Placebo tablets dosages: 5 mg, 10 mg, 25 mg, 40 mg.

The Aberrant Behavior Checklist (ABC) is a 58-item questionnaire with 5 subscales derived by factor analysis: Irritability, Social Withdrawal, Stereotypy, Hyperactivity, and Inappropriate. It has been extensively used in psychopharmacological studies of autism and assesses many symptoms that are either central to autism (Social Withdrawal, Stereotypy, and Inappropriate Speech) or frequently a target of treatment (Irritability). Each item of the 58-item scale is scored on a 4-point scale (0=never a problem to 3=severe problem). The interpretation of the tool and its sub-scales is that a greater number of items, indicates greater severity. The range of scores per subscale are: Social Withdrawal/Lethargy 0-48; Stereotypy 0-21; Irritability 0-45; Hyperactivity 0-48; Inappropriate Speech 0-12. Parent ratings occur every 2 weeks during the study. Estimates are adjusted for baseline score, study stratum, and site, which were set at their sample means.

The Social Responsiveness Scale (SRS) is completed by the parent in order to assess whether additional improvements in social functioning occur with atomoxetine, as observed in our pilot study. This 65-item questionnaire will be completed at baseline and at the end of 8 weeks. The SRS is a standardized measure of the core symptoms of autism. Each item is scored on a 4-point Likert scale. The score of each item is summed to create a total score. Total score results as follows: 0-62: within normal limits; 63-79 mild range of impairment; 80-108: moderate range of impairment; 109-149: severe range of impairment. Estimates are adjusted for baseline score, study stratum, and site, which were set at their sample means.

The Vineland Adaptive Behavior Scales, Second Edition (VABS) is used to assess adaptive functioning in four domains: Communication, Daily Living Skills, Socialization, and Motor Skills. This is a well-standardized open-ended interview used to assess the overall functioning of children and adults. This measure is especially important for subjects with PDDs given that their intellectual level is not always comparable to their adaptive functioning. The Vineland Maladaptive Behavior subscales will be included with these measures as these have been shown to be responsive to drug effects in other clinical trials in this population. The VABS will be done at baseline and at the end of 8 weeks. The composite score represents a standard score (mean = 100 and standard deviation of 15; range = 20-160) on which higher scores indicate a higher level of adaptive functioning. Estimates are adjusted for baseline score, study stratum, and site, which were set at their sample means.

Quality of life is assessed with the Pediatric Quality of Life Inventory (PedsQL 4.0). This instrument is well-validated and widely used for measuring health-related quality of life in children and adolescents. It also appears to be a valid instrument for use with children with psychiatric disorders. The Generic Core scales include 23 items. The health related and family functioning scores range from 0 to 100, with higher scores indicating better quality of life. The Family Impact module will be included to assess any change in family functioning. This will be completed at baseline and at the end of 8 weeks. Estimates are adjusted for baseline score, study stratum, and site, which were set at their sample means.

Since the ABC does not have items which directly assess anxiety, the Pediatric Anxiety Rating Scale (PARS) is administered at week 8 during the study as an exploratory measure. The PARS is a clinician-rated instrument that assesses anxiety symptoms that are commonly associated with social anxiety, separation anxiety, and generalized anxiety disorders. Scaled score ranges form 0-25 with higher scores indicating more severe anxiety symptoms. Estimates are adjusted for baseline score, study stratum, and site, which were set at their sample means.

The Clinical Global Impressions Global Improvement (CGI-I) is designed to take into account all factors to arrive at an assessment of response to treatment. The CGI-I scale ranges from 1 to 7 (1=very much improved; 2= much improved; 3=minimally improved; 4=no change; 5=minimally worse; 6=much worse; 7=very much worse), with lower scores indicating improvement (1=very much improved and 2=much improved). Participants with a CGI-I score of 1 or 2 were classified as improved. Odds of improvement at 8 weeks were estimated using a repeated measures logistic regression model adjusting for baseline severity, study stratum, and site. The CGI-I was administered biweekly during the study. The CGI was focused on the target symptoms of inattention, hyperactivity, and impulsivity.

The ADHD Rating Scale (ADHD-RS) is an 18-item scale directly derived from DSM-IV criteria for Attention Deficit Hyperactivity Disorder with established reliability, validity and sensitivity to change. The ADHD-RS-IV is investigator-administered biweekly during the 8-week double-blind, placebo-controlled phase of the study. The scale consists of 2 subscales: inattention (9 items) and hyperactivity-impulsivity (9 items). If 3 or more items are skipped, the clinician should use extreme caution in interpreting the scale. Results from this rating scale alone should not be used to make a diagnosis. The total score can range form 0 to 54, with a higher score indicating greater severity. Change will be determined from the start of the open-label trial to 8 weeks post-start.

The ADHD Rating Scale (ADHD-RS) is an 18-item scale directly derived from DSM-IV criteria for Attention Deficit Hyperactivity Disorder with established reliability, validity and sensitivity to change. The ADHD-RS-IV is investigator-administered biweekly during the 8-week open-label phase of the study. The scale consists of 2 subscales: inattention (9 items) and hyperactivity-impulsivity (9 items). If 3 or more items are skipped, the clinician should use extreme caution in interpreting the scale. Results from this rating scale alone should not be used to make a diagnosis. The score fro each subscale ranges from 0-27, with a higher score indicating greater severity. Change will be determined from the start of the open-label trial to 8 weeks post-start.

The Aberrant Behavior Checklist (ABC) is a 58-item questionnaire with 5 subscales derived by factor analysis: Irritability, Social Withdrawal, Stereotypy, Hyperactivity, and Inappropriate. It has been extensively used in psychopharmacological studies of autism and assesses many symptoms that are either central to autism (Social Withdrawal, Stereotypy, and Inappropriate Speech) or frequently a target of treatment Irritability). Each item of the 58-item scale is scored on a 4-point scale (0=never a problem to 3=severe problem). The interpretation of the tool and its sub-scales is that a greater number of items, indicates greater severity. The range of scores per subscale are: Social Withdrawal/Lethargy 0-48; Stereotypy 0-21; Irritability 0-45; Hyperactivity 0-48; Inappropriate Speech 0-12. Parent ratings occur every 2 weeks during the study. Change will be determined from the start of the open-label trial to 8 weeks post-start.

The Social Responsiveness Scale (SRS) is completed by the parent in order to assess whether additional improvements in social functioning occur with atomoxetine, as observed in our pilot study. This 65-item questionnaire will be completed at baseline and at the end of 8 weeks. The SRS is a standardized measure of the core symptoms of autism. Each item is scored on a 4-point Likert scale. The score of each item is summed to create a total score. Total score results as follows: 0-62: within normal limits; 63-79 mild range of impairment; 80-108: moderate range of impairment; 109-149: severe range of impairment. This 65-item questionnaire will be completed at the start of and at the end of 8 weeks of the open-label trial.

The Vineland Adaptive Behavior Scales, Second Edition (VABS) is used to assess adaptive functioning in four domains: Communication, Daily Living Skills, Socialization, and Motor Skills. This is a well-standardized open-ended interview used to assess the overall functioning of children and adults. This measure is especially important for subjects with PDDs given that their intellectual level is not always comparable to their adaptive functioning. The Vineland Maladaptive Behavior subscales will be included with these measures as these have been shown to be responsive to drug effects in other clinical trials in this population. The composite score represents a standard score (mean = 100 and standard deviation of 15; range = 20-160) on which higher scores indicate a higher level of adaptive functioning. Change will be determined from the start of the open-label phase to 8 weeks

Quality of life is assessed with the Pediatric Quality of Life Inventory (PedsQL 4.0). This instrument is well-validated and widely used for measuring health-related quality of life in children and adolescents. It also appears to be a valid instrument for use with children with psychiatric disorders. The Generic Core scales include 23 items. The health related and family functioning scores range from 0 to 100, with higher scores indicating better quality of life. The Family Impact module will be included to assess any change in family functioning. This will be completed at the start of the open-label trial and at the end of 8 weeks. Change will be determined from the start of the open-label trial to 8 weeks post-start.

Since the ABC does not have items which directly assess anxiety, the Pediatric Anxiety Rating Scale (PARS) is administered at week 8 during the study as an exploratory measure. The PARS is a clinician-rated instrument that assesses anxiety symptoms that are commonly associated with social anxiety, separation anxiety, and generalized anxiety disorders. Scaled score ranges form 0-25 with higher scores indicating more severe anxiety symptoms.Change will be determined from the start of the open-label trial to 8 weeks post-start.

Participants will receive blinded, matched placebo for 8 weeks. Dosage can be increased over the first 4 weeks of study participation and will then be held constant for the remainder of the 8-week trial.

Placebo: Placebo tablets dosages: 5 mg, 10 mg, 25 mg, 40 mg.

Total of all reporting groups

All children attended school. Children not attending public school attended private, specialized, or home school, or a specialized classroom with type of school unspecified.

Tanner stage (1-5) was used to characterize physical development in children, adolescents, and to stratify randomization.The stage was based on external primary and secondary sex characteristics, such as the size of the breasts, genitalia, and development of pubic hair. Tanner stage increases as the child develops. Randomization lists for each of the four strata defined by pubertal status (prepubertal [Tanner Stage I or 2] vs. postpubertal [Tanner Stage 3-5]) and gender (male vs. female) were utilized.

IQ was estimated from the Mullen Scale for 1 child assigned to atomoxetine and the Stanford Binet test for all other children.

IQ scores were not obtained for 2 children assigned to atomoxetine and 1 child assigned to placebo.

The ADHD Rating Scale (ADHD-RS) is an 18-item scale directly derived from DSM-IV criteria for Attention Deficit Hyperactivity Disorder with established reliability, validity and sensitivity to change. The scale consists of 2 subscales: inattention (9 items) and hyperactivity-impulsivity (9 items). The total score can range from 0 to 54, with a higher score indicating greater severity.

The Vineland Adaptive Behavior Scales, Second Edition is a measure of adaptive behavior in children, adolescents and adults. The composite score has a mean of 100 and a standard deviation of 15 (range = 20 to 160). Higher scores suggest a higher level of adaptive functioning. A rise in standard scores from Baseline to the Final Visit indicates improvement.

Scores range from 0-100, with higher scores indicating better quality of life.

Scores range from 0-100, with higher scores representing better family functioning.

Scaled score ranges form 0-25 with higher scores indicating more severe anxiety symptoms.

Participants will receive blinded, matched placebo for 8 weeks. Dosage can be increased over the first 4 weeks of study participation and will then be held constant for the remainder of the 8-week trial.

Placebo: Placebo tablets dosages: 5 mg, 10 mg, 25 mg, 40 mg.

The ADHD Rating Scale (ADHD-RS) is an 18-item scale directly derived from DSM-IV criteria for Attention Deficit Hyperactivity Disorder with established reliability, validity and sensitivity to change. The ADHD-RS-IV is investigator-administered biweekly during the 8-week double-blind, placebo-controlled phase of the study. The scale consists of 2 subscales: inattention (9 items) and hyperactivity-impulsivity (9 items). If 3 or more items are skipped, the clinician should use extreme caution in interpreting the scale. Results from this rating scale alone should not be used to make a diagnosis. The score for each subscale ranges from 0-27 with a higher score indicating greater severity. Estimates are adjusted for baseline score, study stratum, and site, which were set at their sample means.

Two participants from the Atomoxetine arm, and one participant from the Placebo arm did not have post-baseline measures.

The Aberrant Behavior Checklist (ABC) is a 58-item questionnaire with 5 subscales derived by factor analysis: Irritability, Social Withdrawal, Stereotypy, Hyperactivity, and Inappropriate. It has been extensively used in psychopharmacological studies of autism and assesses many symptoms that are either central to autism (Social Withdrawal, Stereotypy, and Inappropriate Speech) or frequently a target of treatment (Irritability). Each item of the 58-item scale is scored on a 4-point scale (0=never a problem to 3=severe problem). The interpretation of the tool and its sub-scales is that a greater number of items, indicates greater severity. The range of scores per subscale are: Social Withdrawal/Lethargy 0-48; Stereotypy 0-21; Irritability 0-45; Hyperactivity 0-48; Inappropriate Speech 0-12. Parent ratings occur every 2 weeks during the study. Estimates are adjusted for baseline score, study stratum, and site, which were set at their sample means.

Two participants from the Atomoxetine arm, and one participant from the Placebo arm did not have post-baseline measures.

The Social Responsiveness Scale (SRS) is completed by the parent in order to assess whether additional improvements in social functioning occur with atomoxetine, as observed in our pilot study. This 65-item questionnaire will be completed at baseline and at the end of 8 weeks. The SRS is a standardized measure of the core symptoms of autism. Each item is scored on a 4-point Likert scale. The score of each item is summed to create a total score. Total score results as follows: 0-62: within normal limits; 63-79 mild range of impairment; 80-108: moderate range of impairment; 109-149: severe range of impairment. Estimates are adjusted for baseline score, study stratum, and site, which were set at their sample means.

Four participants from the Atomoxetine arm, and two participants from the Placebo arm did not have post-baseline measures.

The Vineland Adaptive Behavior Scales, Second Edition (VABS) is used to assess adaptive functioning in four domains: Communication, Daily Living Skills, Socialization, and Motor Skills. This is a well-standardized open-ended interview used to assess the overall functioning of children and adults. This measure is especially important for subjects with PDDs given that their intellectual level is not always comparable to their adaptive functioning. The Vineland Maladaptive Behavior subscales will be included with these measures as these have been shown to be responsive to drug effects in other clinical trials in this population. The VABS will be done at baseline and at the end of 8 weeks. The composite score represents a standard score (mean = 100 and standard deviation of 15; range = 20-160) on which higher scores indicate a higher level of adaptive functioning. Estimates are adjusted for baseline score, study stratum, and site, which were set at their sample means.

Four participants from the Atomoxetine arm, and two participants from the Placebo arm did not have post-baseline measures.

Quality of life is assessed with the Pediatric Quality of Life Inventory (PedsQL 4.0). This instrument is well-validated and widely used for measuring health-related quality of life in children and adolescents. It also appears to be a valid instrument for use with children with psychiatric disorders. The Generic Core scales include 23 items. The health related and family functioning scores range from 0 to 100, with higher scores indicating better quality of life. The Family Impact module will be included to assess any change in family functioning. This will be completed at baseline and at the end of 8 weeks. Estimates are adjusted for baseline score, study stratum, and site, which were set at their sample means.

Four participants from the Atomoxetine arm, and two participants from the Placebo arm did not have post-baseline measures.

Since the ABC does not have items which directly assess anxiety, the Pediatric Anxiety Rating Scale (PARS) is administered at week 8 during the study as an exploratory measure. The PARS is a clinician-rated instrument that assesses anxiety symptoms that are commonly associated with social anxiety, separation anxiety, and generalized anxiety disorders. Scaled score ranges form 0-25 with higher scores indicating more severe anxiety symptoms. Estimates are adjusted for baseline score, study stratum, and site, which were set at their sample means.

Four participants from the Atomoxetine arm, and two participants from the Placebo arm did not have post-baseline measures.

The Clinical Global Impressions Global Improvement (CGI-I) is designed to take into account all factors to arrive at an assessment of response to treatment. The CGI-I scale ranges from 1 to 7 (1=very much improved; 2= much improved; 3=minimally improved; 4=no change; 5=minimally worse; 6=much worse; 7=very much worse), with lower scores indicating improvement (1=very much improved and 2=much improved). Participants with a CGI-I score of 1 or 2 were classified as improved. Odds of improvement at 8 weeks were estimated using a repeated measures logistic regression model adjusting for baseline severity, study stratum, and site. The CGI-I was administered biweekly during the study. The CGI was focused on the target symptoms of inattention, hyperactivity, and impulsivity.

Two participants from the Atomoxetine arm, and one participant from the Placebo arm did not have post-baseline measures.

The ADHD Rating Scale (ADHD-RS) is an 18-item scale directly derived from DSM-IV criteria for Attention Deficit Hyperactivity Disorder with established reliability, validity and sensitivity to change. The ADHD-RS-IV is investigator-administered biweekly during the 8-week double-blind, placebo-controlled phase of the study. The scale consists of 2 subscales: inattention (9 items) and hyperactivity-impulsivity (9 items). If 3 or more items are skipped, the clinician should use extreme caution in interpreting the scale. Results from this rating scale alone should not be used to make a diagnosis. The total score can range form 0 to 54, with a higher score indicating greater severity. Change will be determined from the start of the open-label trial to 8 weeks post-start.

Participants randomized to placebo who are not responders at the end of 8 weeks will be treated with atomoxetine for 8 weeks during an open-label trial

The ADHD Rating Scale (ADHD-RS) is an 18-item scale directly derived from DSM-IV criteria for Attention Deficit Hyperactivity Disorder with established reliability, validity and sensitivity to change. The ADHD-RS-IV is investigator-administered biweekly during the 8-week open-label phase of the study. The scale consists of 2 subscales: inattention (9 items) and hyperactivity-impulsivity (9 items). If 3 or more items are skipped, the clinician should use extreme caution in interpreting the scale. Results from this rating scale alone should not be used to make a diagnosis. The score fro each subscale ranges from 0-27, with a higher score indicating greater severity. Change will be determined from the start of the open-label trial to 8 weeks post-start.

Participants randomized to placebo who are not responders at the end of 8 weeks will be treated with atomoxetine for 8 weeks during an open-label trial

The Aberrant Behavior Checklist (ABC) is a 58-item questionnaire with 5 subscales derived by factor analysis: Irritability, Social Withdrawal, Stereotypy, Hyperactivity, and Inappropriate. It has been extensively used in psychopharmacological studies of autism and assesses many symptoms that are either central to autism (Social Withdrawal, Stereotypy, and Inappropriate Speech) or frequently a target of treatment Irritability). Each item of the 58-item scale is scored on a 4-point scale (0=never a problem to 3=severe problem). The interpretation of the tool and its sub-scales is that a greater number of items, indicates greater severity. The range of scores per subscale are: Social Withdrawal/Lethargy 0-48; Stereotypy 0-21; Irritability 0-45; Hyperactivity 0-48; Inappropriate Speech 0-12. Parent ratings occur every 2 weeks during the study. Change will be determined from the start of the open-label trial to 8 weeks post-start.

Placebo-treated subjects that don't respond to placebo will be offered an 8-week open-label trial of atomoxetine, and the open-label extension will be a continuation phase for those subjects that respond to 8 weeks of atomoxetine during the double-blind phase.

The Social Responsiveness Scale (SRS) is completed by the parent in order to assess whether additional improvements in social functioning occur with atomoxetine, as observed in our pilot study. This 65-item questionnaire will be completed at baseline and at the end of 8 weeks. The SRS is a standardized measure of the core symptoms of autism. Each item is scored on a 4-point Likert scale. The score of each item is summed to create a total score. Total score results as follows: 0-62: within normal limits; 63-79 mild range of impairment; 80-108: moderate range of impairment; 109-149: severe range of impairment. This 65-item questionnaire will be completed at the start of and at the end of 8 weeks of the open-label trial.

Participants randomized to placebo who are not responders at the end of 8 weeks will be treated with atomoxetine for 8 weeks during an open-label trial

The Vineland Adaptive Behavior Scales, Second Edition (VABS) is used to assess adaptive functioning in four domains: Communication, Daily Living Skills, Socialization, and Motor Skills. This is a well-standardized open-ended interview used to assess the overall functioning of children and adults. This measure is especially important for subjects with PDDs given that their intellectual level is not always comparable to their adaptive functioning. The Vineland Maladaptive Behavior subscales will be included with these measures as these have been shown to be responsive to drug effects in other clinical trials in this population. The composite score represents a standard score (mean = 100 and standard deviation of 15; range = 20-160) on which higher scores indicate a higher level of adaptive functioning. Change will be determined from the start of the open-label phase to 8 weeks

Participants randomized to placebo who are not responders at the end of 8 weeks will be treated with atomoxetine for 8 weeks during an open-label trial

Quality of life is assessed with the Pediatric Quality of Life Inventory (PedsQL 4.0). This instrument is well-validated and widely used for measuring health-related quality of life in children and adolescents. It also appears to be a valid instrument for use with children with psychiatric disorders. The Generic Core scales include 23 items. The health related and family functioning scores range from 0 to 100, with higher scores indicating better quality of life. The Family Impact module will be included to assess any change in family functioning. This will be completed at the start of the open-label trial and at the end of 8 weeks. Change will be determined from the start of the open-label trial to 8 weeks post-start.

Participants randomized to placebo who are not responders at the end of 8 weeks will be treated with atomoxetine for 8 weeks during an open-label trial

Since the ABC does not have items which directly assess anxiety, the Pediatric Anxiety Rating Scale (PARS) is administered at week 8 during the study as an exploratory measure. The PARS is a clinician-rated instrument that assesses anxiety symptoms that are commonly associated with social anxiety, separation anxiety, and generalized anxiety disorders. Scaled score ranges form 0-25 with higher scores indicating more severe anxiety symptoms.Change will be determined from the start of the open-label trial to 8 weeks post-start.

Participants randomized to placebo who are not responders at the end of 8 weeks will be treated with atomoxetine for 8 weeks during an open-label trial

Participants will receive blinded, matched placebo for 8 weeks. Dosage can be increased over the first 4 weeks of study participation and will then be held constant for the remainder of the 8-week trial.

Placebo: Placebo tablets dosages: 5 mg, 10 mg, 25 mg, 40 mg.

Placebo-treated subjects from the randomized trial that don't respond to placebo will be offered an 8-week open-label trial of atomoxetine