The continually increasing prevalence of overweight and obesity is a serious health problem around the world and also in Finland (King, 2011; Vuorela et al., 2011). Unhealthy eating habits and taste preference for palatable foods have been viewed as risk factors contributing to obesity and overweight (Lanfer et al., 2012). Hyperpalatable foods, particularly ultra-processed foods with a high content of carbohydrates, sugar, fat and/or salt may possess addictive qualities, inducing strong craving and difficulties in controlling food intake in susceptible individuals (Gearhardt et al., 2011). With more understanding of the role of psychological factors and specific behaviors in the development of obesity, the concept of food addiction has gained increasing popularity among researchers in recent years (Penzenstadler et al., 2019). Regarding the measurement for food addiction, one of the most commonly used tool is the Yale Food Addiction Scale (YFAS). This is the first formal instrument to assess the addictive-like food intake originally designed based on the Diagnostic and Statistical Manual of Mental Disorders IV (DSM-IV-TR) diagnostic criteria for substance use disorders (Gearhardt et al., 2009). Several versions of the YFAS have been developed based on the original one for measuring food addiction in different contexts. For instance, a modified brief version of the YFAS (only one item assessing each symptom) and a child version of the YFAS have been proposed and applied (Flint et al., 2014). In addition, the YFAS has been translated into different language with sufficient validity and reliability in different countries, such as Chinese (Chen et al., 2015), French (Brunault et al., 2014), and Italian (Innamorati et al., 2015). However, there is no published research testing the psychometric properties of a Finnish version of the YFAS to measure food addiction in Finland.

Although there are still controversies about the terminology when referring to food addiction, mainly regarding whether to categorize it to substance-based or behavioral-based addictions, the presence of addictive-like eating is reasonably well established (Hebebrand et al., 2014; Gordon et al., 2018). For example, previous research has indicated that individuals with food addiction are characterized by food craving and a lack of control over their eating behavior (Meule, 2015). Moreover, findings from neurobiological studies have shown relations between intake of palatable foods (e.g., foods high in sugar and/or fat) and activity in the brain reward pathways (Grosshans et al., 2012). In addition, emotional functioning plays a role in addiction. For instance, emotion regulation difficulties are reportedly associated with substance use and addictive disorders, which may be attributed to problems in emotional stress regulation and inhibitory control related to frontal-limbic dysfunction (Li and Sinha, 2008; Tang et al., 2015).

Alexithymia, a personality trait characterized by difficulties in emotional awareness and expression, is a relevant factor in the context of addiction research. It was first proposed in 1973 with dimensions mainly involving difficulties in identifying and describing feelings, and externally oriented style of thinking (Sifneos, 1973). Alexithymia has long been reported to be associated with substance-related and addictive disorders such as alcohol use problems and pathological gambling (Parker et al., 2005; Kajanoja et al., 2019). It has been further suggested that alexithymia is linked to hypothalamic–pituitary–adrenal (HPA) axis dysfunction (Alkan et al., 2013), which may account for the role of alexithymia in addictions. Based on limited evidence, obese patients with food addiction seem to have more difficulties in emotion regulation and a higher level of alexithymia (El Archi et al., 2021). Nevertheless, research on associations between alexithymia and food addiction remains scarce.

The main aim of this study was to explore the relationship between alexithymia and food addiction, considering covariates including sample characteristics and psychological symptoms. As the psychometric properties of the Finnish YFAS (YFAS-F) are heretofore uninvestigated, we first intended to preliminarily validate the YFAS-F through determining the structural validity and internal consistency. Here, the YFAS-F was expected to be a valid and reliable tool for evaluating food addiction in our study sample of a Finnish general population. Furthermore, alexithymia was hypothesized to have associations with food addiction after taking into account potential confounders including sex and current psychological symptoms, and we also expected to establish a structural model confirming the link of alexithymia to food addiction.

The current study determined a significant association between alexithymia and food addiction. A higher level of alexithymia was related to more food addiction symptoms (controlling for the covariates including anxiety symptoms) and food addiction “diagnosis” (excluding anxiety symptoms). The structural relationship between alexithymia and food addiction was confirmed by conducting SEM. No sex differences were observed in their relationships. In addition, this study preliminarily validated the Finnish YFAS (YFAS-F) through testing the structural validity and internal consistency in general populations from a birth cohort.

The prevalence rate of food addiction in this study (5%) was lower than in other studies (around 10%) (e.g., Gearhardt et al., 2009; Brunault et al., 2014). However, the median symptom score was higher than the median scores reported in these studies, which indicated that the participants in our study may experience more food addiction symptoms but less clinical impairment or distress. In addition, although food addiction tends to be linked to obesity, it also occurs among normal-weight individuals (Pedram et al., 2013; Flint et al., 2014), which likely to some degree accounted for our finding that no correlations were found between BMI and food addiction.

Importantly, individuals with higher alexithymic traits experienced more food addiction symptoms, even after controlling for covariates including anxiety symptoms. The finding echoes previous evidence that alexithymia was associated with addictive behaviors such as pathological gambling (Parker et al., 2005). One of the potential mechanisms underlying the associations between alexithymia and addictions may be HPA axis dysfunction in alexithymic individuals, which reflects problematic regulation of stress responses (Alkan et al., 2013). Alexithymia was found to interact with perceived stress thus contributing to the development of stress-related mental health problems (Li et al., 2022). It is suggested that alexithymia may function as defense mechanism and addictive behaviors as maladaptive strategies to buffer psychological distress (Selfhout et al., 2009). Additionally, the proneness to addictions in alexithymic individuals can be speculated from the aspect of emotion regulation and social engagement. For instance, people may use addictive substances such as alcohol for alleviating their difficulties in emotional awareness and interpersonal function (Cruise and Becerra, 2018).

El Archi et al. (2021) have reported a higher alexithymia level in obese patients who were diagnosed as having food addiction, which is in line with the result of group comparison in our study. However, this association did not survive in the logistic regression analysis when adjusting for anxiety symptoms, suggesting that alexithymia had a relatively weak relation to the presence of food addiction compared to symptom levels of food addiction. Nevertheless, the results indicated that alexithymia still showed a trend toward the “diagnosis” score of the YFAS-F. In light of previous research showing that negative emotions were antecedents of binge eating (Dingemans et al., 2017), psychological distress such as anxiety may be involved in the link between alexithymia and food addiction. In addition, the relatively low detection rate (5%) in our sample may to some extent hinder detection of associations between alexithymia and “diagnosis” of food addiction. Notwithstanding, as there were no previous studies that investigated the link between alexithymia and food addiction considering the role of emotional distress, it is strongly recommended to consider the interaction effects of emotion factors when studying on the role of alexithymia in food addiction in future research.

According to previous studies, sex differences have been found in the prevalence of food addiction with mixed results (Pedram et al., 2013; Stojek et al., 2021). In our study, more females met the “diagnosis” of food addiction as measured by the YFAS-F compared to males, but no sex differences were observed in terms of the food addiction symptom levels. Moreover, the current results indicated no sex differences in the association between alexithymia and food addiction. Given that evidence on the role of alexithymia in food addiction is still lacking, more studies are needed to explore whether alexithymic traits as a developmental factor for food addiction varies across sex, which would improve understanding of food addiction and provide possibilities for personalized intervention.

By conducting SEM, we further confirmed the significant relationship between alexithymia and food addiction explained by the 3 dimensions of the TAS-20 and the 8 criteria of the YFAS-F with good overall fit of the measurement and structural model. To ensure correct modeling in the case of a small negative value of residual variance, a strategy reasonably used is to fix the residual variance of DDF to zero. We tried an ad hoc EFA and found that most of the items for DDF cross-loaded onto the other two distinct dimensions DIF and EOT, which is one important reason for leading to this phenomenon in our study. This is also in accordance with a previous study indicating that different types of alexithymia could be differentiated by either high DIF or EOT, but not DDF (Kajanoja et al., 2017). Overall, this is the first study determining their associations in general population with using SEM, a multivariate statistical method commonly applied for providing model-based causal evidence (Ullman, 2006). Our findings highlight the significance of alexithymia as a potential risk factor for food addiction, especially the symptom levels of food addiction, in a sample of Finnish general population.

Regarding the structural validity of the YFAS-F, in line with previous studies, a single-factor model for the criteria of the YFAS-F was supported by CFA with a sufficient model fit. However, a three-factor model for the items was identified in the present study, which is inconsistent with the original version of the YFAS (Gearhardt et al., 2009). Similarly, previous studies on the validation of the French and Italian versions of the YFAS have also shown a poor fit of the one-factor model for the items by using CFA or suggested a two-factor solution by EFA (Brunault et al., 2014; Innamorati et al., 2015). Considering differences in sample characteristics and the nature of the component, the three-dimension food addiction are explicable. For instance, the second dimension, consisting of four items measuring a criterion “Giving up social, occupational, or recreational activities” and item 16 (problems in the ability to function effectively such as daily routine, job/school, and social activities), represents social damage. The study by Brunault et al. (2014) suggested a two-factor model with high loading values of these items in the second component as well. Additionally, the third dimension consisted of two items (20 and 21) measuring the criterion “Tolerance,” which was reported by only 3.3% of the participants in our study. Unlike other substances such as tobacco and alcohol, the amount of daily food intake is relatively limited. Moreover, it is suggested that tolerance is not a central element in terms of behavioral addictions as increased specific behaviors are not for reducing negative emotions or increasing pleasure (Blaszczynski et al., 2008; Starcevic, 2016). Hence, the current results may to some extent reflect a behavioral-related feature of food addiction. Future research exploring potential subtypes of food addiction with a clinical sample would be beneficial to better understand its nature. In the present and previous studies, few items of the scale have shown low psychometric qualities, such as item 24 (Gearhardt et al., 2009; Meule et al., 2012; Chen et al., 2015). Hence, advanced approach such as Rasch modeling would be beneficial to the development of the instrument via detecting the properties of the items, improving the accuracy of measurements, and computing the performance of respondents (Boone, 2016).

Some limitations of this study should be acknowledged. Firstly, the YFAS-F was validated in the study sample of Finnish parents, and thus the generalization of the findings was limited by the sample characteristics. Secondly, regarding the validation of the YFAS-F, test–retest reliability and convergent/discriminant validity were not determined in the current study, which warrants future studies, ideally with concurrent measuring eating disorders such as binge eating and bulimia nervosa. Thirdly, due to the limited sample size, we did not randomly split the sample for EFA and CFA each or test the 3-factor solution for the items by using CFA in an additional group, which is worth considering in further studies. Fourthly, the YFAS 2.0 developed based on the DSM-5 criteria has not been translated into Finnish yet. Nevertheless, given the large overlap with the DSM-IV criteria and the reported less relevance of the added criteria in DSM-5 in the context of food, it is suggested that the YFAS can still be effectively used for screening and investigating food addiction (Meule and Gearhardt, 2014). The comparability of these two versions in Finnish population is worth exploring in the future. Lastly, although SEM suggested the model-based effect of alexithymia on food addiction, causal inferences cannot be made due to the use of cross-sectional data. Notwithstanding, given limited evidence in this regard, our findings provide insights into the role of alexithymia in food addiction. The current understanding would benefit from further follow-up investigation.

Individuals with a higher level of alexithymia experienced more food addiction symptoms, and no sex differences were observed in this relationship. Simultaneously, sufficient structural validity and internal reliability of the YFAS-F were determined, suggesting that this tool can be reasonably used in our sample. To the best of our knowledge, this is the first study examining the associations between alexithymia and food addiction using regression analysis and SEM among general population. Moreover, this is the first study testing the psychometric properties of the Finnish version of YFAS. The current findings suggest that alexithymia may play a role in food addiction, providing implications for implementing preventive and psychotherapeutic interventions such as emotion regulation training in individuals with food addiction. Given that this is only preliminarily validation and has some limitations regarding the generalizability beyond pregnancy cohort population, more studies are recommended to assess the overall validity of the YFAS-F in more diverse populations also in Finland.

The datasets presented in this article are not readily available because of restriction imposed by the Finnish law, and the study’s ethical permissions do not allow sharing of the data used in this study. Requests can be directed to the Board of the FinnBrain Birth Cohort Study. Requests to access the datasets should be directed to Hasse Karlsson, hasse.karlsson@utu.fi.

The studies involving human participants were reviewed and approved by the Ethics Committee of the Hospital District of Southwest Finland (14.6.2011 ETMK:57/180/2011 § 168). The patients/participants provided their written informed consent to participate in this study.

RL: conceptualization, methodology, formal analysis, investigation, data curation, writing – original draft, and writing – review and editing. JK: conceptualization, supervision, methodology, investigation, data curation, and writing – review and editing. JT: investigation, data curation, and writing – review and editing. LK: investigation, data curation, project administration, resources, funding acquisition, and writing – review and editing. HK: investigation, data curation, project administration, resources, and writing – review and editing. MK: supervision, funding acquisition, methodology, and writing – review and editing. All authors contributed to the article and approved the submitted version.

The work was supported by the Signe and Ane Gyllenberg Foundation and Academy of Finland. Funds for the open access publication fee were received from the State Research Grants of the Hospital District of Southwest Finland.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.