In this study, the validity of the parent form of the Turkish NIAS scale in children was investigated, and its psychometric properties were examined. The Turkish NIAS scale was shown to be valid and reliable. Through the reliability analysis, the NIAS demonstrated good internal consistency. The three-factor model of the Turkish NIAS was found to be in an acceptable structure by CFA analysis. NIAS scores were shown to be higher in underweight participants. The NIAS-parent version subscales showed expected convergent and divergent validity with the CEBQ, EDEQ-S, and RCADS scales in children.

In the study that the NIAS was developed and first validated, Cronbach's α for the Total score was 0.90 [7], 0.86 in the Chinese NIAS validity study [10], 0.88 in the Polish study [29], 0.84 in the Mexican study [30], and alpha 0.81 in the Turkish NIAS. In addition, the three-factor structure in our study also exhibited internal consistency similar to other studies.

In addition, the three-factor NIAS model was acceptable in the CFA analysis. NIAS-picky eating, NIAS-appetite, and NIAS-fear subscales of the goodness of fit indices were compatible. This model supports the separation of ARFID in the DSM-5 as in the original study [7]. In the scale, NIAS-picky eating indicated the presentation of selective/neophobic ARFID, NIAS-appetite the presentation of lack of interest, and NIAS-fear the presentation of fear of negative consequences [31].

In factor covariances and correlations between NIAS subscales, Picky eating, Appetite, and Fear subscale scores were found to be correlated with each other. These relationships show that the scale harmoniously evaluates the three presentations to be measured. Good inter-item correlation coefficients show that the three subscales measure the ARFID presentation in a standard way [7]. In addition, as in the studies conducted in the USA (adults), China (college undergraduates), and Mexico (adolescents), no gender differences were found in the NIAS total score and subscale scores [10, 30].

In our study, only NIAS-appetite was associated with a low BMI percentile, and underweight children had higher scores on this subscale. Still, no such relationship could be shown between NIAS-picky eating and NIAS-fear. The results for Turkey, located at the intersection of Europe and Asia, were in line with the literature on the relationship between selective eating and body weight in Western countries, as expected [USA (e.g., [7, 32], United Kingdom (e.g., [33]), and Australia (e.g., [34]]. In China, NIAS-appetite and NIAS-picky eating were associated with lower BMI in young adults, which the authors speculated was due to a less obesogenic food environment in China compared with the US, Australia, and Europe [10]. The literature for the USA/Europe/Australia is somewhat mixed, but systematic reviews suggest no relationship between children's picky eating and BMI in these countries [35]. Understandably, being less motivated to eat and restricting volume (i.e., Appetite ARFID symptoms) is protective against obesity in most countries, but restricting variety may only be protective against obesity in certain food environments. In highly obesogenic food environments such as the USA and Mexico, there is sometimes a slight positive correlation between picky eating and BMI [36]. Picky eaters, particularly young children, may indeed prefer certain processed foods with high calories. These foods are often highly palatable due to their high fat, sugar, and salt content, making them more appealing to picky eaters [37].

As hypothesized, there was evidence of divergent validity with the EDEQ-S, a measure of eating disorder symptoms maintained by cognitive restraint, weight and shape concerns, and fear of weight gain [11]. The presence of a small positive correlation between NIAS-picky eating and the EDEQ-S is consistent with findings from the adult NIAS. NIAS validation studies from treatment-seeking samples with eating disorders suggest that while the NIAS is valid and reliable in this population, it has a relatively poor ability to discriminate between ARIFD and other EDs due to the tendency of patients with non-ARFID EDs to endorse ARFID symptoms on the NIAS [31]. As a result, it is recommended to use NIAS and EDE-Q together to evaluate both ARFID and other eating disorders for a more complete picture of the eating behaviors and motivations behind them than is offered by either measure alone [31].

On the other hand, the CEBQ is a scale with different subscales that measures appetitive traits, early-emerging physiological, emotional, and behavioral responses to food and eating that are linked to weight gain (food approach traits) or protection against obesity (food avoidance traits) [15]. The CEBQ measures traits that may represent risk factors for ARFID. As hypothesized, NIAS-appetite showed positive correlations with food avoidance subscales and negative correlations with food approach subscales of the CEBQ. This relationship of NIAS-appetite with Appetite's physiological, motivational, emotional, and behavioral dimensions shows that it is consistent with the ARFID symptoms defined in DSM-5 [1]. NIAS-picky eating was negatively associated with enjoyment of food, while it was positively associated with satiety responsiveness, slowness in eating, emotional undereating, and especially food fussiness [38]. When EAT-26 scores were controlled, the direction and significance of the relationship between NIAS-appetite and CEBQ food approach and food avoidance did not change. Similarly, NIAS-picky eating and FF maintained a positive and strongly associated relationship. As in the adult validation study, the NIAS-fear subscale had weak or null relationships with appetitive traits. Whereas picky eating and appetite disturbances are early-emerging traits often first identified before age 5 in patients who develop selective and Appetite ARFID symptoms, fear ARFID has an acute onset often associated with a conditioning event like choking or vomiting [39, 40]. Temperamental risk factors for this ARFID presentation are likely to overlap with risk for anxiety and affective disorders [3].

The rates of psychiatric disorders accompanying ARFID are high, ranging between 57 and 95% [6, 41]. Anxiety disorders are most common in 36–72% [6, 42], and generalized anxiety disorder is the most common comorbid anxiety disorder in youth with ARFID (although this may reflect the base rate of GAD being higher than that of other anxiety disorders) [41, 43]. Mood disorders accompany ARFID with the second frequency between 17 and 33% [42, 44]. In the relationship with RCADS scores, there was a positive relationship between NIAS-appetite, NIAS-picky eating, NIAS-fear, total anxiety, and depression scores. When non-ARFID eating disorder symptoms were controlled, the relationship between Appetite, Picky eating and Fear subscales, and anxiety and depression measures did not change. Picky eating frequently contributes to the symptoms in clinical samples of those diagnosed with ARFID [39, 40, 45, 46]. Besides, picky eating has been hypothesized to be a transdiagnostic indicator of psychopathology in children, as it is associated with high emotional lability, cognitive rigidity, and concurrent symptoms of anxiety and depression [8, 47, 48]. A recent study showed that picky eating is associated with the symptoms of many concurrent psychopathologies in children, and the basis of this relationship is its association with OCD [49]. The present study also had a stronger relationship between picky eating and OCD symptoms compared to other RCADS subscales. This association may be clinically valuable in improving diagnosis and evaluation in the OCD group, which was generally diagnosed lately. Notably, there was also a stronger association between NIAS-fear and RCADS panic disorder compared to other NIAS/RCADS subscales. Although the association between fear-ARFID and panic disorder is poorly understood, there is evidence of symptom and functional overlap between the diagnoses. In the adult NIAS validation sample, there was a strong correlation between NIAS-fear and a measure of visceral sensitivity analogous to anxiety sensitivity in panic disorder [7].

However, it is noteworthy that the GAD scores of RCADS did not correlate with picky eating and appetite scale scores of NIAS. This lack of correlation may be attributed to the complex and heterogeneous presentation of ARFID, which often includes multiple physical symptoms and comorbid psychiatric [6, 50]. Furthermore, the lack of correlation may also be influenced by the unique eating behaviors and attitudes associated with ARFID. Individuals with ARFID may exhibit selective eating patterns, food avoidance based on sensory characteristics, and limited interest in eating, which may not be fully captured by the GAD scores of the RCADS [3, 51]. Future research should continue to explore the specific relationships between anxiety symptoms, eating behaviors, and comorbidities in individuals with ARFID to better understand the lack of correlation observed in this context. The NIAS subscales were diversely correlated with comorbid psychopathology symptoms, highlighting the importance of comorbidity in children with ARFID. Fink et al.’s study [52] at the gastroenterology clinic found that anxiety and depression scores were higher in cases with more severe ARFID symptoms screened by NIAS. The contribution of this data to clinical practice may be for the use of mirtazapine in treatment. Evidence that mirtazapine, which indicates treating adult anxiety and depression, can be used to treat ARFID is increasing daily [53,54,55]. Although mirtazapine contributes to weight gain by increasing appetite, it may facilitate the treatment of ARFID by treating anxiety and depression, which are often comorbid with ARFID. It may even support cognitive behavioral therapy (CBT) interventions used in treating ARFID. Similarly, a recent study has shown that selective serotonin reuptake inhibitors (SSRIs) and/or hydroxyzine show promise by reducing anxiety in treating ARFID [56].

To the best of our knowledge, this is the first study to investigate the relationship between ARFID symptoms and anxiety, depression, OCD symptoms, and eating behaviors through the NIAS, in addition to the validity study of the Turkish NIAS scale. The results of the study are significant because the study provides a first step towards providing tools to assist in the assessment of ARFID symptoms in young children aged 6–12 years and shows that the NIAS-parent version is a powerful measurement tool in the evaluation of symptoms in Turkish children. The limitations of the study include cross-sectional design, geographic limitations, and data being obtained from a non-clinical sample. In addition, since the age of the children is not suitable for filling out the scales, the parents' filling in the scales may create a bias. This study obtained the children's height and weight based on parental reporting. Studies have demonstrated that the inaccuracy in parents' reporting of their children's height and weight is generally due to underreporting, although it varies depending on the child's age and the country [57]. Studies need to be conducted to calculate the NIAS cut-off value for ARFID, including clinical samples.