Identity Verified Thinker in Lifestyle / Family & Parenting / Children
Marc Weissbluth
Marc Weissbluth
Dr. Weissbluth graduated from Stanford University and Washington University Medical School. A father of four sons and a Board Certified physician for more than thirty years, he completed his pediatric training at St. Louis Children's Hospital and is a Professor of Clinical Pediatrics at Northwestern

Sleep & Bedtimes: Old & New Data

Sep. 21, 2009 3:00 am
Categories: Uncategorized
Keywords: None

Here's what came out on CNN yesterday: Enforcing Bedtimes Improves Kids' Health

Here's what was written ten years ago:

Journal of Developmental & Behavioral Pediatrics

Issue: Volume 20(3), June 1999, p 164–169

Copyright: © 1999 Lippincott Williams & Wilkins, Inc.

Publication Type: [ORIGINAL ARTICLE]

ISSN: 0196-206X

Accession: 10393073

Keywords: Child Behavior Disorders, Child Behavior Disorders, Child, Preschool, Female, Humans, Male, Psychiatric Status Rating Scales, Psychological Tests, Severity of Illness Index, Sleep Disorders, Sleep Disorders[ORIGINAL ARTICLE] Sleep and Behavior Problems Among Preschoolers



This study described the relationship between amount of sleep and behavior problems among preschoolers. Participants were 510 children aged 2 to 5 years who were enrolled through 68 private pediatric practices. Parents reported on the amount of sleep their child obtained at night and in 24-hour periods. With demographic variables controlled, regression models were used to determine whether sleep was associated with behavior problems. The relationship between less sleep at night and the presence of a DSM-III-R psychiatric diagnosis was significant (odds ratio = 1.23, p = .026). Less night sleep (p < .0001) and less sleep in a 24-hour period (p < .004) were associated with increased total behavior problems on the Child Behavior Checklist; less night sleep (p < .0002) and less 24-hour sleep (p = 90th percentile on the Total Behavior Problems scale of the CBCL), the child's mother was invited to a second-stage evaluation. Subsequently, two children who screened low were matched with each child who screened high on age, sex, and race/ethnicity and were invited to participate. There were 56.3% (N = 510) of the eligible families who completed the second-stage evaluation. Those completing that evaluation did not differ from those who did not participate on the measures of age, sex, race/ethnicity, social class, marital status, or Total Behavior Problems score on the CBCL. The main reason given for not participating seemed to be distance from the study site at which evaluations were being conducted. Of the 510 children who completed the second-stage evaluation, 37% (N = 191) had screened high and 63% (N = 319) had screened low (< 90th percentile on the CBCL Total Behavior Problems scale). Age distribution was relatively consistent (2 years, 25.8%; 3 years, 24.3%; 4 years, 31.9%; 5 years, 18.0%), and approximately two thirds were non-Hispanic white (67.1 %, hereafter referred to as “white”; African Americans, 18.8%; Hispanic, 7.2%; other racial/ethnic origin, 6.8%) and male (59.9%; female, 40.1%). For data analyses, “minority” referred to the combined group of African American, Hispanic, and “other” participants. The majority (60.8%) were from the two lower Hollingshead 16 social classes, but all classes were represented. Informed consent was obtained. Further details on inclusion criteria are available elsewhere.14


Demographic Information

A questionnaire concerning the child's age, sex, racial/ethnic group, family size and membership, socioeconomic status (SES), and parents' marital status was completed by the mothers.

Achenbach CBCL.8,9

The CBCL, a widely used measure of children's behavior problems, was completed by the mothers. The scale yields estimates of internalizing problems (e.g., anxiety, depression), externalizing problems (e.g., defiant, oppositional, overactive behaviors), and total behavior problems.

Rochester Adaptive Behavior Inventory (RABI).17,18

The RABI is a semistructured interview administered to the mother. The scale yields measures of child cooperation, friendships, timidity, fearfulness, activity level, imaginary play, symptomatic behavior, whininess, demanding attention, depression, and persistence. Interrater reliability across scales was .84 to .99.14

Play Observation

Children were observed during a play session following the procedures described by Forehand and McMahon.19 Videotapes were reviewed by the two clinical child psychologists who provided diagnoses.

Developmental Evaluation

Each child was administered an age-appropriate developmental evaluation using the Bayley Scales of Infant Development 20 or the McCarthy Scales of Children's Abilities.21

Sleep Characteristics

During the interview, the mothers were asked to report the usual time at which their child went to sleep and awoke each day, the number of naps per week, and the average length of naps. Amount of night sleep was calculated from estimates of usual time to sleep and time to wake up. An estimate of average sleep in a 24-hour period combined usual night sleep plus average daily nap time.


Mothers were invited to participate in the study at the time of a pediatric office visit. Those who consented to participate then completed a demographic questionnaire and the CBCL. Participants who screened high on the CBCL and matched screen-low children were seen for the second-stage visit. At that visit, a psychology graduate student administered the CBCL, RABI, and sleep interview to the mother and conducted the play observation and developmental evaluation with the child.

After the information was collected at the second-stage evaluation, two Ph.D.-level, licensed clinical child psychologists reviewed each protocol. The protocol included the results of the RABI, CBCL, developmental evaluation, and play session; the process paralleled that which would be used in clinical situations in which the clinician might use a parental interview, questionnaire, testing, and observation to assess the child and assign a diagnosis. Diagnoses were assigned after a “best estimate” procedure.22 In this procedure, each psychologist independently assigned a DSM-III-R diagnosis 23 by following the guidelines of the DSM-III-R as closely as possible (DSM-III-R was used because data collection was initiated before DSM-IV was established, and it was continued to be used throughout the longitudinal study to maintain continuity in diagnoses). If the independent diagnoses disagreed, the psychologists met and reached consensus on the diagnoses to be assigned. The psychologists could assign multiple Axis I diagnoses, but two diagnoses proved sufficient for this age group. The reliability of assigning diagnoses was moderately high (Yule's Y, .62–.99 for specific Axis I diagnoses) and was comparable to that of studies of older children.11

Statistical Procedures

Univariate statistical procedures were used to describe the relationships among sleep variables and between sleep characteristics and demographic variables (age, sex, race/ethnic group, SES). Multivariate statistical analyses were then conducted. These analyses were conducted to determine whether sleep characteristics were related to psychopathology when the effects of demography were removed. Outcome variables included the three-dimensional characteristics of psychopathology (CBCL Total, Internalizing, and Externalizing Behavior Problems scores) and the presence versus absence of a DSM-III-R diagnosis.

Forward stepwise multiple regression analyses were used when the outcome measure was one of the continuous CBCL scales. A total of six multiple-regression analyses were conducted. One set of analyses examined the strength of the relationship between demographic variables, night sleep, and each of the CBCL outcome measures separately; another set of analyses examined the strength of the relationship between demographics, 24-hour total sleep, and each CBCL scale individually. When the outcome measures were dichotomous (presence vs absence of a DSM-III-R diagnosis), logistic regression analyses were used. Two separate logistic regression analyses were conducted, one for night sleep and a second in which the amount of 24-hour sleep was entered. All demographic variables were entered in each equation, with all variables entered in a single step.

Finally, if the sleep variable was significantly associated with behavior problems in any of the analyses, one-way analyses of variance were conducted to determine whether there was a “dose-response” relationship between the amount of sleep and behavior problems.


Bivariate Analyses

Relationship Between Sleep Characteristics. There were high correlations between the amount of night sleep and total (24-hr) sleep (r - .73, p < .001) and moderate correlations between total sleep in a 24-hour period and early bedtime (r = -.39, p < .001), later awakening (r = .45, p < .001), more naps (r = .44, p < .001), and longer naps (r = .50, p < .001). Length of nighttime sleep was correlated with earlier bedtime (r = -.52, p < .001) and later awakening (r = .62, p < .001) and nap frequency (r = -.19 p < .001), but it was unrelated to nap length (r = -.07, p < .001). Nap length was significantly related to nap frequency (r = .68, p < .001).

Demographics and Sleep Patterns

Age differences in the time that children go to bed (range, 8:58 p.m. for 3 year olds to 9:14 p.m. for 2 year olds), time that children awaken (range, 7:24 a.m. for 4 year olds to 7:29 a.m. for 3 and 5 year olds), and the amount of nighttime sleep (range, 10:16 hr for 2 year olds to 10:32 hr for 3 year olds) were not significant. Estimates of nap length (2 year olds, 1:44 hr; 3 year olds, 1:17 hr; 4 year olds, 1:04 hr; 5-year-olds, :32 hr; F = 33.41, p < .001) and nap frequency (2 year olds, 5:12 hr; 3 year olds, 3:20 hr; 4 year olds, 2:16 hr; 5 year olds, :31 hr; F = 67.12, p < .001) were significant, reflecting a decrease with age. Total amount of 24-hour sleep (2 year olds, 11:45 hr; 3 year olds, 11:28 hr; 4 year olds, 10:54 hr; 5 year olds, 10:40 hr; F = 19.55, p < .001) was significantly related to age, with 2 year olds reportedly sleeping an average of 1 hour per day longer than 5 year olds.

There were no significant sex differences on any of the sleep variables. White versus minority differences reflect significantly later bedtimes for minority children (whites, 8:55 p.m.; minorities, 9:25 p.m.; t = 5.84, p = .001), significantly later awakenings for minority children (whites, 7:23 a.m.; minorities, 7:36 a.m.; t = 2.15, p = .033), and significantly less night sleep (whites, 10:29 hr; minorities, 10:11 hr; t = 2.75, p = .006), but minority children experienced longer naps (whites, 1:05 hr; minorities, 1:26 hr; t = 3.99, p = .001) and more naps per week (whites, 2.67; minorities, 4.15; t = 5.61, p = .001). Overall, with white children obtaining more sleep at night and minority children having more naps, the total amount of sleep in a 24-hour period did not differ (whites, 11:11 hr; minorities, 11:17 hr; t = .83, p = .41).

Socioeconomic status (SES) differences were examined by comparing the social class group at greatest risk for emotional problems (Hollingshead Class 5, N = 69) with children from the four higher classes. Minority children tended to be overrepresented in that social class, so bivariate results paralleled those for race/ethnicity in certain areas. Children from the lowest SES class went to sleep later (lowest SES, 9:29 p.m.; higher SES, 9:00 p.m.; t = 3.7, p = .001), awakened later (lowest SES, 7:49 a.m.; higher SES, 7:23 a.m; t = 3.02, p = .003), took longer naps (lowest SES, 1:27 hr; higher SES, 1:10 hr; t = 2.34, p = .022), and took more frequent naps (lowest SES, 4.33; higher SES, 2.95; t = 3.80, p = .001). There was a trend in the direction of more total sleep in 24 hours among the lowest SES children (lowest SES, 11:26 hr; higher SES, 11:12 hr; t = 1.69, p = .094), whereas SES groups did not differ in amount of nighttime sleep (lowest SES, 10:22 hr; higher SES, 10:23 hr; t = .27, p = .787).

Multivariate Analyses: The Relationship of Sleep' and DSM-III-R Diagnosis

In a logistic regression model including the amount of sleep at night and demographic variables, the amount of night sleep was significantly related to the presence of a diagnosis, with children obtaining less sleep at night more likely to receive a DSM-III-R diagnosis. Demographic variables were not significantly related to the presence of a diagnosis (Table 1). In contrast, the logistic regression model for the presence versus absence of a DSM-III-R diagnosis showed no significant effects for age, sex, race/ethnicity, SES, or amount of sleep in a 24-hour period.

Table 1

Multivariate Analyses: The Relationship Between Sleep and Quantitative Estimates of Psychopathology

Child Behavior Checklist (CBCL) Total Behavior Problems

A stepwise multiple regression model including the amount of night sleep indicated that age (B = 3.31, SE = .50, [beta] = .29, t = 6.72, p < .001, R2 = .08) and amount of night sleep (B = -2.02, SE = .51, [beta] = -.17, t = -3.99, p < .001, R2 = .10) were significantly associated with Total Behavior Problems reported on the CBCL (multiple R = .32), with children receiving less sleep exhibiting more behavior problems. Similarly, in the model including total 24-hour sleep rather than night sleep, older children (B = 2.82, SE = .54, [beta] = .24, t = 5.26, p < .001, R2 = .08) and children who obtained less sleep (B = -1.31, SE = .45, [beta] = -.13, t = -2.90, p < .004, R2 = .09) exhibited more behavior problems at a statistically significant level (multiple R = .31).

CBCL Internalizing Behavior Problems

Neither amount of night sleep nor amount of 24-hour sleep was associated with internalizing behavior problems.

CBCL Externalizing Behavior Problems

Age (B = 2.26, SE = .49, [beta] = .30, t = 4.64, p < .001, R2 = .04) and length of night sleep (B = -1.85, SE = .49, [beta] = -.16, t = -3.77, p < .001, R2 = .06) were both significantly associated with the number of externalizing behavior problems the child displayed. For total 24-hour sleep, older children (B = 1.88, SE = .52, [beta] = .17, t = 3.59, p < .001, R2 = .04), children receiving lower amounts of sleep in 24-hour periods (B = -1.25, SE = .43, [beta] = -.13, t = -2.89, p < .004, R2 = .05), and children from lower social classes (B = 3.02, SE = 1.52, [beta] = .09, t = 1.99, p < .05, R2 = .06) were more likely to exhibit externalizing behavior problems.

Threshold Effects

Multivariate analyses established that there were statistically significant overall relationships between sleep and behavior problems, with children who slept less having higher levels of behavior problems as reported on CBCL scales. Comparisons were then made to examine whether there was a threshold or “dose” effect (i.e., what amount of sleep was most strongly associated with behavior problems) for those sleep variables that were significant on the multivariate analyses. This was done by conducting one-way analyses of variance with the amount of sleep categorized into groups consisting of children receiving similar amounts of sleep. For night sleep, the three categories consisted of children with less than 10 hours of sleep per night, 10 to 10.9 hours per night, or 11 or more hours per night. For 24-hour sleep, the categories were less than 10 hours, 10 to 10.9 hours, 11 to 11.99 hours, and 12+ hours of sleep in a 24-hour period. Analyses were conducted separately for 2 to 3 year olds and for 4 to 5 year olds.

In these analyses, significant differences were noted for the 2 to 3 year olds, but significant group differences or threshold effects were not found for the 4 to 5 year olds. Significant effects were found for 2 to 3 year olds for amount of night sleep on both Total Behavior Problems (F = 6.89, 2,247; p < .001) and Externalizing Problems (F = 7.80, 2,247; p < .0005) scales. For both outcome variables, children with the lowest amount of sleep (less than 10 hr) were having significantly more problems than children with the most sleep (11+ hr), but they did not differ from those having a moderate (10 hr) amount of night sleep. For amount of sleep in 24 hours among the 2 to 3 year olds, significant effects were found for Total Behavior Problems (F = 6.05, 3, 244, p < .0005) and for Externalizing Problems (F = 4.59, 3 244, p < .0045). For Total Behavior Problems, the children in the two lowest groups (< 10 and 10.0–10.99 hr) differed from those getting the most sleep (11+ hr). For 24-hour sleep, those with the least sleep (< 10 hr) were having more behavior problems than were those with the most sleep (12+ hr). Viewed in this manner, those 2 to 3 year olds receiving less than 10 hours of sleep per night or per 24-hour period were consistently at greatest risk for behavior problems; children with less than 11 hours of sleep in 24 hours were also at increased risk for externalizing problems. Although 4 to 5 year olds with less sleep were also at risk, a threshold difference could not be identified.


Although developmental characteristics of sleep and sleep problems of young children have been of interest to researchers in the past, little attention has been paid to the relationship between sleep and behavior problems among young children. Prior reports have noted a relationship between difficulties with children falling asleep and daytime behavior problems, but we have been unable to find previously published studies concerning the relationship between the amount of sleep young children obtain and their daytime behavior problems.

This report shows a clear relationship between the amount of sleep children obtain at night, the amount of sleep they obtain during a 24-hour period, and daytime behavior problems. When the outcome measure was the presence versus the absence of a DSM-III-R diagnosis, children who sleep less at night were more likely to exhibit a diagnosis. The relationship between the amount of sleep in 24 hours and diagnosis, however, was not significant. An alternate approach to assessing behavior problems, in which problem behaviors were thought to differ in amount rather than in being present or absent, was more sensitive to the relationships between sleep and daytime behavior. When the outcome was a continuous measure of children's total behavior problems from the Child Behavior Checklist (CBCL), lower amounts of night sleep and 24-hour sleep were both related to increased behavior problems. There was also a significant relationship between lower amounts of sleep and increasing numbers of externalizing behavior problems. Within the preschool age group, externalizing behaviors primarily include “acting out” behaviors, such as hyperac-tivity, oppositional or noncompliant behavior, and aggression. The relationship between sleep and internalizing behavior problems, including problems associated with anxiety and depression, was not significant.

Further analyses established that among children aged 2 to 3 years, very low amounts of sleep (less than 10.0 hr per night or in a 24-hr period) were most closely associated with higher levels of behavior problems. Children aged 2 to 3 years who were receiving less than 11 hours of sleep in a 24-hour period also were more likely to exhibit externalizing behavior problems than were children sleeping 12 or more hours per day. Overall differences were present for all children regardless of age on the multivariate analyses, but a clear threshold effect was not present for 4 to 5 year olds.

As noted in the introduction, there are relatively few studies of the relationship between sleep and behavior problems. The results of this study are consistent with those of Bates et al 6 in finding a relationship between less sleep and more behavior problems in preschoolers and with Weissbluth 7 in that more difficult children slept less. The report extends the findings of Bates et al by examining a larger sample and by extending the approach to diagnosis to include both taxonomic and dimensional approaches to assessing behavior problems. It further extends the Weissbluth report by including a more specific examination of behavior problems rather than behavioral style characterized by concepts of temperament.

Along with the relative paucity of data pertaining to sleep and daytime behavior is the absence of a theory or model to explain such a relationship. The data collected for this study, along with those of Bates et al 6 and Weissbluth,7 are essentially correlational data, examining naturally occurring patterns of sleep and behavior in cross-sectional designs rather than being “true” experiments in which sleep was manipulated experimentally and the effects on behavior were observed. The multivariate models used in data analyses were structured to examine the degree to which the amount of children's sleep was related to the presence of behavior problems independent of the effects of demographic variables including age, sex, race/ethnicity, and socioeconomic status but cannot in themselves establish a causal relationship between sleep and behavior problems. Thus, the relationship between sleep and daytime behavior problems may exist because less sleep causes children to have daytime behavior problems, because daytime behavior problems cause children to sleep less, because a third variable exists (e.g., temperament, parental ability to structure sleep arrangements and daytime behavior) that contributes to both lower amounts of sleep and daytime behavioral difficulties, or because there is some interaction effect that produces a reciprocal influence between sleep and behavior problems. If the amount of sleep has a causal relationship to daytime behavior problems, there may be a direct (e.g., biochemical) effect or a psychological mediator (e.g., increased daytime irritability producing more tantrums, etc.) between sleep and daytime behavior.

This report presents data concerning night sleep and total amount of sleep in a 24-hour period. As expected, these two variables were highly correlated, so it is not surprising that they yielded similar results. Analyses of both variables are included herein because there was a lack of prior information indicating whether they would yield similar or divergent results. In future studies, 24-hour sleep amounts or amount of night sleep may serve as surrogates for one another. Variability in daytime sleep was too small in this sample to allow for independent analyses.

Similarly, measures of psychopathology used herein are not completely independent of one another. The factor structure of the CBCL is such that there is some correlation between Internalizing and Externalizing scales even though they are conceptually quite different; of course, both scales correlate with the Total Behavior Problems scale.8,9 Results again are presented for all three scales because of the lack of models regarding sleep and daytime behavior to guide the choice of variables. It is noteworthy that the relationship between sleep and externalizing problems was significant in this age group, whereas the relationship was not significant with internalizing problems. This may be related to the relatively small number of internalizing problems that are present in the preschooler age group compared with the number of externalizing problems, such as the symptoms associated with oppositional defiant disorder (ODD) (tantrums, defiance, disobedience, etc.). Although many of these symptoms are common in preschoolers, high levels of such problems can be reliably diagnosed as ODD' as the results of this study indicate, moderate-to-high levels of such symptoms may be related to the amount of sleep the child obtains. Because the most common problems in this age group involve externalizing problems, it is not surprising that Total Behavior Problems scale scores parallel the results for Externalizing Behavior Problems on the CBCL.

Two limitations of this study are worth noting. First, data gathered on sleep patterns involved parental report. Independent assessment of sleep patterns (e.g., night waking) and length of sleep may yield different results, and such data would be of particular interest in future studies. Second, reliability of sleep reports could not be directly assessed in this study because of the constraints resulting from the collection of other data for the project. Although this is a limitation, if the sleep data were too unstable, it is highly unlikely that interpretable, statistically significant results would have been obtained because highly unreliable results contain so much error that the chances of obtaining significance are reduced.

Overall, these results provide an interesting early look at the relationship between sleep and the behavior problems of preschool children. It is unlikely that “true” experiments will be conducted in which amounts of sleep are systematically varied, but the relationship between sleep and behavior merits further investigation.


1. Adair RH, Bauchner H: Sleep problems in childhood. Curr Probl Pediatr 23:147-170, 1993 Bibliographic Links [Context Link]

2. Weissbluth M: Naps in children: 6 months-7 years. Sleep 18:82–87, 1995 [Context Link]

3. Richman N: Sleep problems in young children. Arch Dis Child 56:491–493, 1981 Bibliographic Links [Context Link]

4. Jenkins S, Bax M, Hart H: Behaviour problems in pre-school children. J Child Psychol Psychiatry 21:5–19, 1980 [Context Link]

5. Richman N: Prevalence and treatment of sleep problems in young children, in Stevenson JE (ed.): Recent Research in Developmental Psychopathology. Oxford, England, Pergamon Press, 1981, pp 45–63 [Context Link]

6. Bates JE, Viken RJ, Alexander D, Beyers J, Stockton L: Sleep and adjustment in preschool children. Presented at the meeting of the Society for Research in Child Development. Indianapolis, IN, 1995 [Context Link]

7. Weissbluth M: Sleep duration, temperament, and Conners' ratings of three-year-old children. J Dev Behav Pediatr 5:120–123, 1984 Ovid Full Text Bibliographic Links [Context Link]

8. Achenbach TM: Manual for the Child Behavior checklist/4-18 and 1991 Profile. Burlington, VT, University of Vermont, Department of 18. Psychiatry, 1991 [Context Link]

9. Achenbach TM: Manual for the Child Behavior checklist/2–3 and 1992 Profile. Burlington, VT, University of Vermont, Department of 19. Psychiatry, 1992 [Context Link]

10. Lavigne JV, Gibbons RD, Christoffel KK, et al: Prevalence rates and correlates of psychiatric disorders among preschool children. J Am Acad Child Adolesc Psychiatry 35:204–214, 1996 Ovid Full Text Bibliographic Links [Context Link]

11. Lavigne JV, Arend R, Rosenbaum D, et al: Interrater reliability of the DSM-III-R with preschool children. J Abnorm Child Psychol 22:679–690, 1994 Bibliographic Links [Context Link]

12. Lavigne JV, Arend R, Rosenbaum D, Binns HJ, Christoffel KK, Gibbons RD: Psychiatric disorders with onset in the preschool years: I. Stability of diagnoses. J Am Acad Child Adolesc Psychiatry 37:1246–1254, 1998 Ovid Full Text Bibliographic Links [Context Link]

13. Lavigne JV, Arend R, Rosenbaum D, Binns HJ, Christoffel KK, Gibbons RD: Psychiatric disorders with onset in the preschool years: II. Correlates and predictors of stable case status. J Am Acad Child Adolesc Psychiatry 37:1255–1261, 1998 Ovid Full Text Bibliographic Links [Context Link]

14. Lavigne JV, Binns HJ, Christoffel KK, et al: Behavioral and emotional problems among preschool children in pediatric primary care: Prevalence and pediatricians' recognition. Pediatrics 91:649–655, 1993 Bibliographic Links [Context Link]

15. Christoffel KK, Binns HJ, Stockman JA, et al: Practice-based research: Opportunities and obstacles. Pediatrics 82:399–406, 1988 [Context Link]

16. Hollingshead AB: Four-Factor Index of Social Status. New Haven, CT, Yale University Press, 1975 [Context Link]

17. Jones FH: The Rochester Adaptive Behavior Inventory: A parallel series of instruments for assessing social competence during early and middle childhood and adolescence, in Strauss KS, Babigian HM (eds): The Origins and Course of Psychopathology, New York, NY, Plenum, 1977, pp 249–281 [Context Link]

18. Seifer R, Sameroff AJ: Multiple determinants of risk and invulnerability, in Anthony EJ, Cohler BJ (eds): The Invulnerable Child. New York, NY, Guilford, 1981, pp 51–69 [Context Link]

19. Forehand RL, McMahon RJ: Helping the Noncompliant Child. New York, NY, Guilford, 1981 [Context Link]

20. Bayley N: Manual for the Bayley Scales of Infant Development. New York, NY, The Psychological Corporation, 1969 [Context Link]

21. McCarthy D: Manual for the McCarthy Scales of Children's Abilities. New York, NY, The Psychological Corporation, 1972 [Context Link]

22. Weissman MM, Fendrich M, Warner V, Wickramaratne P: Incidence of psychiatric disorder in offspring at high and low risk for depression. J Am Acad Child Adolesc Psychiatry 32:640–648, 1992 [Context Link]

23. American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 3rd ed, revised. Washington, DC, American Psychiatric Association, 1987 [Context Link]

Index terms: sleep; behavior problems; preschool children



There are currently no comments.
Latest Ebooks