Continuity of behavioral and emotional problems from pre-school years to pre-adolescence in a developing country Luciana Anselmi, 1 Fernando C. Barros, 2 Maycoln L.M. Teodoro, 3 Ce ´sar A. Piccinini, 4 Ana Maria B. Menezes, 2 Cora L. Araujo, 2 and Luis A. Rohde 5 1 Post-Graduate Program in Psychiatry, Federal University of Rio Grande do Sul, and Post-Graduate Program in Epidemiology, Federal University of Pelotas, Brazil; 2 Post-Graduate Program in Epidemiology, Federal University of Pelotas, Brazil; 3 Post-Graduate Program in Psychology, Unisinos, Brazil; 4 Post-Graduate Program in Psychology, Federal University of Rio Grande do Sul, Brazil; 5 Child and Adolescent Psychiatric Division, Federal University of Rio Grande do Sul, Brazil Background: All previous longitudinal community studies assessing the continuity of child behavioral/ emotional problems were conducted in developed countries. Method: Six hundred and one children randomly selected from a Brazilian birth cohort were evaluated for behavioral/emotional problems through mother interview at 4 and 12 years with the same standard procedure – Child Behavior Checklist (CBCL). Results: CBCL Total Problem score presented a medium stability (r ¼ .42) with externalizing problems showing higher stability and more homotypic continuity than internalizing problems. Of the children presenting deviant scores at the age of 4, only 31% remained deviant at the age of 12 (p < .001). A deviant CBCL Total Problem score at 12 years old was predicted by Rule-Breaking Behavior [OR ¼ 7.46, 95% CI 2.76–20.19] and Social Problems [OR ¼ 3.56, 95% CI 1.36–9.30] scores at 4 years of age. Either Rule-Breaking or Aggressive Behavior – externalizing syndromes – were part of the predictors for the three broad-band CBCL scores and six out of the eight CBCL syn- dromes. Conclusions: Behavioral/emotional problems in preschool children persist moderately up to pre-adolescence in a community sample. Externalizing problems at the age of 4 comprise the devel- opmental history of most behavioral/emotional problems at pre-adolescence. Our findings concur with findings from developed countries and are quite similar for continuity, stability and predictability. Keywords: Behavior problems, continuity, longitudinal studies, Third World children, child develop- ment, externalizing disorders. Typical child development is characterized by sta- bility, but it also involves behavioral changes and reorganizations (Piaget, 1970). Similarly, the atypical development may present continuities and disconti- nuities throughout the life cycle (Rutter, Kim-Cohen, & Maughan, 2006). There is evidence that symptom patterns change as individuals get older – including diagnostic categories – what is also known as heterotypic continuity or sequential comorbidity (Angold, Costello, & Erkanli, 1999). This may be caused by changes in expressions of biological development, cognitive level, social status, duration and type of experiences (Rutter, 1988). Evidence of continuity between child and adult psychopathology has been revealed by studies using clinical samples and retrospective methods. It has also been found in prospective studies and investi- gations using random samples from birth cohorts (Colman & Jones, 2004; Feehan, McGee, & Williams, 1993). Prospective studies have demonstrated that the origin of many adult mental disorders can be found in behavioral characteristics present since the first years of life (Caspi, Moffitt, Newman, & Silva, 1996; Maughan & Kim-Cohen, 2005) and that externalizing problems are the strongest predictors of disorders in adulthood (Hofstra, van der Ende, & Verhulst, 2002). Prospective studies on psychiatric disorders in children and adolescents also have shown that between 23% and 61% of children who had received a positive diagnosis at any follow-up point were diagnosed as having psychiatric disorders, although not necessarily the same one, in the next follow-up (Costello, Mustillo, Erkanli, Keeler, & Angold, 2003). Medium stability of behavioral/emotional problems in children and adolescents assessed with the Child Behavior Checklist (CBCL; Achenbach, 1991) has also been found in longitudinal community studies. The correlation coefficient between the Total Prob- lems scores at different time points ranged between .58 (McConaughy, Stanger, & Achenbach, 1992) and .56 (Verhulst & van der Ende, 1995) in follow-ups with time intervals of 3 and 6 years, respectively. Stability was lower when the time intervals between the assessments were longer. For example, Verhulst and van der Ende (1995) found coefficients of .40 for girls and .42 for boys with initial ages from 4 to 6 years old in an 8-year time interval assessment. Finally, externalizing problems had higher stability than the internalizing problems. It is important to note that all previous longitudi- nal studies assessing the continuity of behavioral/ Conflict of interest statement: No conflicts declared. Journal of Child Psychology and Psychiatry **:* (2007), pp **–** doi:10.1111/j.1469-7610.2007.01865.x Ó 2007 The Authors Journal compilation Ó 2007 Association for Child and Adolescent Mental Health. Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA CHILD PSYCHOLOGY AND PSYCHIATRY THE JOURNAL OF