Abstract
In recent years, there has been growing interest in explaining why some authors have found increased risk for certain recognized immune-mediated diseases, including asthma (Bernsen, de Jongste, & van der Wouden, 2003; Zekveld et al., 2006), type 1 diabetes mellitus (Cardwell, Carson, & Patterson, 2005), atopy (Bernsen et al., 2003; Karmaus & Botezan, 2002), and some cancers (Hemminki & Mutanen, 2001; Sanderson, Daling, Doody, & Malone, 2006) in firstborn children. The following common features of these diseases are important: They are immune mediated, they have been hypothesized to have a basis in early developmental damage, and (most importantly) their increasing worldwide prevalence over the past decade has not been explained to date (Becker, 2007; Mauny et al., 2005; Nystad, Magnus, & Gulsvik, 1998; Sultesz, Katona, Hirschberg, & Galffy, 2010). Immune programming hypotheses (Ogbuanu et al., 2010) and hygienic hypotheses have been proposed to explain the increase in these diseases, but the results have differed (Becker, 2007). An increased risk in the firstborn has also been described for some psychiatric disorders, such as autism (Gardener, Spiegelman, & Buka, 2009; Glasson et al., 2004) and schizophrenia (Haukka, Suvisaari, & Lonnqvist, 2004). Currently, however, no hypotheses to explain these findings have been proposed. Recently, Berger published a study showing that birth order does not affect ADHD risk (Berger & Felsenthal-Berger, 2009), and our group replicated this finding in a preliminary study that found an increased, but not statistically significant, prevalence of ADHD in the firstborn child (Masana Marin, Lopez Seco, Marti Serrano, & Acosta Garcia, 2011).
ADHD is the most frequently diagnosed infantile neurodevelopmental disorder. Despite increasing ADHD diagnoses, the prevalence is attributed to different diagnostic criteria in some countries (Montejano, Sasane, Hodgkins, Russo, & Huse, 2011). In some cases, ADHD is comorbid with autism (Rommelse, Franke, Geurts, Hartman, & Buitelaar, 2010), and comorbidity with asthma (Fasmer et al., 2011), type 1 diabetes mellitus (Gelfand et al., 2004), and atopy (Roth, Beyreiss, Schlenzka, & Beyer, 1991) has also been described. These findings suggest common pathways and/or common risk factors for these diseases (Becker, 2007; Brookes et al., 2006; Ficks & Waldman, 2009; Lasky-Su et al., 2008). Although previous reports have found no birth-order influence on ADHD risk, we hypothesize that being the firstborn is a risk factor for developing ADHD. Our objective was to compare the prevalence of ADHD diagnoses in firstborns with that of other birth-order positions while controlling for confounding factors, such as sex and parental age.
We selected all of the currently treated ADHD outpatients with a stable diagnosis for a minimum of 6 months (n = 748) from our database. Proband information, such as ADHD diagnosis, birth date, sex, birth order, maternal and paternal birth date, nuclear or stepfamily, biological or adopted, and number of siblings, were collected from the clinical records. Sibling data, such as psychiatric diagnoses, birth date, sex and birth order, were also collected from the clinical records. In addition, we included information on family features, such as maternal and paternal birth date, nuclear or stepfamilies, biological or adopted children, and number of siblings. Families with adopted sons, nonnuclear families, and families with only one child were excluded. To minimize the influence of individuals outside of our diagnostic range, families with sons (affected or unaffected) younger than 6 or older than 18 years old were also excluded. A total of 181 families met the inclusion criteria. We used all siblings without a clinical diagnosis of ADHD and who had no contact with our service as our unaffected controls. In total, 213 cases and 173 controls were identified. The present study was approved by the local ethics committee and was performed in accordance with the ethical standards of the 1964 Declaration of Helsinki and its later amendments. The demographic variables of ADHD and unaffected children are described in Table 1.
Demographic Variables of ADHD Cases and Unaffected Children.
At the time of childbirth.
Pearson chi-square.
The bivariate analysis showed that ADHD was associated with birth order (Table 2; p = .013). We then measured the degree of association by two binary logistic regression models: Model I did not adjust for confounding variables, and Model II adjusted for sex. The first model showed that the firstborn son had a 0.89 major risk of ADHD compared with children of other birth orders (Table 3, Model I). The firstborn child’s risk for ADHD was 0.77 and 2.1 times greater than the risks for the second and third born, respectively. Birth orders with frequencies less than 2 were not included in the analysis. The second model, which included confounding variables, analyzed sex as the only covariate, as we did not find any significant differences in maternal or paternal age between the ADHD patients and the unaffected controls. Although sibling age was significantly different between the ADHD patients (M = 11.59, SD = 2.49) and the unaffected participants (M = 10.38, SD = 3.68; p = .003), age was not included as a covariate, as we found a strong positive correlation between age and birth order (p < .001). The patient and control groups differed by sex, as there were more women (49.4%) than men in the control group and more men (78.4%) than women in the patient group (p < .001). The statistical power (1−β) of our sample to detect differences was 85.72%. When controlling for sex, the ADHD risk of the firstborn compared with other birth orders increased to 0.90 (Table 2, Model II).
ADHD Frequencies According to Birth Order.
Degrees of freedom.
Pearson chi-square.
p = .013.
Logistic Regression Models Used To Assess the Association Between ADHD Diagnoses and Birth Order While Adjusting for Sex.
Note: OR = odds ratio; CI = 95% confidence interval.
Binary logistic regression adjusted for sex.
Nagelkerke R2.
Our results suggest that birth order can be an ADHD risk factor in clinical samples and that firstborn children have nearly twice the ADHD risk of children with other birth orders. These results are in contrast to findings by Berger and Felsenthal-Berger (2009) that there is no relationship between birth order and ADHD, and they are in contrast to our prior preliminary study (Masana Marin et al., 2011). In that study, we found a nonsignificant increase in ADHD among firstborn children; those findings, along with the current findings, suggest the importance of replicating results in a selected sample with a case-control comparison. The small number of sons and age selection made our sample homogeneous enough to obtain reliable results. We also think that not assuming people above the age of 18 are unaffected by a disorder that was underdiagnosed in Spain until recent decades and not assuming that children less than 6 years old are unaffected can minimize the risk of excluding unknown cases. There were some families in our sample with more than one child with ADHD. As we used an extensive public mental health database (all of the cases involving some contact with a provincial area mental health service were included), it is unlikely that nondiagnosed cases and cases currently involving the treatment of a sibling were excluded. In addition, considering cases and controls inside the same nuclear family can minimize some of the biases from environmental risk. In contrast to the Berger et al. (Berger & Felsenthal-Berger, 2009) sample, the present study had few sons. Some authors have found that sibship size can contribute to or moderate the birth order risk in some disorders (Bernsen et al., 2003; Karmaus & Botezan, 2002), which can explain divergent results. Our study also has a few limitations, including the sample size and the sibship size. Therefore, we could not analyze the risks associated with fourth-born or higher birth orders, and our third-born results may not be representative, despite our statistically significant results. Finally, the ADHD cases and controls were obtained by clinical reports and were not compared with interviews. Future studies should control for these aspects. In conclusion, we found a 1.77-fold increased risk for ADHD in firstborn compared with second born children in a clinical sample. We emphasize that natality is decreasing in developed countries, and families with one or two children are the most prevalent. If being born first is an ADHD risk factor, then populations with a higher proportion of firstborn children will have an increased prevalence of this disorder. The increased risks for the firstborn child can partially explain the increasing incidences and changing geographical distributions of these modern “epidemic” diseases, confirming that birth order as a risk factor can open new and epidemiologically relevant fields, particularly in disorders with unexplained increases in worldwide prevalence (Heinrich, Richter, Magnussen, & Wichmann, 1998; Thomas, Birgit, Edith, & Austrian Diabetes Incidence Study Group, 2008).
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.