Introduction
Avoidant/restrictive food intake disorder (ARFID) is characterized by reduced dietary intake leading to medical and/or psychosocial impairment, independent of body image concerns. Its prevalence in gastrointestinal (GI) disorders remains unclear despite frequent symptom-induced restrictive eating behaviors.
Aims & Methods
We aim to assess the prevalence and characteristics of ARFID in patients with various GI disorders compared to healthy volunteers (HV), and its associations with symptom severity, extraintestinal somatic symptoms and depression.
Data from 92 HV and 467 GI patients across seven tertiary care diagnostic categories —including irritable bowel syndrome (IBS, n=88), functional dyspepsia (FD, n=83), (refractory slow transit) constipation (n=59), reflux (n=107), globus (n=45), (non-obstructive) dysphagia (n=37), and achalasia (n=48)—were analyzed. Questionnaires included Rome IV criteria, ARFID screening, extraintestinal symptoms and depression scores (PHQ-12/9). Positive NIAS screening (NIAS+) was defined as meeting any subscale cutoff (≥10 for Picky Eating, ≥9 for Appetite, and ≥10 for Fear). Medical and/or psychosocial impairment in NIAS+ individuals was evaluated using the ARFID DSM-5 checklist. ARFID+ was defined as NIAS+ plus DSM-5 impairment. Odds ratios were calculated using a multiple logistic regression model with three-level hierarchical approach. Odds of fulfilling ARFID or NIAS criteria between DGBI diagnoses (step 1) were adjusted for z-scores of symptom severity (step 2) and extra-intestinal somatic symptoms and depression (step 3). Esophageal and non-esophageal disorders were analyzed separately.
Results
Table 1 shows demographics and characteristics of patients. Compared to HV, ARFID+ was significantly associated with IBS (OR: 15.00, p < 0.001), FD (OR: 10.59, p = 0.002), constipation (OR: 18.21, p < 0.001), dysphagia (OR: 7.03, p = 0.024) and reflux (OR: 5.16, p = 0.036), but not with achalasia and globus. Within disease categories, there were no differences in the odds of ARFID+ after adjusting for symptom severity (step 2) and for comorbid extraintestinal symptoms and depression (step 3). Increase in PHQ9, contributed independently to the presence of ARFID in non-esophageal disorders (OR: 1.11, p = 0.009).
| | Non-Oesophageal | Oesophageal |
| Demographics | HV, N = 92 | IBS, N = 88 | FD, N = 83 | Constipation, N = 59 | Reflux, N = 107 | Globus, N = 45 | Dysphagia, N = 37 | Achalasia, N = 48 |
| Gender (% female) | 57% | 82% | 86% | 95% | 54% | 61% | 60% | 42% |
| Age (years) | 38 (14) | 33 (11) | 38 (14) | 46 (15) | 48 (16) | 52 (12) | 52 (16) | 54 (11) |
| BMI (kg/m²) | 23.4 (21.0-25.4) | 23.7 (20.4-26.3) | 21.7 (19.7-23.8) | 22.8 (20.7-25.3) | 24.7 (22.2-27.4) | 24.6 (21.0-27.7) | 23.7 (21.4-28.2) | 24.5 (22.4-26.3) |
| ARFID+ | 2 (2%) | 22 (26%) | 16 (19%) | 17 (29%) | 11 (10%) | 4 (9%) | 5 (14%) | 5 (10%) |
| Model A step 1 Uncorrected vs. HV | | 15.00 (3.41–66.02)*** | 10.59 (2.36–47.61)** | 18.21 (4.02–82.47)*** | 5.16 (1.11–23.90)* | 4.29 (0.76–24.33) | 7.03 (1.30–38.06)* | 5.23 (0.98–28.06) |
| Model B step 1 Uncorrected vs. disease mean | | 0.051 (-0.36-0.47) | -0.297 (-0.74-0.14) | 0.246 (-0.20-0.69) | -0.035 (-0.66-0.59) | -0.220 (-1.07-0.63) | 0.275 (-0.53-1.08) | -0.020 (-0.81-0.77) |
| Model B step 2 Adjusted for Severity | | -0.421 (-0.90-0.06) | 0.066 (-0.45-0.59) | 0.355 (-0.15-0.86) | -0.122 (-0.75-0.51) | -0.082 (-0.96-0.80) | 0.227 (-0.60-1.05) | -0.023 (-0.82-0.77) |
| Model B step 3 Adjusted for Severity & Extraintestinal Symptoms and depression | | -0.548 (-1.158, 0.061) | 0.123 (-0.430, 0.677) | 0.425 (-0.132, 0.982) | -0.163 (-0.848, 0.523) | -0.049 (-0.939, 0.842) | 0.160 (-0.679, 1.000) | 0.051 (-0.812, 0.914) |
Table 1. Data are presented as n(%), mean(SD), median(IQR), or OR(95%). HV = Healthy Volunteers. Unadjusted ORs compare each disorder to HV (Model A). In Model B, each condition is compared to the disease mean within esophageal and non-esophageal groups. Negative values in adjusted models indicate log-odds differences from the group-specific disease mean. **p < 0.01, ***p < 0.001, ****p < 0.0001.
Conclusion
ARFID symptoms are prevalent across various GI disorders. While mechanical factors may drive ARFID in esophageal conditions, psychological factors (e.g., depression) may play a role in ARFID in non-esophageal disorders. Further research is needed to establish the contribution of mechanical factors to ARFID in esophageal conditions. Routine ARFID screening is crucial to address psychosocial and nutritional impacts in GI patients.