Comment on: Social anxiety disorder in Saudi adolescent boys: Prevalence, subtypes, and parenting styles as a risk factor

Sir,

This is in reference to the article entitled, “Social anxiety disorder in Saudi adolescent boys: Prevalence, subtypes, and parenting styles as a risk factor.”[1]

Considering the lack of literature on the social anxiety disorders (SADs) in adolescents, the current study is commendable. With a relatively higher sample size compared to previous studies, this study has made precise estimates of SAD. The use of Liebowitz Social Anxiety Scale (LSAS) which was validated in Arabic version among adolescents is appreciated as this would have minimized measurement error.[2] Moreover, the Arabic version of the questionnaire was further subjected to internal consistency through Cronbach's alpha. This actually enabled the author to obtain total score for each check domain from various items collected. Further, anonymous data collection would have further increased the credibility of the estimates by encouraging the students to respond freely. However, we felt that a clear mention of some of the following methodological issues would help the readers understand better and make it easier to replicate the study in their own context.

Although, in the background, the authors had clearly mentioned that SAD was common in the age group of early to late adolescence (10-19 years), restricting the samples to a narrow age range (17-18 years) may not yield the precise estimates of SAD among adolescents. In the bivariate analysis, the authors had further compared the prevalence of SAD in 17-year-olds versus 18 years of age. Since the information related to age was collected through self-report, there could have been misclassification error. Since this study is about SAD in Saudi boys, it would have been worth mentioning school or college drop-outs in this age group.

There is a lack of information on the calculation of sample size and the sampling method adopted to arrive at the two government schools in the study. As students in private schools can differ in their characteristics from students of government schools, inclusion of a private school in the study could have given a more precise estimate of SAD. Differential type of training and focus given toward extra-curricular activities and self-development opportunities provided in these two different types of schools will have an impact on the prevalence of SAD. Since this prevalence of 11.7% is an estimate of a sampled population, giving confidence interval will give the reader a more precise understanding.

As the two versions of LSAS, namely, clinician assessing scale and self-report scale, are available, the version of LSAS which was used in this study was not mentioned. Further, there could be differences in outcome measurement when two different formats of LSAS are used. If assessed by clinicians, situation-specific examples related to the items of the questionnaire might have been provided by the clinician, whereas in the self-reported form, the children have to orient their thoughts toward the particular situation provided before arriving at the responses.[3]

In this type of study where there is a probability of social desirability bias, the study setting will have an influence on outcome measures. No mention was made of where the questionnaires were administered whether in the classroom or at home. If it was administered among selected boys in class, how the authors managed the practical challenges of ethical issues and hindrance response from nonparticipating students was not indicated.

Since there are previous studies available regarding the prevalence of SAD, the authors clearly state in their rationale that their main focus was toward classifying parenting style and its association with SAD. Hence, giving collateral evidence on the impact of severity of SAD on mental health and association of parenting style with this particular issue in the discussion section would have been more helpful. Apart from the risk of social desirability bias, a few of the parenting style factors may not have temporal relationship with SAD. Although cross-sectional studies are not meant to prove the lack of permanence mentioning it under limitations, they would have added value to the article.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.