TY - JOUR
T1 - Longitudinal exploration of biopsychosocial profiles in individuals with anorexia nervosa
AU - Halls, Daniel
AU - Batchelor, Rachel
AU - Holetic, Victoria
AU - Leppanen, Jenni
AU - Williams, Steve
AU - Tchanturia, Kate
N1 - Funding Information:
One of our main findings, of persistent psycho-social difficulties in individuals with AN, is supported by numerous cross-sectional (Martín et al., 2019; Sander et al., 2021; Amianto et al., 2022) and follow up studies (Li et al., 2020; Castro et al., 2021; Pleplé et al., 2021; Di Lodovico et al., 2023). These findings highlight the prominent role that psycho-social difficulties play in AN. Previous work has documented that when individuals with AN weight recover, anxiety scores remain higher than those of HC participants (Castro et al., 2021). Following treatment anxiety has been suggested to not improve while BMI did (Tchanturia et al., 2019; Li et al., 2020; Toppino et al., 2022). Moreover, work, and social functioning along with anxiety, but not BMI, have also been shown to be predictors of food avoidance (Di Lodovico et al., 2023), while changes in depression scores were positively correlated to ED symptoms, but not changes in BMI (Pleplé et al., 2021). Though this study did not directly compare BMI and psycho-social difficulties, the magnitude of difference and probabilities from the cross-validation shown in this study, support the notion that psycho-social difficulties are central in AN.Our null finding of no group differences regarding autistic traits is surprising as previous work has suggested greater likelihood of autistic traits in individuals with AN (Sedgewick et al., 2019; Kerr-Gaffney et al., 2021). However, advantages of using Bayesian frameworks are inferences regarding potential reasons for null results can be made (Quintana and Williams, 2018). The cross-validation suggested that neither hypotheses can be fully supported and this uncertainty likely reflects an aspect of study design (Quintana and Williams, 2018). Arguably, the main possible reason for this uncertainty could be the use of the AQ-10 in measuring autistic characteristics. The AQ-10 has been criticised for several failings such as high false negatives rate (Ashwood et al., 2016), poor internal reliability and measuring multiple factors beyond autism (Tchanturia et al., 2019; Taylor et al., 2020). The AQ-10 has also been shown to be weakly correlated to other clinical measures that are considered more robust in measuring autism (Ashwood et al., 2016; Sedgewick et al., 2019). This has led to questions regarding the appropriateness of the AQ-10s role in research (Ashwood et al., 2016; Taylor et al., 2020). Another possibility is, as the variance in the autistic component is higher in the AN group than the HC group, any effect maybe too small to be detected by this study. More data could overcome this as well as clearing uncertainty in the cross-validation comparison.This work was supported by supported by MRC-MRF Fund (MR/S020381/1; BiomaRkers for AnorexIa NErvosa and autism spectrum Disorders-longitudinal study And MR/R004595/1 The Triple A study [Adolescents with Anorexia and Autism]: A search for biomarkers).
Funding Information:
This work was supported by supported by MRC- MRF Fund (MR/S020381/1; BiomaRkers for AnorexIa NErvosa and autism spectrum Disorders-longitudinal study And MR/R004595/1 The Triple A study [Adolescents with Anorexia and Autism]: A search for biomarkers).
Publisher Copyright:
© 2023
PY - 2023/11
Y1 - 2023/11
N2 - Background: Previous work in individuals with Anorexia Nervosa (AN) has demonstrated a range of psycho-social difficulties such as increased anxiety, depression, obsessive-compulsive symptoms, as well as difficulties in work and with interpersonal interactions. However, making inferences regarding the stability of these psycho-social difficulties from previous studies is challenging, due to lack of a control group and known frequentist statistical issues. Methods: 134 participants, 40 healthy controls (HC) and 94 participants with AN, completed self-reported measures designed to explore eating disorder concerns, body mass index, mood symptoms, work and social functioning as well as traits associated with autism at two time points, two years apart. A principal component analysis and Bayesian mixed effects models were used to build and explore group differences in bio-psychosocial profiles at time points. Results: The Bayesian models demonstrated evidence for individuals with AN having higher scores for a component representing psycho-social difficulties and lower scores for a component representing biological difficulties compared to HC, at both time points. There was no evidence of a group difference for a component representing autism. Conclusions: Our results demonstrate that persistent psycho-social difficulties are a feature in individuals with AN.
AB - Background: Previous work in individuals with Anorexia Nervosa (AN) has demonstrated a range of psycho-social difficulties such as increased anxiety, depression, obsessive-compulsive symptoms, as well as difficulties in work and with interpersonal interactions. However, making inferences regarding the stability of these psycho-social difficulties from previous studies is challenging, due to lack of a control group and known frequentist statistical issues. Methods: 134 participants, 40 healthy controls (HC) and 94 participants with AN, completed self-reported measures designed to explore eating disorder concerns, body mass index, mood symptoms, work and social functioning as well as traits associated with autism at two time points, two years apart. A principal component analysis and Bayesian mixed effects models were used to build and explore group differences in bio-psychosocial profiles at time points. Results: The Bayesian models demonstrated evidence for individuals with AN having higher scores for a component representing psycho-social difficulties and lower scores for a component representing biological difficulties compared to HC, at both time points. There was no evidence of a group difference for a component representing autism. Conclusions: Our results demonstrate that persistent psycho-social difficulties are a feature in individuals with AN.
KW - Anorexia nervosa
KW - Bayesian
KW - Follow up
KW - Mixed effects models
KW - Psycho-social
UR - http://www.scopus.com/inward/record.url?scp=85173133366&partnerID=8YFLogxK
U2 - 10.1016/j.jpsychires.2023.09.001
DO - 10.1016/j.jpsychires.2023.09.001
M3 - Article
C2 - 37806284
AN - SCOPUS:85173133366
SN - 0022-3956
VL - 167
SP - 16
EP - 22
JO - Journal of psychiatric research
JF - Journal of psychiatric research
ER -
