The Arabic Mood and Feelings Questionnaire: Psychometrics and Validity in a Clinical Sample

Abstract

The purpose of this study was to provide clinicians in the Arab World with a child and adolescent depression screening tool. Child and parent versions of the Mood and Feelings Questionnaire (CMFQ and PMFQ respectively) were translated to Arabic and administered along with the Strengths and Difficulties Questionnaire (SDQ) to 30 children and adolescents and with mood disorders and 76 children and adolescents with other psychiatric disorders seeking treatment at a child and adolescent psychiatry clinic. DSM-IV diagnoses were generated through clinical interviews by a psychiatrist blinded to self-reports. Internal consistency for both versions was excellent with moderate inter-informant agreement and good convergent validity with the SDQ emotional symptoms subscales on the child and parent forms. The CMFQ and PMFQ significantly differentiated between currently depressed participants and those with other psychiatric disorders. CMFQ scores were a stronger predictor of categorization into depressed and non-depressed groups than the PMFQ. Two modes of cutoffs were calculated with one favoring sensitivity (a score of 26 for the CMFQ and 22 for the PMFQ) and another favoring specificity (a score of 31 for the CMFQ and 28 for the PMFQ).

Screening efficiency of the Mood and Feelings Questionnaire (MFQ) and Short Mood and Feelings Questionnaire (SMFQ) in Swedish help seeking outpatients

“…Parent MFQs have discriminated as well as child ratings in most [11,12,14] but not in all [16] comparisons. However, the parent MFQ did worse when the diagnosis was established with just an interview with the child [16] as opposed to studies which used separate interviews with child and parents [11,12] or a clinical diagnosis [14]. On the contrary, MFQ-parent was better than MFQ-child at discriminating any mood disorder.…”

“…There are matching versions for self (with 33 items) and parent report (with 34 items) [10] and a short MFQ with 13 items [17]. Internal consistency has been high with Cronbach's alphas for MFQ about .90-.96 [10,11,[13][14][15]26] and for SMFQ about .85-.89 [10,13,15,19]. Correlations between the long and short versions have been very strong (r = .95) [13].…”

“…The Mood and Feelings Questionnaire (MFQ) was developed to screen for depression in epidemiological samples [10]. It has also been validated in clinical samples [11][12][13][14][15][16] and is recommended as a screening tool [5]. The largest clinical validation contained ongoing patients from ADHD and mood disorder clinics along with high risk and epidemiological samples in the U.S. MFQ performed well for child and parent ratings and in both genders and in children as well as in adolescents with area under curve (AUC) mostly good (.80-.90).…”

“…Earlier studies have suggested cut-offs based on maximum separation of 27-29 (MFQ-child) and [21][22][23][24][25][26][27] (MFQ-parent) [11][12][13]16]. Suggested cut-offs favouring sensitivity were 26-28 (MFQ-child) and 22 (MFQ-parent) [14,15].…”

“…SMFQ-child and SMFQ-parent has correlated strongly to MFQ and has discriminated as MFQ in adolescents but just fair in younger samples. MFQ and SMFQ have usually been validated in epidemiological, high risk or among already diagnosed and clinical samples but not in unselected help seeking child psychiatric outpatients except the MFQ Arabic version in a small study of mostly adolescents [14]. Another recent study included help seeking adolescents but they were recruited and selectively referred for a cognitive behaviour intervention for depression [15].…”