s i m r t t c t v t t e g c l a

T l

W

F

F

F

F

a

b

m * * * A

J

A particularly interesting finding is the valuation that outh place on their parents’ opinions when making erious decisions, notably valuing their parents’ opin- ons more than that of their peers. Although some outh showed more interest in being with their friends han spending time with family, parents may have more nfluence on their teens’ behaviors than may be appar- nt to them. Valuing parents’ opinions appears to be a rotective factor against unhealthy behavioral and emo- ional health indicators among both girls and boys. ealthcare providers and school personnel working ith teens may want to share study results with parents

o help them feel empowered by the knowledge that heir opinions matter. In order to promote teens’ ttention to and respect for parents’ opinions, parents f youth may wish to practice having discussions on ensitive topics without lending judgment to their een’s ideas until asked, and then strategizing solutions o the problem in a collaborative manner.

One particular area of concern was the significance f adolescents’ perceptions of maternal caring, and its ssociation with both behavioral and emotional health. nterventions can enhance mother– child caring. For amilies in which the mother– child relationship is trained, youth may benefit from developing positive elationships with other adult female role models, such s through Big Sister programs, other female relatives,

able 2. Boys: parent– child connectedness and behavioral h evelsb

Unhealthy weight control

AP (SE) p value

hose opinion is valued for serious decisions? Parents’ opinion 3.41 (0.65) Friends’ opinion 9.31 (2.09) <0.001***

eel you can talk to mother about problems? Quite a bit or very much 3.02 (0.68) Somewhat 3.25 (1.00) 0.834 Not at all or a little 8.79 (1.76) <0.001***

eel you can talk to father about problems? Quite a bit or very much 2.48 (0.67) Somewhat 3.82 (1.08) 0.226 Not at all or a little 7.40 (1.43) <0.001***

eel your mother cares about you? Quite a bit or very much 3.63 (0.63) Somewhat 9.74 (3.89) 0.019* Not at all or a little 25.45 (7.13) <0.001***

eel your father cares about you? Quite a bit or very much 3.62 (0.66) Somewhat 5.87 (2.14) 0.198 Not at all or a little 12.38 (3.25) <0.001***

Adjustments were made for grade level, socioeconomic status, race/ Significance levels are reported with the following comparison: valuin uch about, and feeling mother/father cares quite a bit or very mu

p � 0.05; *p � 0.01; **p � 0.001 (all bolded). P, adjusted probability; SE, standard error.

r community leaders, while continuing to ameliorate H

anuary 2006

he parent– child relationship. A longitudinal study of dolescents found that youth who experience a rela- ionship deficit in one source (e.g., the family) may be ble to compensate for that void by forming a positive elationship from another source (e.g., school, eighborhood).23

Increased perceived communication and caring by ither mother or father were consistently associated ith adolescent well-being. These associations under-

core the importance of the parent– child relationship n promoting overall health among youth, and empha- ize that while it is valuable for either parent to be nvolved in enhancing the parent–teen relationship, it

ay be ideal for both mother–teen and father–teen elationships to be fostered through open communica- ion and caring. Professionals working with youth and heir families, such as family therapists and school ounselors, should promote parent– child communica- ion and find opportunities to enhance parental in- olvement in addressing primary and tertiary preven- ion of compromising behaviors among teens. One way o promote parent– child connectedness may be to ncourage parents and adolescents to spend time to- ether, such as at family mealtimes, as higher frequen- ies of family meals also have been associated with ower substance use, depressive symptoms, and suicid- lity, even after controlling for family connectedness.24

: adjusted probabilities,a standard errors, and significance

Substance use Suicide attempts

P (SE) p value AP (SE) p value

9.19 (1.45) 3.77 (0.57) 6.42 (2.90) <0.001*** 10.21 (1.81) <0.001***

5.73 (1.77) 2.80 (0.57) 0.10 (2.60) <0.001*** 4.85 (1.09) 0.052 7.45 (2.41) <0.001*** 8.83 (1.40) <0.001***

4.99 (1.97) 3.74 (0.76) 6.12 (2.46) <0.001*** 2.85 (0.79) 0.403 7.33 (2.14) <0.001*** 7.73 (1.18) 0.001***

1.66 (1.41) 3.97 (0.56) 8.72 (5.83) 0.229 11.29 (3.65) 0.003** 0.75 (6.35) 0.157 21.28 (4.92) <0.001***

1.20 (1.48) 4.00 (0.58) 0.73 (4.41) 0.036* 8.95 (2.40) 0.006** 5.29 (4.11) 0.340 10.11 (2.34) <0.001***

city, and parent marital status, and were clustered for school. ents’ opinion, feeling able to talk to mother/father quite a bit or very

ealth

A

3 5

3 5 4

3 4 4

4 4 5

4 5 4

ethni g par

ch.

ealthcare professionals should also assess the pres-

Am J Prev Med 2006;30(1) 63

e a h m i r I b a

P s s o a s b c e e y p p p c p s s

c i b h w b f t n m f b s a r n t a s a m y

u c g p

T l

W

F

F

F

F

a

b

m * * * A

6

nce of health problems in parents, as there are strong ssociations between fathers’ and mothers’ mental ealth and that of their offspring.25,26 For example, ale and female adolescents of mothers with suicidal-

ty, compared to their peers whose mothers did not eport suicidality, indicated greater suicide attempts.25

n addition, depression in a mother has been found to e associated with depression recurrence, chronicity, nd severity in young adult sons and daughters.26

Several practical implications result from this study. rofessionals designing teen interventions may want to pecifically target enhancing the parent– child relation- hip. One parent-based intervention targeted the devel- pment of communication skills and strengthening of dult– child relationships through lessons such as re- olving conflicts in a respectful manner by avoiding lame and criticism, exploring mutual needs within ommunication, solving problems constructively, and ncouraging adolescent emancipation.27 Toward the nd of educating parents on suicide risk factors among outh, children of mothers who received the informal rofessionally led intervention groups reported im- roved perception of maternal caring and reduced arent– child conflicts after the 3-month intervention, ompared to youth in the control condition.27 These romising results indicate that mother– child relation- hips are malleable and able to be improved within a

able 3. Girls: parent– child connectedness and emotional h evelsb

Body dissatisfaction

AP (SE) p value

hose opinion is valued for serious decisions? Parents’ opinion 27.75 (2.40) Friends’ opinion 43.40 (1.32) <0.001***

eel you can talk to mother about problems? Quite a bit or very much 26.32 (1.49) Somewhat 32.83 (2.35) 0.013* Not at all or a little 41.98 (2.37) <0.001***

eel you can talk to father about problems? Quite a bit or very much 28.10 (2.19) Somewhat 24.49 (2.26) 0.237 Not at all or a little 35.95 (1.63) 0.004**

eel your mother cares about you? Quite a bit or very much 29.18 (1.23) Somewhat 48.05 (4.67) <0.001*** Not at all or a little 56.91 (5.19) <0.001***

eel your father cares about you? Quite a bit or very much 28.64 (1.31) Somewhat 46.40 (4.05) <0.001*** Not at all or a little 41.60 (3.48) <0.001***

Adjustments were made for grade level, socioeconomic status, race/ Significance levels are reported with the following comparison: valuin uch about, and feeling mother/father cares quite a bit or very mu

p � 0.05; *p � 0.01; **p � 0.001 (all bolded). P, adjusted probability; SE, standard error.

hort period of time. Interventions targeting parent– t

4 American Journal of Preventive Medicine, Volume 30, Num

hild connections and communication may be effective n preventing and reducing the use of health risk ehaviors, while at the same time enhancing emotional ealth. Importantly, results from the current study, in hich over one half of girls (55.4%) and one third of oys (36.6%) did not feel that they could talk to their ather about their problems, underscore the impor- ance of including both parents in improving commu- ication. It is possible that interventions that target others form the foundation, and those targeting

athers construct the protective walls against health risk ehaviors. Furthermore, although there were no con- istent results across parent– child connectedness vari- bles with respect to race/ethnicity, lower SES was egularly associated with lower perceived family con- ectedness (ability to talk to mother or father, percep-

ion of caring from mother or father). While a full ssessment to identify at-risk youth is critical, profes- ionals conducting broader outreach programs (such s school based or as part of a community intervention) ay choose to target interventions toward lower SES

outh, as low SES may be associated with greater need. This study had several strengths that increase the

tility of study findings. The use of a diverse, nonclini- al, population-based sample of girls and boys allows for eneralization to larger populations, exploration of atterns across race/ethnicities and socioeconomic sta-

: adjusted probabilities,a standard errors, and significance

Low self-esteem Depression

P (SE) p value AP (SE) p value

0.10 (2.18) 34.11 (1.38) 4.93 (0.98) <0.001*** 51.54 (2.41) <0.001***

2.44 (1.07) 31.55 (1.57) 1.12 (1.99) <0.001*** 40.46 (2.46) 0.001** 9.72 (2.14) <0.001*** 51.15 (2.39) <0.001***

2.68 (1.58) 30.77 (2.24) 4.64 (1.83) 0.407 29.82 (2.41) 0.765 2.87 (1.37) <0.001*** 45.07 (1.69) <0.001***

6.04 (0.93) 35.67 (.01) 4.88 (4.46) <0.001*** 58.21 (4.61) <0.001*** 7.15 (5.25) <0.001*** 63.52 (5.00) <0.001***

5.87 (0.99) 34.44 (1.37) 0.82 (3.79) <0.001*** 52.02 (4.11) <0.001*** 9.59 (3.29) <0.001*** 55.53 (3.51) <0.001***

city, and parent marital status, and were clustered for school. ents’ opinion, feeling able to talk to mother/father quite a bit or very

ealth

A

3 1

1 2 2

1 1 2

1 3 4

1 3 2

ethni g par

ch.

us, and identification of groups at potential high risk

ber 1 www.ajpm-online.net

f t a b c

s c u p a

c T b l a t c a p r

c b p t n t i t y

T B a a i

T l

W

F

F

F

F

a

b

m * * * A

J

or lower levels of parent– child connectedness. Fur- hermore, the broad range of outcomes assessed yields

greater understanding of the diversity of health risk ehaviors potentially associated with parent– child onnectedness.

The present results need to be contextualized with everal limitations. The questions assessing parent– hild connectedness were brief (and thus cannot be sed in an in-depth and qualitative exploration of the arent– child relationship), and prone to source vari- nce due to self-report by only one reporter (adoles-

able 4. Boys: parent– child connectedness and emotional h evelsb

Body dissatisfaction

AP (SE) p value

hose opinion is valued for serious decisions? Parents’ opinion 11.68 (0.85) Friends’ opinion 18.07 (2.13) 0.002**

eel you can talk to mother about problems? Quite a bit or very much 10.56 (1.02) Somewhat 11.77 (1.54) 0.501 Not at all or a little 17.82 (1.70) <0.001***

eel you can talk to father about problems? Quite a bit or very much 8.36 (1.02) Somewhat 14.07 (1.60) 0.002** Not at all or a little 16.84 (1.46) <0.001***

eel your mother cares about you? Quite a bit or very much 12.31 (0.82) Somewhat 15.61 (3.82) 0.354 Not at all or a little 21.99 (5.80) 0.016*

eel your father cares about you? Quite a bit or very much 11.45 (0.83) Somewhat 21.42 (3.33) <0.001*** Not at all or a little 17.84 (2.86) 0.014*

Adjustments were made for grade level, socioeconomic status, race/ Significance levels are reported with the following comparison: valuin uch about, and feeling mother/father cares quite a bit or very mu

p � 0.05; *p � 0.01; **p � 0.001 (all bolded). P, adjusted probability; SE, standard error.

What This Study Adds . . .

The current study of a large, diverse population of youth expands on previous research using smaller, homogenous samples, and explores the relationship between a broad range of behavioral and emotional health indicators and specific char- acteristics of parent– child communication and caring.

Valuing parent opinions for serious decisions, feeling able to discuss problems with parent(s), and perceiving parent’s caring were related in- versely to substance use, unhealthy weight con- trol, suicide, body dissatisfaction, and depression, and directly to self-esteem.

o

anuary 2006

ent). Some cells of interest yielded small sample sizes. he cross-sectional design cannot address causation etween variables. While parent– child connectedness

ikely has an impact on adolescent well-being, it may lso be that the health-compromising behaviors among eens lead to deterioration of family relations, per- eived levels of parental communication and caring, nd increased reliance on peers’ opinions. Future rospective studies can address the direction of these elationships.

In conclusion, adolescents’ perceptions of parent– hild relationships were significantly associated with ehavioral and emotional health indicators. Healthcare rofessionals and school personnel should familiarize hemselves with the perceived parent– child connected- ess of youth. Interventions may target the mother–

een relationship as a foundation for change, but deally should include both parents when possible for he most comprehensive effort toward preventing outh health risk behaviors.

his study was supported by the Maternal and Child Health ureau (Title V, Social Security Act), Health Resources nd Service Administration, U.S. Department of Health nd Human Services (grant MCJ-270834, DN-S, principal nvestigator).

No financial conflict of interest was reported by the authors

adjusted probabilities, standard errors, and significance

Low self-esteem Depression

P (SE) p value AP (SE) p value

7.47 (0.74) 17.83 (0.98) 4.18 (1.98) <0.001*** 30.80 (2.52) <0.001***

6.16 (0.82) 14.94 (1.17) 6.49 (1.20) 0.815 21.80 (1.96) 0.002** 5.16 (1.65) <0.001*** 28.19 (1.97) <0.001***

5.77 (0.88) 15.26 (1.34) 7.41 (1.24) 0.258 17.27 (1.73) 0.355 2.82 (1.38) <0.001*** 27.68 (1.74) <0.001***

7.50 (0.67) 18.79 (0.94) 9.58 (4.30) <0.001*** 38.21 (5.32) <0.001*** 4.56 (5.07) <0.001*** 33.35 (5.46) 0.002**

7.31 (0.71) 18.02 (0.98) 6.00 (3.06) <0.001*** 34.56 (3.91) <0.001*** 4.91 (2.73) <0.001*** 27.54 (3.39) 0.003**

city, and parent marital status and were clustered for school. ents’ opinion, feeling able to talk to mother/father quite a bit or very

ealth;

A

1

1

1

1 2

1 1

ethni g par

ch.

f this paper.

Am J Prev Med 2006;30(1) 65

R

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

6

eferences 1. Resnick MD, Bearman PS, Blum RW, et al. Protecting adolescents from

harm: findings from the National Longitudinal Study on Adolescent Health. JAMA 1997;278:823–32.

2. Sale E, Sambrano S, Springer JF, Turner CW. Risk, protection, and substance use in adolescents: a multi-site model. J Drug Educ 2003;33:91–105.

3. Allen JP, Hauser ST, Eickholt C, et al. Autonomy and relatedness in family interactions as predictors of expressions of negative adolescent affect. J Res Adolesc 1994;4:535–52.

4. Aydin B, Oztutuncu F. Examination of adolescents’ negative thoughts, depressive mood, and family environment. Adolescence 2001;26:77– 83.

5. Delaney ME. Across the transition to adolescence: 1ualities of parent/ adolescent relationships and adjustment. J Early Adolesc 1996;16:274 –300.

6. Ackard DM, Neumark-Sztainer D. Family mealtime while growing up: associations with symptoms of bulimia nervosa. Eat Disord 2001;9:239 – 49.

7. Fergusson DM, Woodward LJ, Horwood LJ. Risk factors and life processes associated with the onset of suicidal behaviour during adolescence and early adulthood. Psychol Med 2000;30:23–39.

8. Hollis C. Depression, family environment, and adolescent suicidal behav- ior. J Am Acad Child Adolesc Psychiatry 1996;35:622–30.

9. King RA, Schwab-Stone M, Flisher AJ, et al. Psychosocial and risk behavior correlates of youth suicide attempts and suicidal ideation. J Am Acad Child Adolesc Psychiatry 2001;40:837– 46.

0. O’Donnell L, Stueve A, Wardlaw D, O’Donnell C. Adolescent suicidality and adult support: the Reach for Health study of urban youth. Am J Health Behav 2003;27:633– 44.

1. Herman MR, Dornbusch SM, Herron MC, Herting JR. The influence of family regulation, connection, and psychological autonomy on six mea- sures of adolescent functioning. J Adolesc Res 1997;12:34 – 67.

2. Ackard DM, Neumark-Sztainer D. Health care information sources for adolescents: age and gender differences on use, concerns, and needs. J Adolesc Health 2001;29:170 – 6.

3. Neumark-Sztainer D, Story M, Ackard D, et al. The “family meal”: views of

adolescents. J Nutr Educ 2000;32:329 –34.

6 American Journal of Preventive Medicine, Volume 30, Num

4. Neumark-Sztainer D, Story M, Ackard D, et al. Family meals among adolescents: findings from a pilot study. J Nutr Educ 2000;32:335– 40.

5. Neumark-Sztainer D, Story M, Perry C, Casey M. Factors influencing food choices of adolescents: findings from focus-group discussions with adoles- cents. J Am Diet Assoc 1999;99:929 –37.

6. Resnick M, Harris LJ, Blum RW. The impact of caring and connectedness on adolescent health and well-being. J Paediatr Child Health 1993;29:1–9.

7. Neumark-Sztainer D, Wall MM, Story M, Perry CL. Correlates of unhealthy weight-control behaviors among adolescents: Implications for prevention programs. Health Psychol 2003;22:88 –98.

8. Minnesota Department of Children, Families and Learning. Minnesota Student Survey. Roseville: Minnesota Department of Human Services, 2001.

9. Pingitore R. Gender differences in body satisfaction. Obes Res 1997;5:402–9.

0. Rosenberg M. Society and the adolescent self-image. Princeton NJ: Princeton University Press, 1965.

1. Kandel DB, Davies M. Epidemiology of depressive mood in adolescents; an empirical study. Arch Gen Psychiatry 1982;35:1205–12.

2. Breiman L, Friedman JH, Olshen RA, Stone CJ. Classification and regres- sion trees. Belmont CA: Wadsworth International Group, 1984.

3. Barber BK, Olsen JA. Socialization in context: connection, regulation, and autonomy in the family, school, and neighborhood, and with peers. J Adolesc Res 1997;12:287–315.

4. Eisenberg ME, Olson RE, Neumark-Sztainer D, et al. Correlations between family meals and psychosocial well-being among adolescents. Arch Pediatr Adolesc Med 2004;158:792– 6.

5. Lieb R, Bronisch T, Hofler M, Schreier A, Wittchen HU. Maternal suicidality and risk of suicidality in offspring: findings from a community study. Am J Psychiatry 2005;162:1665–71.

6. Rohde P, Lewinsohn PM, Klein DN, Seeley JR. Association of parental depression with psychiatric course from adolescence to young adulthood among formerly depressed individuals. J Abnorm Psychol 2005;114: 409 –20.

7. Toumbourou JW, Gregg ME. Impact of an empowerment-based parent

education program on the reduction of youth suicide factors. J Adolesc Health 2002;31:277– 85.

ber 1 www.ajpm-online.net