The Role of Maternal Education and Nonmaternal Care Services in the Prevention of Children's Physical Aggression Problems

Please download to get full document.

View again

of 8
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Similar Documents
Information Report

Short Stories


Views: 35 | Pages: 8

Extension: PDF | Download: 0

The Role of Maternal Education and Nonmaternal Care Services in the Prevention of Children's Physical Aggression Problems
  ORIGINAL ARTICLE The Role of Maternal Education and NonmaternalCare Services in the Prevention of Children’sPhysical Aggression Problems Sylvana M. Coˆte´, PhD; Michel Boivin, PhD; Daniel S. Nagin, PhD; Christa Japel, PhD; Qian Xu, MSc;Mark Zoccolillo, MD; Marianne Junger, PhD; Richard E. Tremblay, PhD Context : Physicalviolenceisanimportanthealthprob-lem, and low maternal education is a significant risk forthedevelopmentofchronicphysicalaggression(PA).Wehypothesizedthatnonmaternalcare(NMC)servicescouldpreventthedevelopmentofchildhoodPAproblems,de-pending on the age at which the services are initiated. Method :  Children who followed a trajectory of atypi-callyfrequentPAbetween17and60monthsofageamongapopulationsampleof1691Canadianfamilieswereiden-tified. Maternal education and NMC were considered inpredictinggroupmembershipwhilecontrollingforcon-founding family characteristics. Results :  Children of mothers with low education lev-els (ie, no high school diploma) were less likely to re-ceive NMC. Those who did receive such care had sig-nificantlylowerriskofahighPAtrajectory.ResultsfromlogisticregressionsindicatedthatNMCreducedtheriskofhighPA,especiallywheninitiatedbeforeage9months(odds ratio,0.20; 95% confidence interval, 0.05-0.90).Childrenofmotherswhograduatedfromhighschoolwereless at risk of PA problems, and NMC had no additionalprotective effect. Conclusions :  Nonmaternal care services to children of mothers with low levels of education could substan-tially reduce their risk of chronic PA, especially if pro-vided soon after birth. Because children most likely tobenefitfromNMCservicesarelesslikelytoreceivethem,special measures encouraging the use of NMC servicesamong high-risk families are needed.  Arch Gen Psychiatry. 2007;64(11):1305-1312 H IGH FREQUENCY OF PHYSI -calaggression(PA)isthecentral feature of severeconduct disorder 1 and isassociated with a widerangeofsocial,mental,andphysicalhealthproblems. 2-4 ThesrcinofPAproblemscanbe traced back to early childhood, andstudies have specifically shown that ma-ternal characteristics, especially low lev-els of education, are among the best pre-dictors of high PA from early childhoodto adolescence. 5-9 Cross-fostering animal studies showthatthenegativeconsequencesofpoorma-ternal care can be prevented by provid-ing high-quality surrogate mother-ing. 10-13 Similarly, in humans, a fewexperimental studies offering nonmater-nal care (NMC) services to young chil-dren from high-risk families provide evi-dence for their protective role. 14,15 However, it is not clear that the effects of suchexperimentalprograms,designedtobe of the highest quality, are comparabletotheeffectsofNMCservicesatthepopu-lation level, which typically are of lowerquality. 16 Nonmaternal care services areprovided by a person chosen to care for achild, usually during the day when themother is working. Nonmaternal care in-cludes various types of day care arrange-ments(family-orcenter-based)usedregu-larly during the preschool years.The objective of the present study wasto test, with a representative sample of apopulation, whether NMC services couldprevent the development of childhood PAproblemsandwhetherthemagnitudeoftheeffect varied according to the age at whichthe child began to receive NMC services.Mostchildreninpresent-dayindustri-alizedsocietiesreceiveNMCservicesdur-ing the preschool years. 17,18 Children of motherswithlowlevelsofeducation,how-ever, are less likely to receive NMC be-causetheirmotherislesslikelytobework-ing. 18,19 Childrenofmotherswithlowlevelsofeducationarealsomorelikelytobeex-posed to a wide range of risks in theirhome,suchaspoorparenting, 20,21 lowlev-els of stimulation, 20 and poverty. 22 Thus,childrenofpoorlyeducatedmothers(com-pared with children of better-educated Author Affiliations: Departments of Social andPreventive Medicine (Dr Coˆte´)and Psychology and Paediatrics(Dr Tremblay) and ResearchUnit on Children’s PsychosocialMaladjustment (Ms Xu),International Laboratory forChild and Adolescent MentalHealth Department, Universityof Montreal, Que´bec, Canada,and Inserm U669, Paris, France(Drs Coˆte´ and Tremblay);Department of Psychology,Laval University, Que´bec(Dr Boivin); Carnegie MellonUniversity, Heinz School of Public Policy, Pittsburgh, PA(Dr Nagin); Department of Education, Universite´ duQue´bec à Montre´al (Dr Japel);Department of Psychiatry,McGill University, Que´bec(Dr Zoccolillo); andDepartment of Psychology,Utrecht University, Utrecht,the Netherlands (Dr Junger). (REPRINTED) ARCH GEN PSYCHIATRY/VOL 64 (NO. 11), NOV 2007 WWW.ARCHGENPSYCHIATRY.COM 1305 ©2007 American Medical Association. All rights reserved.  at University of Montreal, on November 6, 2007 www.archgenpsychiatry.comDownloaded from   mothers) may be exposed to higher risks of PA becausethey are more likely to spend time in the context of anat-risk home environment.Therefore, at the most basic level, NMC may be pro-tectivebecauseitprovidesareductioninexposuretofam-ilyrisks.Withinthiscontext,theprotectiveeffectshouldbe more pronounced if NMC experiences are initiatedearlierinthechild’slife.EarlierinitiationofNMCwouldnot only lead to a greater reduction of exposure to fam-ily risk but would do so at younger ages, when the brainand behavior repertoire are more malleable. 23 A recentstudy found that NMC services during the first year of life were related to better receptive language at 4 yearsamong children of low socioeconomic status. 24 How-ever, to our knowledge, no study has examined the ex-tenttowhichthetimingofNMCinitiation(earliervslater)isrelatedtothelong-termdevelopmentofPAamonghigh-risk children.To adequately examine the role of NMC in a popula-tion-based sample, 2 important issues need to be con-sidered.First,theassociationbetweenNMCandPAmayvaryaccordingtochildren’sfamilyriskstatus.FamilyrisksforPAincludelowmaternaleducation,lowincome,pa-rental psychiatric disorders, and poor parenting prac-tices. 5-9,25  WhileNMCmaybeassociatedwithbetterout-comes among the high-risk children, 14,19,26,27 it may beassociated with poorer outcomes among children whoare not from high-risk families, 25,28 especially if it is ini-tiatedininfancy. 29 Indeed,forchildrenwhoarenotfromhigh-risk families, their NMC may be of a lower qualitythan the care they would receive at home.Second, the “social selection” of children into NMCcanconfoundtheassociationbetweenNMCandPAprob-lems. If children of less educated mothers are less likelyto receive NMC services 19 but are more likely to benefitfrom them, then NMC services at the population levelwill not reduce the disparities in PA problems betweenchildren of low- and better-educated mothers. 30 In the present study, we used a population sample of newborns to test the hypothesis that: (1) young chil-dren of mothers with low levels of education (no highschool diploma) benefit more from receiving NMC ser-vices than do children of better-educated mothers and(2) the protective effects are greater if NMC is initiatedbefore rather than at or after 9 months of age. Maternaleducationwasusedasaproxyvariableforchildren’sriskstatus because it is the most robust and reliably mea-sured predictor of PA across studies 5-9,25 and because ithas practical value in the context of service delivery. METHODS SUBJECTS Asampleof1759infantsrepresentativeofchildrenbornintheprovince of Que´bec (Canada) from 1997 through 1998 wereselectedthroughbirthregistriesandfollowedupannuallyfrom5 to 60 months of age. Yearly home interviews were con-ducted with the mothers to obtain information about family,parent,andchildcharacteristicsandbehaviors.Informedwrit-tenconsentwasobtainedfromallparticipants.Approvalofthestudy was obtained by the Que´bec Institute of Statistics’ insti-tutional review board. Multivariate analyses were conductedon the 1691 children (96.1%) for whom complete data wereavailable. All analyses were weighted to ensure the represen-tativeness of the sample.  Table 1  describes the demographiccharacteristics of the sample. MEASURES The main outcome variable evaluated was children’s PA at 17,30, 42, 54, and 60 months. Mothers were asked to rate theirchild on a frequency scale indicating whether the child never(codedas0),sometimes(codedas1),oroften(codedas2)ex-hibits PA. The following items were included in the PA scale:hits,bites,kicks;fights;andbulliesothers.Scoresrangedfrom0to6.Theseitemswerealsousedinpreviousstudies. 5,8,19 Theinternal consistency of mothers’ PA ratings was assessed overtime (17-60 months) (  =0.83) because the outcome measureincluded all 5 assessments of PA. PREDICTOR VARIABLES Coding is indicated in parentheses for each categorical vari-able.TheNMCvariabledistinguishedamongchildrenwhoini-tiated NMC at different time points between 3 and 31 monthsof age. Age cutoffs were established with reference to the ageof the children at the different data collection points: before 9months of age for time 1 (age range, 3-8 months), before 20months for time 2 (age range, 16-19 months), and before 32months for time 3 (age range, 27-31 months). Most childrenreceived NMC in family day care settings outside the home,eitherbynonrelatives(40.7%)orrelatives(30.9%).Fifteenper-cent received NMC in center-based settings. The remaining(13.6%) received NMC in various other settings. We com-puted a variable for the mean number of hours in NMC thatreflected the mean number of hours in NMC services acrossthe entire preschool period.Maternal education was treated as a dummy variable indi-cating if the mother had a high school diploma (0) or not (1). Table 1. Demographic Characteristics of the Sample CharacteristicNo. (%)of 1691 Subjects SexFemale 844 (49.9)Male 847 (50.1)Maternal age (birth of target child), y  21 1497 (88.5)  21 194 (11.5)Maternal educationHigh school diploma 1395 (82.5)No high school diploma 296 (17.5)Family statusIntact (married or common-law union) 1551 (91.7)Separated, divorced, widowed 65 (3.8)Always single 75 (4.4)Family annual income, $  30000 543 (32.1)30000-60000 674 (39.9)  60000 474 (28.0)Timing of NMCNever in NMC (before school entry) 111 (6.6)Initiated before age 9 mo 234 (13.8)Initiated at or after 9 mo 1346 (79.6)Abbreviation: NMC, nonmaternal care. (REPRINTED) ARCH GEN PSYCHIATRY/VOL 64 (NO. 11), NOV 2007 WWW.ARCHGENPSYCHIATRY.COM 1306 ©2007 American Medical Association. All rights reserved.  at University of Montreal, on November 6, 2007 www.archgenpsychiatry.comDownloaded from   POTENTIAL CONFOUNDING FACTORS(SOCIAL SELECTION FACTORS) Four categories of variables were examined as potential con-founds. Associations between the predictor variables and theoutcome were examined and some variables were categorizedto reflect the best risk cutoff. Maternal Characteristics and Maternal Behavior  Age of the mother   was treated as a dummy variable reflecting 21yearsoryoungeratthebirthofthetargetchild(0)orolderthan21years(1).Becomingamotherbeforeage21yearswasshownto be a valid indicator of risk in the Quebec population. 9 Maternal antisocial behaviors  before the end of high schoolwere assessed via a questionnaire to the mothers, inquiringwhether they had exhibited 5 conduct problems (eg, havingbeen in more than 1 fight that they started or had stolen morethan once). 31 The number of symptom counts was used in theanalyses. Maternaldepression wasassessedthroughalifetimedepres-sion section modified from the Diagnostic Interview Sched-ule. 32,33 Mothers who reported having had at least 1 major de-pressive episode during their lifetime (1) were distinguishedfrom those who did not (0). The maximum number of drinksin 1 occasion (at first assessment) was also examined as a po-tentialconfound.Finally, maternal work reflectedwhetherthemother was working or studying (0) or not (1) at first assess-ment. Child Characteristics The sex of the child  was coded as 0 for girls and 1 for boys.  Childtemperament wasratedbythemotherusing7itemsfromthedif-ficult temperament scale of the Infant Characteristics Question-naire. 34 Each item (for example, how often the child was diffi-culttocalmorsootheorwasfussyandifthechildwaseasilyupset)rangedfrom1to7.Internalconsistencycoefficient(  )was0.84.Thechild’sracewascodedaswhite(0)orother(1).Informationonthebirthofthechildwasobtainedfromhospitalrecordsandcoded as preterm (1) if the child was born before the 37th weekofgestationorontime(0)ifthechildwasbornatorafter37weeks’gestation. Low birth-weight was coded as yes (1) if the infantweighed2500gorlessorno(0)iftheinfantweighedmorethan2500 g. Information on the health of the child was obtainedthrough a question to the mother inquiring whether her child’shealth at birth was good (0) or poor (1). Family Demographics 5 Months After Birth Insufficient household income  was calculated using StatisticsCanada on the basis of the family annual income, the numberof people in the household, and the family zone of residence(urbanvsruralbasedonpopulationdensity).Incomewascodedassufficient(0)ornotsufficient(1). Familysituation wascodedaccording to whether the parents were married or living to-gether(0);separated,divorced,orwidowed(1);orthemotherhasalwaysbeensingle(2). Numberofchildreninthefamily wascodedaccordingtothepresence(1)orabsence(0)ofsiblings. Family Processes 5 Months After Birth Family functioning  was assessed with an 8-item scale measur-ing how well the family functioned (eg, there are lots of badfeelings in our family;  =0.97), 35 with higher values indica-tive of dysfunction in the family.Mothers completed a questionnaire on parenting percep-tionsandbehaviors. 36 Mothersansweredusingascaleof0(“notatallwhatIthinkordid”)to10(“exactlywhatIthinkordid”).The following 4 dimensions reflected the quality of mothers’interactions with their 5-month-old infant: self-efficacy, pa-rental impact, coercive parenting, and overprotection. Inter-nalconsistencycoefficients(  )were0.70,0.69,0.62,and0.58,respectively. ANALYSES Theanalyseswereconductedin3steps:identificationofanatypi-cally elevated trajectory of PA, identification of variables thatcouldpotentiallyconfoundtheassociationbetweenNMCandPA, and testing the modifying role of NMC. IDENTIFYING AN ATYPICALLY ELEVATEDTRAJECTORY OF PA Following previous studies, 5,8,19 distinctive clusters of devel-opmental trajectories were identified using a semiparametricmixture model described by Broidy et al 37 and Nagin. 38 Themodel selection was based on the Bayesian information crite-rion.Modelswith1to6groupswereestimated.ThemodelthatminimizedtheBayesianinformationcriterionwaschosen.Theprocedure yields 2 outcomes: (1) the probability of member-shipinthehighPAtrajectory,whichisacontinuousmeasure,ranging from 0 to 1 and (2) actual membership in the high PAtrajectory, derived from assignment to the high PA trajectory(coded as 1) or to another trajectory (coded as 0). IDENTIFYING CONFOUNDING FACTORS  Wetestedthebivariateassociationsbetweenseveralfamilyandchild risk variables and (1) age at initiation of NMC and (2)the probability of membership in the high PA trajectory andcontrolledforthevariablesassociatedwithbothNMCandPA. TESTING THE MODIFYING ROLE OF NMC Multiple regression analyses were conducted with the prob-ability of belonging to a high trajectory of PA as the outcome.The predictors were entered in a stepwise fashion in the fol-lowing order: (1) the identified confounder variables, (2) themain effect of maternal education and NMC, and (3) the in-teractionbetweentimingofNMCandmaternaleducation.Dif-ferent cutoffs were considered with regard to the timing of theNMC variable. The same model was tested using logistic re-gressionwithmembershipinthehighPAtrajectoryastheout-come. RESULTS IDENTIFYING AN ATYPICALLY ELEVATEDTRAJECTORY OF PA The best trajectory model comprised 3 groups. Seven-teen percent of children were estimated to belong to thehigh PA trajectory group. These children were atypicalin that they represented a small group of children withclearly higher levels of PA during early childhood (be-tween 17 and 60 months).  Figure 1  illustrates the tra- jectory model. (REPRINTED) ARCH GEN PSYCHIATRY/VOL 64 (NO. 11), NOV 2007 WWW.ARCHGENPSYCHIATRY.COM 1307 ©2007 American Medical Association. All rights reserved.  at University of Montreal, on November 6, 2007 www.archgenpsychiatry.comDownloaded from   IDENTIFYING CONFOUNDING FACTORS Among the factors for which we tested the associationsbetweenNMCandthehighPAtrajectory,5variableswerefound to be significantly associated with both NMC andPA:thepresenceofsiblings,insufficientincome,mater-nal depression, maternal work, and family status. Thesevariableswereusedascontrolsintheanalyses. Table2 presents the associations among NMC, the high PA tra- jectory, and the potential confounding variables. THE MODIFYING ROLE OF NMC Table 3  presents the results of the multiple regressionanalysisthatexaminedtheassociationamongNMC,ma-ternaleducation,andPA.Enteringtheconfoundingvari-ablesinstep1revealedsignificantassociationsforallvari-ables.Step2revealedasignificantmaineffectofmaternaleducation but not of NMC. Step 3 revealed a significantinteraction between timing of NMC, as defined by thecutoff of 9 months, and maternal education.  Figure 2 illustrates the interaction. All subsequent analyses usedthe “before 9 months” vs “at or after 9 months” cutoff.Follow-up analyses were conducted to test the differ-encesbetweentheinteractiongroups.Cohen’s d effectsizeswere calculated according to the following equation 39 : d =(X NMC −X MC )/S pooled,where(S is the pooled standard deviations of the maternal care(MC) and NMC groups; x, mean; n, number of subjects;and s, standard deviation of each group.) Negative  d s re-flectthatNMCisassociatedwithalowerprobabilityofhighPA.These analyses revealed 3 main findings. First, chil-drenofmotherswithlowlevelsofeducationwereatsig-nificantlylowerriskofhighPAiftheyreceivedNMCbe-fore 9 months (n=39; mean, 0.14; SD, 0.31) or after 9months (n=221; mean, 0.23; SD, 0.36) compared withthose who never received NMC (n=36; mean, 0.36; SD,0.40).TheeffectsizewaslargeforNMCinitiatedbefore9 months ( d =−0.62; SD, 0.24; 95% confidence interval[CI],−1.09to−0.16)andmoderateforNMCinitiatedator after 9 months ( d =−0.37; SD, 0.18; 95% CI, −0.73 to−0.02).Second,childrenofmotherswhograduatedfromhigh school were not at higher risk for PA if they re-ceived NMC, whether it was initiated before 9 months(n=196; mean, 0.18; SD, 0.30) or at or after 9 months(n=1125;mean,0.17;SD,0.31)comparedwiththosewhonever received NMC (n=75; mean, 0.15; SD, 0.30). WetestedthesamemodelusinglogisticregressionandmembershipinthehightrajectoryofPA(comparedwithanyoftheother2trajectorygroups)astheoutcome.Weobtained the same pattern of results. Children of moth-ers with low education levels who received NMC before9 months were less likely to belong to the high PA tra- jectory(oddsratio[OR],0.20;95%CI,0.05-0.90)com-paredwithchildrenwhoneverreceivedNMC.ChildrenofmotherswithloweducationlevelswhoreceivedNMCatorafter9monthswerealsolesslikelytobelongtothehigh PA trajectory than those who never received NMC(OR, 0.36; 95% CI, 0.13-1.04), but the effect was onlymarginally significant ( P =.06). ADDITIONAL ANALYSES AND RESULTS  We conducted additional analyses to examine whetherthe amount of time spent in NMC could account for thepattern of findings. Introducing a variable reflecting themean number of hours in NMC did not modify the pat-tern of results. The variable is not included in the finalmodel (Table 3), because it was not significantly associ-atedwiththeoutcome( β =.05,SD,1.85; P =.07).Wealsoconducted additional analyses to examine whether thechildren of the high PA trajectory group who benefitedfromearlyNMCweredeflectedtothemediumorthelowPAtrajectorygroup.Todoso,wetestedthesamemodelcomparingdifferenttrajectorygroups.Wefoundthesameinteraction effect for the model comparing membershipin the high vs medium trajectory, but not in the modelopposing the high vs low trajectory groups. This resultindicates that children of the high trajectory group ben-efiting from early NMC were deflected to the mediumtrajectory(andnotthelowtrajectory)group.Finally,wetested all the 3-way interactions among the predictorsusingabackwardregressionprocedure.Noneofthe3-wayinteractions were significant. COMMENT The objectives of the study were to test, at the popula-tionlevel,whetherNMCservicescanpreventearlychild-hood PA problems and whether the protective role is afunction of the age at which NMC services are initiated. We found, using multiple regression, that NMC re-duced the risk of following an atypically elevated trajec-tory of PA among children of mothers with low educa-tionlevels.TheeffectsizewaslargeifNMCwasinitiatedbefore 9 months ( d =−0.62) and moderate if initiated ator after 9 months ( d =−0.37). Nonmaternal care was notassociatedwithanincreaseintheriskofPAamongchil-dren of better-educated mothers even if it was initiated S pooled =√ {[(n NMC − 1 )(s NMC2 ) +  (n MC − 1 )(s MC2 )]/(n NMC   +  n MC − 2 )}. Age, mo     P    h   y   s    i   c   a    l    A   g   g   r   e   s   s    i   o   n    S   c   o   r   e 1729425460Low 32.5%High 17%Moderate 50.5% Figure 1.  Trajectories of physical aggression between age 17 and 60 months(n=1758). (REPRINTED) ARCH GEN PSYCHIATRY/VOL 64 (NO. 11), NOV 2007 WWW.ARCHGENPSYCHIATRY.COM 1308 ©2007 American Medical Association. All rights reserved.  at University of Montreal, on November 6, 2007 www.archgenpsychiatry.comDownloaded from   beforeage9months.Thesamepatternofresultsemergedwith logistic regression, indicating a protective effect of NMC,especiallywheninitiatedbeforeage9months(OR,0.20; 95% CI, 0.05-0.90).TheprotectiveeffectofNMCisconsistentwiththere-sults of experimental studies showing long-term reduc-tionofpooroutcomesinchildrenofmotherswithlowedu-cationlevelswhoreceivedhigh-qualitychildcareservicesduringthepreschoolyears. 14,26 Oneoftheseexperimentalstudies 26 didshowanimpressivereductioninantisocialbe-haviorduringadolescenceandearlyadulthood,butdidnotdocumenttheeffectsonearlydevelopmentofaggression,asinthepresentstudy.Theresultsarealsoinlinewithpre-vious correlational studies showing a protective effect forchildrenfromhigh-riskfamilies, 21,28 aswellaswithadop-tionstudiesshowingthatcross-fosteringofhigh-riskchil- Table 2. Bivariate Associations Between Potential Confounders, PA, and Timing of NMC Probabilityof High PATrajectory a P  ValueTiming of NMC b,c Never(n=111)After 9 mo(n=1346)At or Before 9 mo(n=234) P  Value Sex b Females 0.13 (0.27)  .001 55 (49.5) 674 (50.1) 115 (49.1).96Males 0.23 (0.36) 56 (50.5) 672 (49.9) 119 (50.9)Ethnicity (n=1677)Nonwhite 0.18 (0.32).73 33 (30.0) 422 (31.4) 76 (32.5).84White 0.18 (0.32) 77 (70.0) 915 (68.0) 154 (65.8)Child health at birthPoor 0.38 (0.45)  .05 0 9 (0.7) 1 (0.4).58Good 0.18 (0.32) 111 (100.0) 1336 (99.3) 234 (100.0)Premature birthYes 0.18 (0.34).96 8 (7.2) 78 (5.8) 17 (7.3).60No 0.18 (0.32) 103 (92.8) 1268 (94.2) 217 (92.7)Low birth weightYes 0.15 (0.27).49 7 (6.3) 14 (1.0) 49 (20.9)  .05No 0.17 (0.31) 106 (95.5) 1344 (96.5) 227 (16.9)Maternal educationNo high school diploma 0.23 (0.36)  .05 36 (32.4) 221 (16.4) 39 (16.7)  .001High school diploma 0.17 (0.31) 75 (67.6) 1124 (83.5) 196 (83.8)Age at childbearing, y  21 0.17 (0.31).47 9 (8.1) 146 (10.8) 39 (16.7)  .05  21 0.18 (0.32) 102 (91.9) 1200 (89.2) 195 (83.3)Maternal depressionYes 0.23 (0.35)  .001 34 (30.6) 296 (22.0) 42 (17.9)  .05No 0.17 (0.31) 77 (69.4) 1050 (78.0) 192 (82.1)Maternal work before 9 moYes 0.16 (0.30)  .001 28 (25.2) 952 (70.7) 230 (98.3)  .001No 0.23 (0.35) 83 (74.8) 394 (29.3) 4 (1.7)Presence of siblingsYes 0.25 (0.36)  .001 88 (79.3) 759 (56.4) 125 (53.4)  .001No 0.09 (0.23) 23 (20.7) 587 (43.6) 109 (46.6)Insufficient incomeNo 0.16 (0.30)  .001 56 (50.5) 1006 (74.7) 181 (77.4)  .001Yes 0.22 (0.35) 55 (49.5) 340 (25.3) 53 (22.6)Family status b Always single 0.10 (0.23)  .0511 (9.9) 49 (3.6) 15 (6.4)  .05Separated, divorced, widowed 0.23 (0.37) 4 (3.6) 55 (4.1) 6 (2.6)Married or common-law union 0.18 (0.32) 96 (86.5) 1242 (92.3) 213 (91.0)Maternal conduct problems (n=1639) d 0.09   .001 0.72 (0.78) a 0.84 (0.95) a 0.77 (0.97) a .26Maternal alcohol consumption e 0.08   .001 0.79 (2.10) a 0.93 (2.76) a 1.33 (2.83) a .08Difficult temperament 0.03 .10 2.53 (1.55) a 2.75 (1.65) a 2.58 (1.47) a .16Age at which infant stopped breastfeeding, mo −0.04 .10 4.33 (5.39) a 4.12 (4.74) a 3.05 (4.06) a  .05Family dysfunction (n=1678) 0.10   .001 1.93 (1.41) a 1.76 (1.45) a 1.63 (1.51) a .19Coercive parenting (n=1636) 0.04 .08 1.07 (1.37) a 1.08 (1.47) a 1.11 (1.50) a .95Parental self-efficacy (n=1634) −0.09   .001 8.93 (1.26) a 8.75 (1.14) a 8.77 (1.11) a .35Perception of parental impact (n=1646) −0.01 .77 7.82 (2.49) a 8.25 (1.99) a 8.62 (1.68) a  .05Parental overprotection (n=1652) −0.04 .16 6.05 (2.33) a 5.53 (2.43) a 4.74 (2.27) a  .001Abbreviations: NMC, nonmaternal care; PA, physical aggression. a Data are presented as mean (SD) and  r  . b N=1691 unless otherwise indicated. c Data are presented as number (percentage) unless otherwise indicated. d Before the end of high school. e Maximum number of drinks in one occasion at time 1 (5 mo). (REPRINTED) ARCH GEN PSYCHIATRY/VOL 64 (NO. 11), NOV 2007 WWW.ARCHGENPSYCHIATRY.COM 1309 ©2007 American Medical Association. All rights reserved.  at University of Montreal, on November 6, 2007 www.archgenpsychiatry.comDownloaded from 
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!