Research on feeding and eating disorders in India: A narrative review

Abstract

Despite growing evidence of their prevalence, research on feeding and eating disorders (FEDs) in India has been sporadic. This narrative review aimed to summarize the research on FED in India and set priorities for future research and translation of evidence. An electronic search was conducted in the MEDLINE, PsycINFO, and Google Scholar databases to identify relevant English peer-reviewed articles from April 1967 to July 2023. The extracted data from these studies included author names, publication year, research location, type of intervention (for interventional studies), nature of comparator treatments, and main outcomes or findings. We found a rising trend in the prevalence of EDs in India. Adolescent age group, female sex, higher socioeconomic status, family history of mental illness or disordered eating, and borderline personality pattern were risk factors for EDs. For feeding disorders (FDs), childhood age group, malnutrition, pregnancy, psychosis, intellectual disability (ID), and obsessive-compulsive disorder (OCD) were putative risk factors. Both physical and psychiatric comorbidities were common in FEDs. Culture appears to exert a pathoplastic effect on symptom presentation in FEDs; an illustrative example is the documented nonfat phobic variant of anorexia nervosa (AN) in India. Research on management has focused on using assessment tools, investigations to rule out medical comorbidities, psychosocial and family-based psychotherapies, nutritional rehabilitation, pharmacotherapy, and neuromodulation approaches. Whereas the publication output on FEDs in India has increased over the last decade, it remains an under-researched area, with a striking paucity of original research. Future research priorities in FEDs include conducting country-wide registry-based studies to offer real-world insights, longitudinal research to identify culturally relevant risk factors, and developing brief, culturally sensitive diagnostic instruments for FEDs in the Indian context. This will help generate locally relevant epidemiological data on FEDs and inform treatment and prevention strategies.

INTRODUCTION

Feeding and eating disorders (FEDs) encompass atypical eating or feeding behaviors that cannot be attributed to other medical issues and are not considered developmentally typical or in keeping with the individual’s sociocultural background. Feeding disorders (FDs) encompass a spectrum of conditions, including restricted or limited food intake such as avoidant-restrictive food intake disorder (ARFID) and behaviors such as consuming non-edible substances known as pica or voluntary food regurgitation referred to as rumination-regurgitation disorder (RRD). EDs, which include anorexia nervosa (AN), bulimia nervosa (BN), and binge eating disorder (BED), have abnormal eating behaviors and varying degrees of fixation on diet, body image, and weight as their key clinical features.[1]

Children often face a spectrum of food-related issues, ranging from minor preferences and difficulties with specific foods to more severe FDs that extend beyond typical variations in eating behavior. Research suggests that approximately 18% to 50% of typically developing children encounter a feeding problem. However, the prevalence of FDs is notably elevated in children with developmental disorders, with as many as 89% experiencing these more severe issues.[2] The case histories of patients with ED have long documented feeding problems and selective eating during childhood, dating back to early case reports. However, limited research has explored the connection and continuity between these early feeding issues and the development of EDs later in life. In keeping with the available evidence, a single classification system that can be applied consistently across different age groups and is sensitive to the developmentally specific manifestations of ED would provide a more accurate representation of the course of these disorders.[3]

The International Classification of Diseases, 11^th edition (ICD-11), consistent with the Diagnostic and Statistical Manual of Mental Disorders, 5^th Edition (DSM-5), includes the following feeding or EDs: AN (6B80), BN (6B81), BED (6B82), ARFID (6B83), Pica (6B84), RRD (6B85), and Other Specified Feeding or Eating Disorders (OSFED). The diagnoses of FEDs should not be applied to categorize minor concerns related to eating or behaviors that are prevalent and culturally accepted. These categories addressed three fundamental limitations within the International Classification of Diseases, 10^th edition (ICD-10) ED guidelines. First, the ICD-10 division of FEDs into distinct categories do not align with empirical evidence and contemporary clinical practices. Second, these guidelines lead to inconsistency in the assigned diagnoses for EDs, with a significant number falling under the “atypical” categories or the “residual” other specified or “unspecified” categories. Third, the ICD-10 guidelines do not adequately acknowledge the diverse cultural variations in the clinical presentations of FEDs. The FD classification was problematic because it included one heterogeneous condition with unclear boundaries from EDs (ARFID) and two less common and more specific conditions in children and adults (pica and RRD). Furthermore, the category of psychogenic vomiting is no longer included in FED in ICD-11, possibly because it typically does not include any body image concerns or abnormal eating patterns.[3]

FEDs are a growing concern in South Asia, although they have historically received less attention than in Western countries. To address the rising prevalence, severity, burden, and mortality risks linked to ED and recognize the emerging data on FD, it is crucial to prioritize research on FED in India. This will aid prevention efforts, support effective interventions, and address a significant priority within global health.[4,5]

Traditional customs related to dietary habits and body image in India possibly play a significant role in the lower prevalence of ED compared with the West. Cultural practices in India emphasize the sanctity of food and avoiding wastage. Gratitude and moderation are ingrained in eating habits, often commencing meals with prayers.[6,7,8] This rich diet heritage parallels the prevailing ideals of body esthetics, which was in contrast with Western standards. Figurines from the Mauryan era depict women with attributes signifying fertility, such as full breasts and wide hips. South Indian literature describes beautiful women as possessing skin with a lustrous glow, substantial hips, and voluminous hair. Perhaps this perspective was a protective factor. Depiction of the “slender” physique with low body weight as appealing by the media is associated with an elevated risk of EDs. Exposure to Western norms through media and globalization may undermine the protective influence of traditional South Asian cultures.[9,10]

Earlier reviews report lower prevalence and possible underestimation of FEDs in India compared with the West. They also hint at possible cultural distinctions in clinical presentation, warranting further evaluation.[5] The present review seeks to summarize the existing literature on FED in India and offer an update on the noteworthy research advancements. Furthermore, we aim to provide a blueprint for future research in this domain. We did not perform a quality appraisal of included studies or delve into a comprehensive discussion of global literature as this was an India-focused narrative review.

METHODOLOGY

Search strategy

MEDLINE, Google Scholar, and PsycINFO databases were used to identify relevant English peer-reviewed articles from April 1967 to July 2023. Our search utilized various combinations of text terms, such as “eating disorder”, “anorexia nervosa”, “bulimia nervosa”, “binge eating disorder”, “feeding disorder”, “pica”, “avoidant-restrictive food intake disorder”, “rumination-regurgitation disorder”, “treatment”, “epidemiology”, “comorbidity”, “management”, “medications”, “behavioural intervention”, “psychosocial intervention”, and “India.”

Study selection

Due to the substantial variations in study populations, methodologies, and outcome measures, we decided not to pursue a systematic review. Instead, we opted for a narrative review. We included all published articles, provided they met the following criteria:

1. The study evaluated any of the FEDs described in the literature.

2. The research was conducted in India including the Indian population or animal studies.

3. The studies were published in English-language peer-reviewed journals.

Data extraction

The authors conducted a thorough screening process for studies that met the broad inclusion criteria. First, we reviewed the titles and abstracts of potentially relevant studies. The full texts were examined if the abstracts did not provide sufficient information for an inclusion decision. After eliminating duplicates, a consolidated list of abstracts for inclusion was compiled. We also conducted a manual search of the references of included studies to identify relevant articles. We did not search citation indexing services and conference proceedings due to concerns about incomplete data reporting and uncertainties regarding the quality of such studies.

The data extraction table was prepared from the selected articles to include author names, publication year, the cities in India where the research was conducted, the type of study, assessments performed, and significant findings. Based on the design, focus, and findings of the included studies, we grouped them into various categories, and these results are presented as follows.

RESULTS

Study settings and designs

We identified a total of 141 Indian publications on FEDs of the following types: original research articles (N = 61), case reports and case series (N = 57), and review articles (N = 23). Of the 61 original articles identified, it was seen that the majority (N = 47) of the papers were published in the last decade (2013–2023). There were only four ED articles before 2000 arising from India. Original studies that have evaluated FEDs in the Indian population can be categorized as community-based [Table 1] and clinic based [Table 2].

Table 1.

Original research conducted in community settings

Author, year, and city	Study design; assessment tools	Sample size and characteristics	Major finding
King & Bhugra, 1989, Yamuna Nagar[11]	Cross-sectional observational study; Hindi version of EAT-26	n=574 female school adolescent girls (age range of 14–23 years)	29% had disordered eating
Srinivasan et al., 1995, Chennai[12]	Cross-sectional observational study; EAT-40 BITE DSM-III Criteria	Step1: n=602 Step 2: n=210 The sample comprised medical students of both sexes	EDs were diagnosed in 14.8% of the participants. None had a syndromal ED diagnosis
Srinivasan et al., 1998, Chennai[13]	Cross-sectional observational study; EAT BITE DSM-III criteria SQ-EDS SRQ	Medical students of both sexes. n=210 With a mean age of 18.2 years	A total of 28 students had scores above the thresholds set by EAT or BITE in the initial step. 14.8% of the participants were categorized as having EDS, a condition that did not meet the established diagnostic criteria for major ED. None of the participants had a syndromal AN or BN. 11% of participants received a diagnosis of EDS
Gupta et al., 2000, Bengaluru[14]	Cross-sectional observational study; EDI	n=65 Canadian women n=47 Indian women Age range: 17–24 years	The EDI scores did not differ significantly between the two groups. Preoccupations with abdomen and lower body weight were shared by both groups, albeit upper body weight was an additional concern in participants from India
Bhugra et al., 2000, Town in North India[15]	Cross-sectional observational study; BITE	n=504 female university students	BN had a point prevalence of 0.4%
Shroff & Thompson, 2004, Mumbai[16]	Cross-sectional observational study; EDI POTS SIAQ	n=96 children in the adolescent age group n=93 adult females	The body dissatisfaction subscale of EDI and internalization predicted the drive for thinness in adults, whereas teasing was an important risk factor for adolescents
Lal & Abraham, 2010, Delhi[17]	Translation of the QOL-ED into Hindi	n=95 female students between 14 and 37 years, ranging from low to high socioeconomic status	After controlling for age, the Hindi translation of QOL-ED scores did not differ significantly from the English version
Lal & Abraham, 2011, Delhi[18]	Cross-sectional observational study; QOL-ED Hindi version	n=461 female school students with the age range of 12–17 years	Adolescent development measures were correlated with the psychological feelings subscores
Balhara et al., 2012, Delhi[19]	Cross-sectional observational study; EAT-26, BSQ	n=97 female nursing students with a mean age of 20.2 years	About 4% (n=3) had scored greater than the EAT-26 cutoffs, indicating the presence of abnormal eating attitudes. BSQ scores correlated with BMI
Chellappa & Karunanidhi, 2013, Chennai[20]	Cross-sectional observational study; EAT-26 BDI State-Trait Anxiety Inventory	n=200 undergraduate female students with the mean age of 19 years	EAT-26 scores were more than 20 in 30% of students, suggesting a risk of EDs. These individuals had greater scores on the BDI and State-Trait Anxiety Inventory
Placek, 2013, Tiruvannamalai[21]	Cross-sectional observational study; informal and semi-structured interviews	n=54 nonpregnant women (19–80 years) n=95 pregnant women (19–35 years)	The most common substances consumed were raw rice, ash, and toothpaste. The presence of pica was associated with hunger, inadequate access to food, and the presence of mental health problems such as depression and anxiety
Jugale et al., 2014, Bengaluru[22]	Cross-sectional observational study; SCOFF	n=117 women between 20 and 25 years of age	About 42.7% screened positive for suspected EDs as measured by SCOFF. These participants also had greater occurrences of dental comorbidity, including caries, perimyolysis, and dental hypersensitivity
Upadhyah et al., 2014, Meerut[23]	Cross-sectional observational study; EAT-26	n=120 female school students aged between 13 and 17	26.67% of the participants scored more than the EAT-26 cutoff scores. EAT-26 scores were associated with restricted eating patterns, perceived poor social support, societal pressures, stressors, mood variability, and perfectionistic traits
Palmer, 2014, Online[24]	Cross-sectional observational study; SCOFF, EDI	n=42 individuals responded	About 69.04% had suspected ED as measured by the SCOFF questionnaire
Anandakrishna et al., 2014, Bengaluru[25]	Cross-sectional observational study; decayed, missing, filled surfaces index CEBQ	n=250 46- to 71-month-old children	The children who consumed more when they had no other activities had higher proportions of ECC at 52.6%
Ramaiah, 2015, Bellur[26]	Cross-sectional observational study; EAT-26 BSQ	n=172 medical students, with a mean age of 21 years	The study found that 16.9% of the participants demonstrated abnormal eating attitudes. EAT-26 scores correlated with BSQ
Karkare & Purwar, 2015, Multicentric[27]	Cross-sectional observational study; EDI Hindi version of Eysenck’s PEN inventory	n=1000 women with a mean age of 33.4 years.	Highly neurotic women had significantly more EDs than low neurotic women, while low neurotic women had significantly fewer EDs than those with moderate neuroticism
Mannat et al., 2016, Udupi[28]	Cross-sectional observational study; EAT-26	n=550 preuniversity students in the age range of 15–19 years	Nearly 31.09% (n=171) had an affinity for developing ED
Placek, 2017, Tiruvannamalai[29]	Cross-sectional observational study; Informal and semi-structured interviews	n=54 nonpregnant women (19–80 years) n=95 pregnant women (19–35 years)	85 of the 95 pregnant reported pica. Culture is an important factor in shaping the presentations of pica. Pica is reported to occur in children and nonpregnant women as well
Shashank et al., 2016, Mandya[30]	Cross-sectional observational study; EAT-26 SCOFF	n=134 medical students with a mean age of 21.4 years	EAT-26 scores indicated suspected ED in 29.2% of participants. SCOFF scores were above cutoff scores in 17.2% of students
Gupta et al., 2017, Chandigarh[31]	Cross-sectional observational study; Hindi version of EAT-26 BSQ	n=250 medical students 18–28 years of age range	Higher scores on the BSQ predicted ED. Females had higher scores on the dieting subscale of EAT-26
Vijayalakshmi et al., 2017, Bengaluru[32]	Cross-sectional observational study EAT-26 SCOFF	n=241 medical students n=213 nursing students  Participants with an age range of 18–29 both genders	In both EAT-26 and SCOFF, a greater proportion of men scored above the cutoffs than in women. Binge eating was more common in women, and overexercising (more than 60 minutes every day) was more common in men
Vijayalakshmi et al., 2018, Bengaluru[33]	Cross-sectional observational study; EAT-26 SCOFF	n=454 participants with an age range of 18–29 of both genders.	Variable proportions of participants were identified as at-risk for EDs based on SCOFF (34.1%) and EAT-26 (10.4%). Age (18–20), female sex, and Hindu religion from a semi-urban/urban background were identified as risk factors
Nivedita et al., 2018, Mysore[34]	Cross-sectional observational study; EAT-26 BES	n=1600 students, 16–25 years old	About 12.3% and 13.7% of the participants reported binge eating distress and abnormal eating attitudes, respectively
Shenoy & Praharaj, 2019, Udupi[35]	Cross-sectional observational study; McLean Screening Instrument for borderline personality disorder, BEDS-7	n=500 students with the mean age of 19.9 years.	BED was more common in individuals with borderline personality disorder
Traugott et al., 2019, Kangara[36]	Cross-sectional a qualitative study; semi-structured questionnaires	n=27 women who engage in geophagy of age range from 18 to 80 years	People typically engage in geophagy due to an intense craving for soil despite concerns about its potential health risks. This practice is often secretive and rarely disclosed to local doctors. While some individuals report feeling relief as a positive outcome, geophagy is generally seen as detrimental to health associated with various complaints
Sharma et al., 2019, Delhi[37]	Cross-sectional observational study; EAT-26 SATAQ-3	n=370 students age range of 17–30 years	About 21.1% exhibited abnormal eating attitudes. Media were identified as an influencing source for perceived attractiveness in 42.2% of the participants, and 27.6% felt pressured to conform to societal body standards, according to SATAQ-3
Purwar & Karkare, 2019, Multicentric[38]	Cross-sectional observational study; EDI	n=1000 Indian women with a mean age of 33.4 years	Mean ED scores were highest in women from South India, followed by women from North, West, Central, and East India
Purwar & Karkare, 2019, Multicentric[39]	Cross-sectional observational study; Socioeconomic status inventory EDI	n=1000 Indian women with a mean age of 33.4 years	ED in women of high socioeconomic status was significantly higher than those of low and middle socioeconomic status
Kumar, 2020, Sangrur[40]	Cross-sectional observational study; Self-administered structured knowledge questionnaire	n=50 adolescent girls 14–17 years of age	Knowledge about EDs improved after the intervention, with 82% having average scores and 18% having good scores
Dikshit et al., 2020, Mumbai[41]	Cross-sectional observational study; BES and Eating Behaviors and Pattern Questionnaire	n=2000 school students of both sexes of the age range of 12–18 years	Female sex and upper-class families had greater scores on binge eating measures. About 87% of the participants reported significant binge eating behaviors of moderate-to-severe intensity
Thangaraju et al., 2020, Madurai[42]	Cross-sectional observational study; EDE-Q BSQ	n=199 female medical students with a mean age of 20.4 years	Mean BMI score of 23.78. According to the EDE-Q assessment, 13.6% of students scored above 5, indicating the presence of an ED. Regarding body shape concerns, 61.9% had mild-to-moderate concerns, and 9% had severe concerns. There was a positive correlation between BMI and body shape concerns and the subscales of ED
Kulshreshtha M et al., 2020, Delhi[43]	Cross-sectional observational study; EAT-26, BSQ	n=441 adult women ages 18–45 year	Dancers endorsed more significant dieting behaviors and binge eating episodes than their counterparts. EAT-26 score correlated with BSQ scores
Lewis-Smith et al., 2021, Delhi[44]	Cross-sectional observational study; Validation of EDE-Q	n=1,465 urban English-speaking adolescents aged 11–15 years	EDE-Q demonstrated reliability and validity among this population
Singh & Gadiraju, 2020, Hyderabad[45]	Cross-sectional observational study; EAT-26 SATAQ-3 BSQ	n=262, including both genders aged 18–25 years	The study’s findings revealed several significant correlations and predictors. BMI was predicted by SATAQ-3, EAT-26 scores, and BSQ. SATAQ-3, BSQ, and EAT-26 scores were significantly correlated
Charak et al., 2021, Patiala[46]	Cross-sectional observational study; CEBQ WHO Oral Health Assessment Form 2013	n=2000 children of 36- to 71-month-old children	The prevalence of ECC in the study sample was 42.6%. CEBQ was used to measure eating behaviors, which were problematic and significantly associated with ECC
Iyer & Shriraam, 2021, Chennai[47]	Cross-sectional observational study; EAT-26, BSQ, and Perceived Stress scale	n=332 students in the age range of 18–21 years	Around 13% of students had a high risk of EDs. High-risk status correlated significantly with factors such as elevated stress, severe body image concerns, a history of counseling, peer pressure, excessive exercise, or using laxatives and diet pills
Nagarale et al., 2021, Western India[48]	Cross-sectional observational study; questionnaire consisting of 40 questions about knowledge about EDs and their effects	n=374 students 18–25 years of age range	43.9% of the participants knew about various EDs. Approximately 35.6% of the participants were aware of oral effects, and 36.4% knew about systemic ED symptoms
Raval et al., 2022, Gujarat[49]	Cross-sectional observational study; EAT-26, and BITE Structured clinical interview as per DSM-5 criteria	n=790 college students in the age range of 17–35 years	The prevalence rates for abnormal eating attitudes, BN, and OSFED were 25.2%, 0.2%, and 0.6%, respectively. None of the participants had syndromal AN.
Goswami et al., 2022, Delhi[50]	Cross-sectional observational study; Self-administered questionnaire	n=650 children aged 10–18 years	The likelihood of developing EDs and engaging in practices associated with BN and AN was higher among participants in the age range of 15–18 years compared with 10–14 years
Amirapu & Brady-Van den Bos, 2022, Online[51]	Cross-sectional observational online study; Done on Microsoft teams using semi-structured questionnaires	n=10 female students in the age range of 20–23 years	Five themes emerged: 1. Cultural and familial factors shaped eating and exercise habits 2. Living conditions during the lockdown intensified these influences 3. Participants’ thoughts and emotions were significantly affected 4. The effects persisted in their relationships with parents post-lockdown 5. The experiences raised awareness and motivation to address body image and mental health issues
Kalpana & Khanna, 2022, Faridabad[52]	Cross-sectional observational study; EAT-26 DASS	n=52 kho kho players age 16–31 years	Among the players, 11.5% exhibited disordered eating attitudes. Depression scores, as measured by DASS, were elevated in this group.
Lewis-Smith et al., 2023, Delhi[53]	Cross-sectional observational study; BES Internalization-General Subscale of the SATAQ-3	568 school children of both sexes aged 11–14 years	After the intervention, it was observed that internalization constructs, life disengagement, abnormal eating attitudes, self-esteem, and negative affect had significant results, with sustained improvement at a 3-month follow-up

AN: anorexia nervosa; BDI: Becks Depression Inventory; BED: binge eating disorder; BES: Body Esteem Scale; BITE: Bulimia Investigatory Test; BSQ: Body Shape Questionnaire; BMI: body mass index; BN: bulimia nervosa; CEBQ: Children’s Eating Behavior Questionnaire; DASS: Depression, Anxiety and Stress Scale; DSM-III : Diagnostic and Statistical Manual of Mental Disorders Version III; DSM-5: Diagnostic and Statistical Manual of Mental Disorders Version 5; ED: eating disorder; EAT: Eating Attitudes Test; EDI: Eating Disorder Inventory; EDS: eating distress syndrome; EDE-Q: Eating Disorder Examination Questionnaire; ICD-10: International Classification of Diseases; POTS: Perception of Teasing Scale; QOL-EDs: Quality of Life for Eating Disorders Questionnaire; SATAQ-3: Socio-Cultural Attitude Towards Appearance Questionnaire 3; SCOFF: Sick, Control, One-Stone (14 lbs/6.5 kg), Fat, Food; SQ-EDS: Screening Questionnaire for Eating Distress Syndrome; SIAQ: Socio-Cultural Internalization of Appearance Questionnaire; SRQ: Self-Report Questionnaire

Table 2.

Original studies conducted in clinical settings

Author, year, and place	Study design; assessment methods	Sample size and characteristics	Major findings
Gogte et al., 1991, Delhi[54]	Cross-sectional observational study; Dithizone method of estimation of lead	n=253 children of both sexes and age range of 0.2 to 15 years	Out of the total, 88 children had pica, with an average blood lead level of 23.0 micrograms/dl, notably higher than that of the control group
Singhi et al., 2003, Chandigarh[55]	Case-control study; Atomic absorption spectrophotometer was used to estimate levels of iron, zinc, calcium, magnesium, lead	n=31 children with pica N=60 controls 18–48 months age range	Plasma iron and zinc levels in children with pica were lower than in controls. The zinc and iron levels did not correlate with age at onset, duration, and frequency of pica, and number of inedible objects ingested
Mammen et al., 2007, Vellore[56]	Retrospective chart review	Medical records of n=3274 children Mean (SD) age of patients with ED was 12.6 (3,4) years	A total of 41 cases of ED were detected in the study. Approximately 1.25% of them had an ED. Additionally, in 14.6% of these cases, six individuals were diagnosed with AN. About 44% of the individuals with ED also had a comorbid psychiatric condition
Boatin et al., 2012, Chhattisgarh[57]	Cross-sectional observational study; Semi-structured pro forma and blood for hemoglobin estimation	n=2386 pregnant women	About 27% of the participants consumed chalk, mud pot, or raw rice during pregnancy. Moderate-to-severe anemia was significantly higher among those with pica
Basker et al., 2013, Vellore[58]	Retrospective observational study	n=7 (3 boys and 4 girls of age range 1 of 4–16 years)	Five adolescents underwent treatment involving nutritional rehabilitation and family-based therapy during their inpatient stay, which lasted approximately 3 weeks. Among these individuals, four patients had robust family support and successfully achieved recovery. However, one of them exhibited minimal weight gain during the process, and unfortunately, two patients were lost to follow-up.
Kurpad et al., 2010, Bengaluru[59]	Cross-sectional observational study; Eating Behavior Questionnaire DSM-IV criteria	n=73 patients diagnosed with psychosis in age range of 17 to 65 years	The majority of binge spectrum behaviors were observed in patients who had been undergoing treatment for over 2 years and were concurrently taking antidepressant medication. BED was not present in this patient group
Crasta et al., 2014, Vellore[60]	Prospective cross-sectional observational study; BAMBI, Sensory Profile Questionnaire, CARS, Binet-Kamat Scale of Intelligence, Gesell’s Developmental Schedule	n=41 children with autism and N=56 with ID (3–10 years of age group)	FP was 61% among children with autism and 46.4% among those with intellectual disabilities. Among children with autism, the feeding profile tended to be more severe
Garg, 2015, Multiple sites in India[61]	Prospective, cross-sectional observational study; Assessment of the utility of IMFeD	n=383 children of 2–10 years of age range	124 children, making up 33% of the study group, experienced multiple FP. This demonstrates the effectiveness of the IMFeD tool in identifying FP in Indian children
Lal et al., 2015, Delhi[62]	Prospective, cross-sectional observational study with comparator; QOL-EDS DSM-IV	n=30 Indian patients n=60 Australian patients Age range of 16 to 26 years	The difference in the global ED-QOL score was not statistically significant between Indian and Australian patients. Indians had higher beliefs of overeating frequency but similar frequencies in food restriction, vomiting, and laxative use. They had fewer beliefs related to fear of losing control over food intake and preoccupations with body or food compared with Australians
Malhi, 2017, Chandigarh[63]	Prospective case-control study; CEBI, 3-day food records, anthropometric measures	63 children with ASD and 50 typically developing children (4–10 years of age range)	79% of parents of children with the diagnosis of ASD expressed FPs when compared to 64% in the controls. ASD children had significantly higher CEBI scores and more FPs than controls. ASD children consumed fewer food items, mainly fruits, vegetables, and proteins, and had lower daily potassium, copper, and folate intake.
Jacob et al., 2018, Bengaluru[64]	Retrospective chart review; 10-year chart review of children and adolescents who attended outpatient at a tertiary care center	19,151 patients attended the outpatient services. Twelve patients were diagnosed with ED as per ICD-10, with a mean age (SD) of 14.42 (1.08)	Prevalence rate of ED was 0.063%. EDs were more commonly diagnosed in female children. About 83.3% of the children had a psychiatric comorbidity
Siddiqi et al., 2019, Mysuru[65]	Cross-sectional study; 3-day food records, FFQ, CEBI	n=53 (45 boys and eight girls) in the age group of 2–13 years	The study found that children with ASD had reduced consumption of fruits and vegetables, leading to lower levels of essential micronutrients, particularly B-Complex, calcium, and iron (P≤0.05). Nutritional intervention programs for children with ASD are crucial
Malik et al., 2020, Lucknow[66]	Longitudinal observational study; Physician-administered questionnaire and ROME III checklist	n=50 children 5–18 years of age range	Among children diagnosed with rumination syndrome, 40% experienced a relapsing and remitting course, while 60% had chronic and persistent symptoms. Before receiving the correct diagnosis and appropriate treatment, the majority of these children, specifically 87%, received incorrect diagnoses
Krishnamurthy et al., 2020, Delhi[67]	Cross-sectional observational study; EAT-26 and BES were used in two groups: group 1 consisted of adults with T2DM, and controls comprised of individuals without T2DM	n=256 in each group, the age range of 20–65 years was included	Among participants, 10.9% with T2DM and 14.1% of controls had significant EAT-26 scores. After detailed psychiatric assessments, the prevalence of syndromal ED was estimated at 0.8% in each group
Prasad et al., 2021, Vellore[68]	Retrospective chart review of adolescents diagnosed with AN	n=43 (12 children were males) mean (SD) age at presentation: 13.4 (1.7) years	About 76% of the patients required hospitalization. Among the 15 patients who had telephonic follow-up 1 to 5 years later, one had unfortunately passed away, but 11 had successfully attained a healthy weight for their respective ages
Francis, 2022, Chennai[69]	Cross-sectional observational survey to assess for pica	n=739 pregnant women with an age range of 20–34 years	The study reported that 29.7% of women engaged in pica behavior. Among the 220 women examined, 42.2% consumed food substances, 41% ingested nonfood substances, and 16.8% consumed both. Among the women practicing pica, it was found that 67.3% had anemia, 36% were underweight, and 9.5% were overweight
Sravanti et al., 2022, Bengaluru[70]	Retrospective chart review was conducted in a tertiary care hospital over 10 years	n=40 children (mean age (SD): 13.96 (2.3) years) were identified with a male-to-female ratio of 1:12	Prevalence of AN was estimated at 0.07%
Swarnameenaa et al., 2023, Chennai[71]	Cross-sectional study for translation and validation of Tamil version of EAT-26	n=150 patients with schizophrenia in the age range of 18–65 years	The Tamil version displayed good reliability, with an internal consistency coefficient of 0.71 and a strong test-retest reliability of 0.896

AN: anorexia nervosa; ASD: autism spectrum disorder; BED: binge eating disorder; BES: Binge Eating Scale; BAMBI: Brief Autism Meal-Time Behaviors Inventory; BMI: body mass index; CARS: Childhood Autism Rating Scale; CEBI: Children’s Eating Behaviour Inventory DSM-IV: Diagnostic and Statistical Manual of Mental Disorders Version IV; EAT: Eating Attitudes Test; ED: eating disorder; FP: feeding problems; FFQ: Food Frequency Questionnaire; ICD-10: International Classification of Diseases; ID: intellectual disability; IMFeD: Identification and Management of Feeding Difficulties for Children; OSFED: other specified feeding or eating disorders; QOL-EDs: Quality of Life for Eating Disorders Questionnaire; T2DM: type 2 diabetes

Clinic-based studies (N = 18) have mostly followed retrospective record-based or cross-sectional exploratory designs. These typically include data from registered patients in various clinics or tertiary care psychiatry services, which have been analyzed for evaluating the clinic prevalence of various FEDs. The study population for ED is typically adolescent students in the majority of the studies. Among studies evaluating FD, pica is the most commonly studied. The study population for RRD and ARFID comprises children and adolescents visiting pediatric or child and adolescent psychiatry (CAP) clinic services.[55,60,63] Four studies evaluated pica in pregnant women.[21,29,57,69] There has been a rise in case reports of pica. The earliest case report on pica is from 1971; 24 case reports have been identified since then. Two case reports have been identified in RRD.

Epidemiology, burden, and determinants of EDs

Incidence and prevalence

As a general observation, there is a shortage of epidemiological studies that are methodologically robust compared with the West in India.

Among community-based research in the context of ED, the prevalence rate of AN for males is estimated at 10/100,000, while for females, it was 37.2/100,000, and the combined sex burden was 22.3/100,000.[72] Another study conducted in Gujarat students, using screening tools and DSM-5-based clinical interviews, found that the prevalence rates for suspected EDs, BN, and OSFED were 25.2%, 0.2%, and 0.6%, respectively. None of the participants had syndromal AN.[49] As reported by a study conducted in 2000, BN had a point prevalence of 0.4% among N = 504 female school students, as estimated by the Bulimic Investigatory Test, Edinburgh (BITE) in North India.[16] A Mysore study evaluated “Binge eating distress” in N = 1600 students and reported findings of 12.37% in their study population.[35] The frequency of disordered eating in studies conducted in community settings ranged from 4 to 69.4% [Table 1]. The proportion of disordered eating attitudes with Sick, Control, One-stone (14 lbs/6.5 kg), Fat, Food (SCOFF) varied from 17.2% to 69.04%, and with Eating Attitudes Test (EAT)- 26 varied between 4% and 34.1% [Table 1].

Among hospital-based retrospective reviews, the earliest study conducted from 2000 to 2005 in Vellore, Tamil Nadu, reported a 6-year prevalence of 1.25% for ED. It identified six cases of AN, resulting in a 6-year prevalence of 0.18%.[56] A 10-year (2002–2012) study from a CAP hospital in Bengaluru reported a prevalence rate of ED as 0.063%. A total of N = 19,151 patients attended the outpatient services, 12 patients with a diagnosis of ED, nine of whom had AN.[64] Another study from the same center assessed case records of children diagnosed with ED from 2009 to 2019. AN prevalence in a tertiary care hospital was 0.07% over 10 years among 39,384 children.[70]

Regarding FD, a community-based epidemiological study of child psychiatric disorders in Bengaluru in Karnataka included rural and urban settings. It assessed the prevalence of mental disorders in children and adolescents using the standard 2-step method of screening followed by detailed assessment. The prevalence of pica in children 0–3 years of age group was determined to be 2% and 4–16 years of age group 0.2%.[73] The prevalence in the general adult population is unclear, although two community-based studies in pregnant women report the prevalence as 27–29.7%.[57,69]

Among clinic-based studies, a multicentric study conducted in N = 383 children estimated that 33% of Indian children presented with more than one feeding difficulty measured by the Identification and Management of Feeding Difficulties for Children (IMFeD) tool.[61] The occurrence is higher in children with autism and intellectual disabilities (IDs). This is supported by a 2014 study that identified feeding problems in 61% of children with autism and 46.45 in ID.[60] No studies have specifically diagnosed ARFID or FDs in infancy and childhood.

Time trends

AN’s prevalence rate was observed to rise from 1990 to 2016.[72] A clinic-based study from Bengaluru revealed a significant rise in the number of diagnosed cases of AN. The study observed that between 2009 and 2015, there were only five cases of AN, whereas from 2016 to 2019, this number increased to 21 cases, corresponding to a fourfold increase. The prevalence estimate has grown from 0.09% to 0.2% over a 6-year period.[70] Clinician observation supports a rise in ED prevalence in India. Psychiatrists’ encounters with AN were previously reported as rare, and a 2012 study found that two-thirds of psychiatrists (N = 66) surveyed in Bangalore, India, had seen a case of AN, BN, and ED unspecified within the past year. About 23.5% of psychiatrists responded that the frequency of EDs was rising, 26.5% felt the rates were unchanged, and 42% were unsure.[74]

There are no studies evaluating the time trends of FD. However, original research and case reports of pica have been consistently present through the past 20 years, although N = 18 of 24 case reports and 4 of the 6 original research have been published after 2010. Similarly, all reports of RRD and feeding problems in childhood and infancy are from the past decade of research.

Geographical trend

The majority of the original research on FED is from Karnataka (N = 14 studies), Tamil Nadu (N = 13 studies), and Delhi (N = 10 studies). Overall, there have been 27 studies from South India, N = 20 from North India, N = 4 from West India, and N = 2 from Central India. Two studies were conducted online, and four were multicentric. One study evaluated the variable magnitude of EDs across North, South, East, West, and Central India. Mean ED scores were highest in women in South India, followed by North, Western, Central, and East India.[38] Variations in the clinical manifestation of ED have been reported compared with the West, which has been further elucidated in the clinical manifestations section.

No studies exclusively looked into the pattern of distribution of ED in urban, semi-urban, or rural backgrounds. Five studies revealed a preponderance of ED and pica in urban populations.[22,33,70,73]

Age at presentation

Multiple studies in India have highlighted adolescents as the most susceptible age group for developing ED symptoms. For instance, in one Indian hospital file review involving 43 cases, the mean age at presentation was 13.4 years, with an average age of onset at 12.4 years.[68] The youngest patient recorded was just 10 years old. Other hospital reviews had similar reports of mean age of presentation of 13.96 years (SD 2.3) and 14.42 years (SD 1.08).[64,70] This contrasts with community-based studies where the at-risk age groups are described as late adolescents.[33,42] All three FDs are more often reported in children than adults in Indian research.[60,66,75,76,77,78,79]

Socioeconomic status

While there is a scarcity of longitudinal research investigating the influence of economic upbringing on ED, the available case reports consistently depict a common scenario in India. In particular, these reports often portray adolescent females from affluent or middle-class backgrounds. A study published in 2019 explicitly evaluated ED in Indian women based on their socioeconomic status. In this study, a sample of N = 1,000 Indian women, with an average age of 33.35 years, was chosen across different regions of India, including the northern, western, southern, central, and eastern parts. The selection of the sample was conducted using a random sampling method. The results revealed that ED in women of high socioeconomic status groups was significantly higher than in women of low and middle socioeconomic status.[39] A study conducted in Mumbai, India, involving N = 2000 students, discovered that a significant proportion of the participants, hailing from upper- and lower-middle-class households, exhibited pronounced tendencies toward binge eating.[41] In contrast, among FD, pica has often been reported in lower socioeconomic groups in several reports.[80,81]

Sex

Indian survey data consistently demonstrate a significant female predominance among individuals with EDs. These findings align with studies in clinical settings in AN reporting a male-to-female ratio of 1:5 to 1:12.[56,64,70]

Similarly, in a study from Mumbai, females reported higher binge eating than males.[41]

However, an interesting finding is that, in two studies in 2017 and 2019, males had greater scores on EAT-26 than females and media pressure to attain an ideal body.

Among feeding problems in children with autism, sex was not reported to be a risk factor.[60]

Religion

One potential explanation for the influence of religion on FED may be attributed to the prevalent fasting practices observed in both Hinduism and Islam, the two most widely practiced religions in India, which are poorly evaluated and have contradictory results. One study reported a significant association between Hindu religion and ED as measured by the SCOFF score; however, the same association was not significant as measured by the EAT-26 measure.[33] Few case reports indicate the same finding.[5] Religious diversity was evident in the 2022 Bengaluru hospital-based study, with 57.7% (N = 15) of participants adhering to Hinduism, 23.1% (N = 6) practicing Christianity, and 19.2% (N = 5) following Islam with no significant associations.

Family history

A significant history of psychiatric disorders was reported in 20.9% of families, and a family history of EDs or excessive preoccupation with health was reported in 18.6% of patients.[68] Similar findings were noted by Sravanti et al. (2022).[70] Both retrospective chart reviews from Bengaluru reported dysfunctional family dynamics as a significant trigger for AN.[64,70] The same was reflected in an online qualitative study conducted during the coronavirus disease 2019 (COVID-19) that revealed five critical themes. To summarize, cultural and familial factors shape eating and exercise habits; living conditions during lockdown intensified these influences; participants’ thoughts and emotions were significantly affected, and these effects persisted in their relationships with parents post-lockdown, and lastly, the experiences raised awareness and motivation to address body image and mental health issues.[51]

Sexuality

No data studying the effect of sexuality on FED have been reported in India.

Vocation

Kho kho players, certain regional dance groups, such as Kathak dancers in North India, and medical students have been identified as being at heightened risk in India.[33,42,43,52]

Miscellaneous determinants of FEDs

Other significant determinants reported as predictors of disordered eating in the Indian literature encompass the influence of Western media and culture, body mass index (BMI) in the overweight or obese range, experiencing more peer pressure, susceptibility of feeding patterns to mood, encountering stressors, and poor social support [Table 1]. There were several risk factors identified for pica. These included having fewer years of education, being involved in agricultural work, and belonging to historically disadvantaged caste groups. Additionally, women who chewed tobacco had a higher prevalence of pica compared with those who did not (46.5% vs. 24.6%, P < 0.0001). However, factors such as age, parity (number of children), home ownership, use of iron supplements, presence of malaria parasites in the blood, chronic diseases, and smoking were not found to be associated with pica.[57] Other reports on risk factors for pica include malnutrition, pregnancy, psychosis, ID, and obsessive-compulsive disorder (OCD).[82,83]

Neurobiology of FEDs

Although a family history of psychiatric and EDs is reported to be higher in children and adolescents with ED, no genetic studies have been conducted. From a neurobiological perspective, a 2018 study assessed the pharmacological impact of piracetam (administered at 200 mg/kg intraperitoneally) on BED in female rats. BED was induced by giving the rats free access to palatable cookies for 2 hours on alternate days. The study revealed several significant findings: increased levels of plasma corticosterone and glutamate in the nucleus accumbens (NAC), hypothalamus (HYP), and prefrontal cortex (PFC), suggesting stress and excitotoxicity; elevated levels of dopamine in NAC and PFC, and lower levels in HYP, potentially contributing to increased motivation for palatable feeding and cognitive deficits; and changes in feeding behavior-regulating hormones: an increase in leptin levels and a decrease in ghrelin levels in BED. Notably, piracetam administration resulted in a significant reduction in binge eating behavior, accompanied by a decrease in associated body weight. Additionally, piracetam helped regulate the levels of neurotransmitters in the relevant brain regions.[84]

Clinical presentations

Type of presentation

Most children reported an insidious onset with a duration greater than 1 year. The typical presentation of AN often involves adolescent females complaining of gastric issues such as loss of appetite or abdominal pain. Common triggers include peer pressure, bullying, a family history of EDs, or the influence of health-conscious adults. Menstrual irregularities and underdeveloped secondary sexual characteristics are frequently observed. Notably, the restrictive pattern of AN is more frequently reported than the binge-purge type.[68,70] It is worth noting that, in contrast to Western studies, patients in India do not exhibit body image disturbances, a strong drive for thinness, or a fear of becoming fat. They may be unconcerned about their weight loss, even rationalizing their reduced food intake with due explanations.[85,86] This phenomenon is known as nonfat phobic AN (NFP-AN). Further studies are needed to evaluate this variant as there are contradictory reports in 2016, where 100% (N = 12) patients reported a restrictive diet and 91.7% (N = 11) reported a dread of fatness.[5,64]

A few BN case reports in India have described typical and atypical variants. Some individuals with BN have been reported to use isabgol husk, diuretics, laxatives, and orlistat.[87,88,89,90,91] Additionally, there is a category called eating distress syndrome (EDS), which represents subsyndromal forms of AN or BN. EDS is characterized by distressing thoughts about body shape and eating habits, including strict dieting and, in some cases, bingeing. However, it does not typically involve significant weight loss or extreme weight loss measures such as diet pills, starvation, purging, or vomiting[5,13]

Of the cases of pica described in Indians, most are associated with one or more risk factors.[57,80,92] There are two kinds of presentations reported: one where the individual directly presents to psychiatry with complaints of having irresistible urges to consume nonfood items, and another usually presents as a complication of pica, and pica is diagnosed on specifically clarifying the history.[75,93,94] The second group commonly presents with diarrhea, abdominal pain, loss of appetite, and dental problems. There are reports of consumption of chalk, brick, clay, mud, pebbles, paper, uncooked rice or wheat, erasers, and burnt matchsticks.[76,77,92,93,94] The onset of pica is usually reported as insidious and gradually progressive over one or more years of duration and is often associated with the full knowledge of the oddity and is done in secrecy.[83,92,94] There is one report of an abrupt onset of sand pica following tramadol use within 3 days that completely recovered on discontinuation of tramadol.[95]

Presentations of RRD are reported in detail by a prospective study in Lucknow. While initially complaining of “vomiting,” these children experienced painless postprandial regurgitation. They regurgitated after around three meals, with vomiting starting soon after eating and lasting up to 15 minutes, unrelated to specific foods. Associated symptoms were observed in the following proportions: nausea in 30%, constipation in 23.3%, weight loss in 16.6%, anorexia in 2%, and abdominal pain in 23.3%. Additionally, 10% of the children met the Rome III criteria for functional dyspepsia. About 40% of the children reported specific stressors around symptom onset. Notably, 93.3% sensed gastric contents rising to their mouths, and 70% decided what to do with the food bolus once it reached their oral cavity.[66]

In this study and the two other case reports concerning RRD, there was a significant delay in diagnosis, several failed drug trials, and a surprisingly good response to supportive therapy.[66,78,96]

Studies regarding presentations specific to ARFID are lacking in India. Disruptive meal-time behaviors and food over-selectivity were more common in children with autism than complete food refusal. This often leads to several micronutrient deficiencies, including B-complex, iron, copper, vitamin C, and calcium.[60,65]

Comorbidities and complications

Regarding comorbidities, depression, OCD, anxious-avoidant traits, and anankastic traits are commonly associated with EDs.[58,68,70,97] In a medical chart review in Vellore, India, 44% of diagnosed ED cases had psychiatric comorbidities. Most individuals had one psychiatric comorbidity, while a smaller percentage had two or three comorbid diagnoses. Mood disorders were the most common comorbidity, followed by ID and dissociative disorders. Other comorbidities included substance abuse, sleep disorders, elimination disorders, adjustment disorders, oppositional defiant disorder, and specific learning disorders (SLDs).[56]

EDs can also co-occur with physical illnesses such as systemic lupus erythematosus, occur secondary to traumatic brain injury, or result from adverse drug reactions, as seen with zolpidem consumption (referred to as nocturnal sleep-related ED). Furthermore, AN can mask physical illnesses such as carcinoma. Medical complications associated with EDs include microcardia, sinus bradycardia, QT prolongation, T-wave changes, electrolyte abnormalities, anemia, hemodynamic instability, and refeeding syndrome. Less common complications involve low bone mineral density, cerebral atrophy, pituitary changes, and Wernicke-Korsakoff syndrome.[5,64,68,70] In the context of BED, there appears to be a higher prevalence among individuals with borderline personality disorder.[35]

Furthermore, EDs are reported to be associated with higher self-injurious behaviors. A study conducted in Bengaluru reported that approximately half of the sample, around 50% (n = 12), had at least one comorbidity, and 20% (N = 5) had two or more coexisting conditions. Among these, major depressive disorder was the most prevalent comorbidity, affecting 40% (N = 10) of the sample, followed by OCD (N = 5) and anxiety (N = 4). Additionally, at baseline, 30.8% (N = 8) of the sample reported experiencing suicidal ideation with the intent to die.[70] An earlier study in the same setting reported that 58.3% had suicidal ideations, and 33% had a history of suicide attempts.[64]

In pica, common comorbidities are iron deficiency anemia, psychosis, OCD, and IDD. Complications of pica can be malabsorption syndromes, dental attrition, loss of teeth, lead toxicity, iron and zinc deficiencies, electrolyte abnormalities, parasitic infections, intestinal and urinary obstruction, and bezoars.[79,93,98,99] In RRD, 16.6% had a diagnosis of other conditions at presentation, including anxiety disorder, SLD, mild IDD, facial tic disorder, and operated unilateral pelviureteric junction obstruction.[66]

Course and outcome

There is a shortage of controlled observations and long-term follow-up data in FED in India. Furthermore, though most patients resumed their regular diets, weight gain was typically modest.[85] As measured by the Clinical Global Impressions-Improvement (CGI-I) scale, a 2022 study from Bengaluru reported that upon discharge, 30% (N = 6) of the patients exhibited significant improvement, while 50% (N = 10) showed minimal improvement. Moreover, the patients receiving outpatient-based treatment at 4 weeks were categorized as “much improved” on the CGI-I scale.[70]

In the Vellore hospital-based study, of the 15 patients followed up 1–5 years later, one had died, and 11 had achieved normal weight for age.[68]

About RRD, the median duration of symptoms was 11 months (ranging from 2 to 72 months), and two distinct clinical patterns emerged. In the first pattern (n = 18), symptoms persisted continuously from onset until presentation without any asymptomatic periods. In the second pattern (n = 12), a relapsing and remitting course was observed. Among the symptomatic children, 80% experienced daily symptoms, 13.4% had symptoms at least three times a week, and only 6.6% had symptoms less than three times a week. Overall, resolution after treatment was seen in 26 (87%) with a relapse in 8 (27%) children.[66]

Assessment tools

Measures

In clinical and research settings, besides having standardized diagnostic criteria, there is a need for objective methods to measure the presence and severity of specific symptoms of FED. For this purpose, several rating scales for ED have been devised worldwide, of which for screening purposes, Eating Disorder Examination Questionnaire (EDE-Q), SCOFF, EAT-26, Bulimia Test (BULIT), and Bulimia Investigatory Test, Edinburgh (BITE) are popularly used worldwide and in India.[100]

For diagnosis by structured clinical interviews, Eating Disorder Assessment for DSM-5 (EDA-5), Structured Interview for Anorexic and Bulimic Syndromes (SIAB-EX), Eating Pathology Symptoms Inventory-Clinician Rated Version (EPSI-CRV), Eating Disorder Examination (EDE), Questionnaire for Eating Disorder Diagnoses (QEDD), and Eating Disorder Diagnostic Scale (EDDS) are used globally. There is a dearth of Indian studies using structured clinical interviews for diagnostic purposes.[100]

Binge Eating Scale (BES), Eating Disorder Inventory (EDI), Binge Scale Questionnaire, Body Shape Questionnaire (BSQ), Socio-Cultural Attitudes Towards Appearance Scale-3 (SATAQ-3), and Eating Disorders Quality of Life Scale ((EDQOL) are some of the other commonly used items to capture different dimensions of symptoms of ED and monitor treatment progress. Quality of Life for Eating Disorders Questionnaire (QOL-EDs), Screening Questionnaire for Eating Distress Syndrome (SQ-EDS), and Socio-Cultural Internalization of Appearance Questionnaire (SIAQ) have been used in studies by Indian researchers [Table 1].[100]

Scales such as EAT-26, QOL-ED, and EDE-Q have been validated in the Indian population. EAT-26 has been translated and used in Hindi, Kannada, and Tamil.[12,28,71] EAT-26, Hindi translation, also has 26 items; however, its rating format was changed to a “forced choice” (no/yes = 1/2) type. Hence, the total score ranges from 26 to 52 and does not have a specified cutoff. QOL-ED and BSQ have been validated in Hindi.[17]

Children’s Eating Behavior Inventory (CEBI), Children’s Eating Behavior Questionnaire (CEBQ), 3-day food records, and IMFeD have been used to identify feeding problems in children [Table 1]. No other specific tools were used in Indian research for FD.

Investigations

Among the adolescents studied by Prasad et al. (2021), 13.9% displayed electrocardiogram (ECG) changes, including sinus bradycardia, QT prolongation, and T-wave changes. Of the five adolescents who underwent echocardiography, all had normal results. However, when seven adolescents underwent brain magnetic resonance imaging (MRI), abnormal findings were present in 71.4% of them. These abnormalities encompass cerebral atrophy, white matter volume loss, periventricular hyperintensities, and pituitary changes. Lastly, bone mineral density tests were conducted in four patients, and half (50%) had low mineral density. These findings suggest a range of cardiac and neurological issues in this group of adolescents, highlighting the need for thorough medical evaluation and care. It is reported that 66.7% (eight individuals) of individuals with an ED have medical comorbidities, with anemia being the most common. Hypotension, bradycardia, hypoglycemia, tetany, and hypothermia are also reported. This highlights the importance of addressing medical comorbidities in child psychiatry and mental health care.[64,68]

The 2003 Chandigarh case-control study compared 31 children with pica and 60 controls matched for age, sex, and nutrition. Plasma iron level in children with pica was about 20% lower than in controls. Plasma zinc levels in the pica group were about 45% lower than in controls. The correlation of zinc and iron levels with pica-related variables such as age at onset, duration and frequency, and number of inedible objects ingested was insignificant.[55] Another study found elevated levels of lead in children who had pica. Children with pica were also significantly more anemic than the controls and showed a higher prevalence of abdominal-neurological symptoms.[54]

Treatment modalities

Pharmacological

Much of the available literature from India has predominantly focused on AN. Pharmacological treatments have included selective serotonin reuptake inhibitors (SSRIs) such as citalopram and fluoxetine (20 mg) and other antidepressants such as mirtazapine and trazodone.[58,101,102] Additionally, various agents have been experimented with to varying degrees of success, including antihistaminic drugs such as cyproheptadine and antipsychotics such as olanzapine and chlorpromazine.[5] Pharmacological intervention played a significant role in the clinician’s treatment approach in a tertiary care hospital in Bengaluru, with SSRIs being the most frequently prescribed medication. SSRIs were administered as monotherapy in 47.6% (N = 10) of cases and in combination with an antipsychotic in 38.1% (N = 8) of the sample.[70]

For individuals with AN and coexisting obsessive traits or OCD, a combination of olanzapine and fluoxetine or olanzapine and fluvoxamine has proven beneficial. As for BN, available reports suggest positive responses to sertraline and fluoxetine (20–80 mg per day).[39] Reported treatments of pica comprise correction of iron deficiency anemia, SSRIs, and antipsychotics. Fluoxetine 40 mg, escitalopram 10 mg, sertraline 75 mg, and fluvoxamine (dose not mentioned) have been utilized with good improvement.[77,94,94,103]

Non-pharmacological approaches

Several reports from India have mentioned using non-pharmacological techniques, such as cognitive behavioral therapy (CBT), supportive psychotherapy, hypnotherapy, and play therapy, often in combination with other treatment modalities for EDs.[8] However, these reports typically lack comprehensive descriptions of the therapy’s structure and components. As these observations were uncontrolled, it became challenging to determine the actual effects of these therapies and the best practice elements involved. Sravanti et al. reported that of the 24 patients included in their study, 21 children (80.8%) required inpatient care, while one patient (3.8%) was managed on an outpatient basis. The average duration of inpatient stays was approximately 30.7 days, ranging from 3 days to 74 days, with a standard deviation of 20.4 days. Inpatient care was delivered by a multidisciplinary team comprising psychiatrists, clinical psychologists, psychiatric social workers, ward nursing staff, nutritionists, and experts from rehabilitation services. For patients needing additional management of medical comorbidities, appropriate collaboration with pediatricians and physicians was pursued. Nearly all patients received individual psychotherapy sessions based on CBT. Dysfunctional family dynamics were identified as a significant concern in most cases. Supportive therapy sessions were administered to all the families by the psychiatric social worker, and 47.6% received structured intensive family therapy.[70]

Existing reports lack detail and conciseness when describing the refeeding protocols used. The components mentioned include psychoeducation regarding a balanced diet, malnutrition, and ideal body weight. Additionally, there is mention of nutritional supplementation, using Ryle’s tube, and, when necessary, parenteral nutrition. The pairing of higher-caloric diets with positive reinforcement is also discussed, but the available details are insufficient for replicability.[94,101,102] An Indian review on nutritional rehabilitation outlined the treatment’s principles, goal-setting process, and potential challenges. The authors emphasized the need to attain a normal weight range before discharge for better outcomes and to reduce the likelihood of rehospitalization.[104]

A community-based interventional study in 2023 in Delhi assessed the effectiveness of an intervention to improve body dissatisfaction for adolescents in India. This randomized controlled trial involved 568 school-going children of both sexes. The intervention comprised five 45-minute sessions led by trained psychologists. Outcomes, including body image measures, were assessed before, immediately after the sessions, and at a 3-month follow-up. The analysis revealed improved body image gains, which were sustained at 3-month follow-up in comparison with controls. Notably, there were significant improvements in eating attitudes, self-confidence, mood, and internalization constructs at the post-intervention stage, with effects sustained at the 3-month follow-up, albeit with some sex-specific variations.[53]

Another study in 2020 in Sangrur employed a quasi-experimental research design involving one group, before-and-after testing, with a randomly selected sample of 50 students. A structured self-administered questionnaire was used to evaluate the student’s knowledge, followed by a structured teaching program. In the pretest, 72% of adolescent girls had poor knowledge scores, and only 28% had average scores. None had a good knowledge score on EDs. However, after the intervention, most adolescent girls (82%) achieved an average knowledge score, and 18% had a good knowledge score. No one scored poorly in the posttest.[40]

Psychotherapy for RRD comprised of diaphragmatic breathing, and habit reversal training is reported to be effective.[66,78] Similarly, pica improvement with psychotherapy has been reported, although specific components are not mentioned.[94]

Neuromodulation treatment

One of the earliest documented cases of AN in India, dating back to 1966, involved treating a middle-aged adult woman. Her treatment regimen consisted of a combination of antipsychotics (100 mg chlorpromazine), nutraceuticals (liver extract, vitamin C), modified insulin therapy, and nine sessions of electroconvulsive therapy. Notably, improvement was reported after 6 weeks of this treatment. Clinicians have successfully used add-on high-frequency repetitive transcranial magnetic stimulation (rTMS) to treat a young adult girl with AN who responded poorly to a combination of antidepressants, antipsychotics, and cognitive behavior therapy.[8] The stimulation was applied to the dorsolateral PFC over 3 weeks, improving body weight and attitudes toward eating, weight, and body shape. Another report of successful application of add-on transcranial direct current stimulation (tDCS) in a patient with schizophrenia reduced their food cravings, contributing to the reversal of weight gain, making this a promising treatment to manage cravings. The patient tolerated tDCS without adverse effects, as ascertained by a structured questionnaire after each session.[105]

FED in special populations

Pregnant women

Women with EDs face unique challenges, with some developing these disorders during pregnancy. While “pregorexia” lacks formal classification, these issues persist postpartum. Symptoms can improve or worsen during pregnancy. Breastfeeding can be challenging for mothers with EDs, affecting both mother and child emotionally. Various factors, such as maternal age and prematurity, may influence the development of EDs in offspring.[106]

Similarly, there is an overrepresentation of pica in pregnant women. Approximately one-quarter of pregnant women admitted to consuming chalk, mud pot, or raw rice during pregnancy. The prevalence of moderate-to-severe anemia (Hgb <9 g/dL) was higher among women with pica (30.9% vs. 20.5%, P < 0.0001).[57] Other studies in Tamil Nadu support the same. Pregnancy in South India is marked as a period of increased “acai,” or desire for unusual foods, nonfoods, and material items.[21,29] Collaborative efforts between specialists in various fields are crucial for pregnant women with FED.

DISCUSSION

This narrative review attempted to collate the significant findings of FED research conducted in India. Historically, case reports formed the majority of the Indian FED research landscape. There has been a noticeable uptick in original research articles over the past decade. The worldwide lifetime prevalence rates for AN are estimated to be as high as 4% among females and 0.3% among males. In the case of BN, the rates are reported to be up to 3% among females and more than 1% among males during their lifetime. DSM-5 BED is estimated to impact approximately 1.5% of women and 0.3% of men globally.[107] Additionally, gender and sexual minority individuals are at a particularly high risk of these disorders.[108] The prevalence rates for suspected EDs, BN, and OSFED were 25.2%, 0.2%, and 0.6%, respectively.[49] Further Indian research is warranted to estimate the prevalence of BED. AN prevalence in a hospital-based population was estimated to be 0.07% over the 10-year period.[70] The recent Indian National Survey of Mental Disorders (2016) does not report the prevalence of the ED.[109]

No general population-based studies exist, resulting in a scarcity of representative data. The global standard involves a two-step assessment, beginning with self-rated questionnaires and followed by structured diagnostic interviews. Most epidemiological research in India focuses on student subpopulation and uses self-report scales without the second step of structured interviews to confirm the diagnosis. Similar to the West, there is a noticeable trend of rising rates of abnormal eating attitudes and body dissatisfaction in both males and females.

India’s linguistic and cultural diversity poses a unique challenge in translating and validating scales in our settings. This could be one reason for the variable rates of disordered eating, summarized in Table 1. Though the EAT-26 questionnaire has been translated into Hindi and Kannada, the methodological rigor in determining the cutoff scores is clouded. The Tamil version has only 21 items, and a cutoff score of ≥20 is used to determine the risk of an ED. Hence, there is a need for the development of culturally appropriate scales for measuring ED.[8,71]

The worldwide prevalence of pica in perinatal women is 27.8%, consistent with Indian reports.[110] Around 3.7% of the participants reported that their children consumed nonnutritious, nonfood substances.[111] As part of a broader population-based study, researchers surveyed 804 children and their parents to gather information about pica and RRD behaviors. The findings indicate that 4.98% of the participants reported recurring pica behaviors, and 1.49% reported recurring RD behaviors.[112] This is higher in comparison with prevalence rates reported in Indian studies.[73] The epidemiological characteristics of ARFID in children are unclear, and there is a lack of validated screening tools in India and internationally.

An amalgamation of diverse risk factors mediates the development of ED reported worldwide and in Indian research. Genetic predisposition, the interaction between gastrointestinal microbiota and autoimmune reactions, traumatic events or family dynamics, personality traits and psychiatric comorbidity, sex, socioeconomic status, elite sports, ethnicity, and societal body image ideals may elevate the risk of developing ED. There is a notable gap in knowledge regarding the risk factors associated with FD. These findings underscore the need for more targeted research efforts to better understand and address the risk factors specific to these newly recognized EDs.[23,33,47,70,113]

Culture partakes in shaping the presentation of EDs in India. A distinctive aspect observed in India is the paucity of preoccupation regarding body fat, a phenomenon termed the “nonfat phobic” variant of AN. These atypical presentations are also reported in other South Asian countries. However, recent studies suggest a growing association between body shape perception and higher scores on ED assessments, possibly influenced by ongoing societal transitions and the increasing prevalence of Western ideals.[8]

It is worth noting that, similar to international research, approximately 50% of individuals with EDs also have coexisting psychiatric conditions, with depression being the most common.[114,115] The management of FEDs in India largely follows Western practices, combining pharmacotherapy and psychotherapy.[116,117,118] Medications such as SSRIs, second-generation antipsychotics, and cyproheptadine have proven effective for AN. In cases of BN, patients have been treated with doses of 20–80 mg/day of fluoxetine in Indian case reports, although higher doses, particularly of fluoxetine, are more common globally. Psychotherapeutic approaches employed for FED in India, such as family-based therapy and CBT, align with international practices.

Setting priorities for research and translation

Based on our review, we identified the following priorities for future research on FED in India.

1. While there has been a rise in the number of hospital-based studies in FED in India, many research gaps still need to be addressed. In a qualitative study in South Asia, researchers investigated the obstacles hindering help-seeking for ED and uncovered several themes. These themes encompassed a lack of awareness regarding ED, the stigma surrounding mental health issues, concerns about privacy during hospital visits, and varying social circumstances.[119] These concerns may lead to limited referrals to tertiary care, where the studies are conducted, and the low utilization of treatment services for these concerns. To ensure representative results, epidemiological surveys on FED should target nonclinical populations. Given the comparatively low prevalence rates of ED in this region, larger sample sizes are crucial for adequate statistical power when comparing subgroups. One possible solution is to have registry-based studies. Studies from well-designed patient registries offer real-world clinical insights, including outcomes, safety, and cost-effectiveness, serving various evidence and decision-making purposes.[120]

2. Culture significantly shapes the manifestation of FED. As globalization and acculturation gain influence, clinical presentations of FED may change. For example, ED manifestations marked by prominent body image disturbances may rise. ARFID, which is barely mentioned in the Indian research landscape, may gain prominence. Also, current research has identified possible correlations between certain biopsychosocial factors and FED. It is unclear what risk factors are important for our settings and how they mediate their effects on disordered eating. Further prospective longitudinal research is required to test these hypotheses and develop targeted prevention strategies.

3. Another concern is the use of current DSM criteria and related tools for identifying FED in Eastern societies such as India. Being validated primarily in Western Caucasian populations may not directly apply to Indian cultures. This could lead to misleading findings, impacting practice and policy. There is a documented “misuse” of these criteria in different cultural settings, leading to inaccurate prevalence estimates.[8] Whether the ICD-11 criteria and developing scales attuned to the cultural variations will address this issue is yet to be seen.

4. The current practice trends in evaluating and managing FED align with Western guidelines. How can the treatments for FED be modified to suit the diverse needs of Indians? The available hospital-based studies discuss broadly the nature of the interventions delivered and outcomes, but there are no descriptions of the specific components and whether they were modified to suit Indian families. Building structured, brief, culturally adapted psychotherapy modules that can be delivered in multiple languages and possibly even in nonspecialist settings for a low resource country like India is needed.

5. What about early identification and prevention? Given that available research points toward a greater prevalence of subsyndromal presentations of disordered eating, how can busy healthcare professionals catering to the large Indian population in primary care settings identify these atypical presentations, and what are some of the critical strategies that can be implemented briefly to prevent the further rise of syndromal FED. Brief, culturally validated screening tools and evidence-based interventions that suit the unique requirements of India are a pressing priority.

6. Research in minority groups: One striking observation is the shortage of research in FED among lesbian, gay, bisexual, transgender, queer or questioning, intersex, asexual, and others (LGBTQIA+) communities in India. LGBTQIA+ individuals face an elevated risk of ED.[121] This underscores the necessity for directing future research toward developing effective screening and treatment approaches in minority groups to ensure equitable access to health care services.

CONCLUSION

In conclusion, there has been a growing research interest and enormous strides in FED in India in the past two decades. The relatively low prevalence of FED may explain the limited focus on FED research in India. Cultural disparities between Eastern and Western contexts have led to variations in FED presentations and diagnostic challenges. While there are reports of the NFP-AN variant of EDs, a growing association between body shape perception and abnormal eating attitudes was observed. The prevalence of pica in children is lower in the Indian population compared with global data. The epidemiological characteristics of other FEDs in children need further evaluation. Therefore, developing culturally sensitive diagnostic tools and gathering locally relevant epidemiological data on FED from both community and hospital settings in India are imperative.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.