We conducted an observational, prospective, noncomparative multicentre study with a follow-up period of four months of duration between April and September of 2010. The study was approved by the ethics committee of the Hospital Clínic i Provincial de Barcelona (2009/5347). All parents/guardians agreed to participate in the study by signing a written informed consent, which was also obtained from children older than 12 years.

The intensity of the ADHD symptoms of the child or adolescent was assessed by means of the Abbreviated Conners Scale (ACS), which was completed by the parent or guardian and referred to the past four weeks. The ACS consists of 10 items with four possible answers (not at all, 0; just a little, 1; pretty much, 2; very much, 3). The sum of the scores of all 10 items yields the total score on the scale (0–30 points).

Family stress or strain was measured by means of the FSI in reference to the past four weeks, which was completed by the parent/guardian. The FSI consists of six items that assess two dimensions, the “emotional” dimension (items 1 and 3: 0–8 points) and the limitations to the family or “restrictiveness” dimension (items 2, 4, 5 and 6: 0–16 points). Each item is scored on a five-point scale (never, 0; almost never, 1; sometimes, 2; almost always, 3; always, 4) with the total score ranging between 0 and 24 points. Higher scores indicate greater impairment in family functioning. The emotional items assess the level of affective or emotional stress, and the restrictiveness dimension the limitations in the family's social activities.

The child psychiatrists and neuropaediatricians that participated as researchers selected the families, which were included by consecutive random sampling. The children could be of any race and sex, were between 6 and 17 years of age, and had a diagnosis of ADHD based on the criteria defined by the DSM-IV-TR.7 Inclusion in the study required not having received any previous treatment for ADHD. The inclusion of cases with depression, anxiety, tics, oppositional-defiant disorder (ODD), conduct disorder, learning disorder or disruptive behaviour disorders was allowed. Patients were given the standard psychopharmacological treatment.

We excluded families with institutionalised children, with more than one child with an ADHD diagnosis, or with at least one child with intellectual disability (IQ<70), psychosis, schizophrenia, bipolar disorder, autism or pervasive developmental disorder.

We performed a descriptive analysis of frequencies and percentages for qualitative variables, and mean±standard deviation, minimum, maximum and 95% confidence interval (95% CI) for quantitative variables. We performed comparisons using the χ2 for comparing proportions or Wilcoxon's test. We compared the mean scores in the ACS and the FSI questionnaire using Student's t-test or one-factor ANOVA with the Bonferroni or Games–Howell multiple-comparison correction when the factor had more than two categories. To analyse the concordance between the mean variations in the intensity of ADHD symptoms (ACS) and in the quality of life of families (FSI), we analysed the intraclass correlation coefficient (ICC) of individual measurements and of their means. We used multiple linear regression analysis to control for the effect of sociodemographic, anthropometric and clinical variables on the variability of symptom intensity and family stress. We set the level of statistical significance at .05. The statistical analysis was performed with the SPSS 14.0 software.

Table 1 summarises the characteristics of the family members that completed the study. The mean age of participants was 41.5 years (95% CI, 40.7–42.2) and the median was 40.8 years (21–70 years).

Table 2 summarises the sociodemographic and clinical data of the children with ADHD. Their mean age was 10.4 years (95% CI, 10.1–10.6), and the median age was 10.1 (6–17 years). We did not observe any differences in age between males and females. Two hundred and thirty-six children (57.1%) were 6–10 years old, and 177 (42.9%) were adolescents aged 11–17 years.

Psychiatric comorbidities were present in 28.7% (123) of the children, with 80.5% (99) having only one comorbidity, 18.7% (23) two comorbidities, and 0.8% (1) three comorbidities. Fig. 1 shows the proportions for each psychiatric comorbidity observed.

Figure 1.

Type of psychiatric comorbidity in the child or adolescent during the period under study for the 123 cases that had a positive mental health history.

(0.09MB).

All children received psychoeducational therapy, and 97.9% (420) were treated with modified-release methylphenidate. Another three patients (0.7%) received methylphenidate combined with risperidone, and yet one other patient (0.2) methylphenidate in combination with tiapride. Five patients (1.2%) received atomoxetine as initial treatment for ADHD. The total daily doses ranged between 10 and 60mg of methylphenidate or 40 and 60mg of atomoxetine.

Forty-one families experienced family events that may have impacted their functioning and quality of life: parental separation (5), death in the family (6), parental disease or accident (8), parent becoming unemployed (7), aggression of family members by affected children (3) and other events (12).

Table 3 summarises the scores of the FSI questionnaire during the followup of the families.

We observed statistically significant decreases in the total score and in every dimension of the FSI questionnaire between the three evaluation times (P<.0001).

In 353 families (87.6%), the FSI score improved between the baseline and the 2-month visits, and in 29 families (7.2%) it did not change. We observed that the total score of the FSI became worse in 21 families (5.2%).

Between the first and the four-month visits, the FSI score improved in 376 families (92.2%), it did not change in 18 families (4.4%), and it became worse in 14 families (3.4%).

We found no statistically significant differences in FSI scores based on the sex of the patient.

We found statistically significant differences in FSI scores between the families of children with psychiatric comorbidities and the families of children without psychiatric comorbidities (P<.0001), and the total scores of families of children with psychiatric comorbidities were significantly higher at every follow-up visit (Fig. 2). In the baseline visit, the average difference was of 3.2 points (95% CI, 1.8–4.6); at two months, the scores differed by 2.4 points (95% CI, 1.1–3.4) and at four months by 2.5 points (95% CI, 1.4–3.7). The differences persisted through the entire follow-up period. The differences continued to be significant when we performed the multivariate analysis to assess several possible risk factors in combination (P<.0001), in which the presence of psychiatric comorbidities was associated with a 2.7-point increase in the total score (95% CI, 1.5–3.9).

Figure 2.

Evolution of the total score in the FSI questionnaire in relation to the presence of psychiatric comorbidity in the specific child or adolescent.

(0.06MB).

We found no interaction between sex and the presence or absence of psychiatric comorbidities. No other factor showed a significant association with the baseline score on the FSI questionnaire.

Table 3 presents the results of the evolution of the scores.

We observed statistically significant improvements in the scores of the ACS in every consecutive follow-up visit (P<.0001). The score in the ACS improved in 377 patients (92.4%), and got worse between the baseline and the 2-month visit in 17 patients (4.2%) with the change amounting to one or two points in 11. The score did not change in 14 patients (3.4%). Between the baseline visit and the 4-month follow-up visit, the score improved in 390 patients (94.4%), got worse in 18 patients (4.4%) with changes amounting to 1 or 2 points in 14, and did not change in 5 patients (1.2%).

We found no differences in the ACS scores between boys and girls. The presence of psychiatric comorbidities was significantly correlated with the ACS scores through the entire follow-up period, with differences of 2.2 points in the baseline visit (95% CI, 1–3.6; P=.001), 1.6 points at 2 months (95% CI, 0.2–3; P=.027) and 1.9 points at 4 months (95% CI, 0.5–3.2; P=.006) between patients with comorbidities and patients without comorbidities, with the former having higher scores at every point in the followup. We did not find an interaction between these two factors and the ACS scores during followup. No other factor had a significant influence on the baseline score in the ACS.

We found a statistically significant correlation between the two instruments (Fig. 3).

Figure 3.

Concordance of the differences between the baseline and the third visits in the Conners scale and the FSI questionnaire. Concordance, 0.784 (95% CI, 0.737–0.822), P<.0001. Positive values indicate an improvement in family stress (FSI score) or ADHD symptoms measured by means of the Abbreviated Conners Scale.

(0.14MB).

The correlation coefficient for the baseline scores was 0.663 (Pearson's r; P<.0001). The correlation was still significant at the 2-month followup (Pearson's r, 0.731; P<.0001) and at 4 months (Pearson's r, 0.74; P<.001).

To study the sensitivity to change between the two assessment scales, we analysed the concordance of the increases in the scores of each questionnaire from the baseline visit to the 2-month followup and from the baseline visit to the 4-month followup, and we found an intraclass correlation coefficient of 0.825 (95% CI, 0.787–0.856; P<.0001) for the differences at 2 months, and of 0.784 (95% CI, 0.737–0.822; P<.0001) for the differences at 4 months.