You are on page 1of 8

Received: 1 May 2018 Revised: 25 May 2018 Accepted: 2 June 2018

DOI: 10.1002/erv.2613

ORIGINAL ARTICLE

Food addiction among men and women in India

Ashley A. Wiedemann1† | Jessica L. Lawson1† | Paige M. Cunningham1 |


Kathryn M. Khalvati1 | Janet A. Lydecker1 | Valentina Ivezaj1 | Carlos M. Grilo1,2

1
Psychiatry Department, Yale School of
Medicine, New Haven, Connecticut, USA
Abstract
2
Yale University, New Haven, Objective: This study aimed to address a cultural gap in the food addiction
Connecticut, USA (FA) literature by examining FA and associated clinical features in a nonclini-
cal group of men and women residing in India.
Correspondence
Ashley A. Wiedemann, PhD, Yale School Method: Participants (N = 415) were recruited from Amazon Mechanical
of Medicine, Program for Obesity, Weight, Turk to complete an online survey about weight and eating. Participants com-
and Eating Research, 301 Cedar Street,
New Haven, CT 06519, USA. pleted self‐report measures assessing FA (Yale Food Addiction Scale [YFAS]),
Email: ashley.wiedemann@yale.edu eating‐disorder psychopathology (Eating Disorder Examination‐Questionnaire
[EDE‐Q]), health‐related quality of life (Short Form Health Survey—12‐item
Funding information
National Institutes of Health, Grant/ version [SF‐12]), and depression (Patient Health Questionnaire‐2).
Award Number: K24 DK070052 Results: The FA symptom mean was 3.53 (SD = 1.90); 32.5% (n = 129) met
FA clinical threshold on the YFAS. Groups categorized with and without FA
on the YFAS did not differ significantly in sex or body mass index. YFAS scores
were significantly correlated with greater frequency of binge eating, higher
severity scores on all EDE‐Q subscales, higher depression, and poorer function-
ing scores on the SF‐12 (all ps < 0.05).
Conclusions:FA, as conceptualized and measured by the YFAS, appears to
be common among individuals residing in India.

KEYWORDS
body mass index, food addiction, India, weight, Yale Food Addiction Scale

1 | INTRODUCTION is considered controversial, with debated discourse


centring around the neurobiological mechanisms of
The concept of food addiction (FA) continues to gain FA and its relation to discriminating features of other
attention, with recent interest directed towards under- addictive and eating disorders (Long, Blundell, &
standing the cross‐cultural presentation of FA Finlayson, 2015). The examination of FA from varied
symptomology and clinical correlates (Brunault et al., cultural perspectives may help address gaps in the liter-
2017; Chen, Tang, Guo, Liu, & Xiao, 2015). The FA ature and contribute to greater understanding of the
model posits that certain individuals are susceptible to construct.
a patterned addictive response to highly palatable foods, FA (generally assessed using versions of the Yale Food
which reinforces problematic eating behaviour Addiction Scale [YFAS]; Gearhardt, Corbin, & Brownell,
(Gearhardt, Davis, Kuschner, & Brownell, 2011). 2009), in both clinical and nonclinical study groups, has
Among the scientific community, the construct of FA consistently been associated with biopsychosocial vari-
ables. Prior findings suggest strong associations between

These authors contributed equally to this work. FA and obesity, depression, health‐related quality of life,

Eur Eat Disorders Rev. 2018;26:597–604. wileyonlinelibrary.com/journal/erv © 2018 John Wiley & Sons, Ltd and Eating Disorders Association. 597
598 WIEDEMANN ET AL.

and disordered eating behaviours, including emotional


eating and binge eating (Burrows, Kay‐Lambkin, Pursey, Highlights
Skinner, & Dayas, 2018; Chao et al., 2017; de Vries &
• This is the first study examining clinical
Meule, 2016; Gearhardt et al., 2009; Gearhardt et al.,
correlates of FA in India.
2012; Gearhardt, Boswell, & White, 2014a; Pursey,
• FA clinical threshold based on the YFAS was
Stanwell, Gearhardt, Collins, & Burrows, 2014; Schulte,
met by 32.5% of participants.
Grilo, & Gearhardt, 2016; Wolz, Granero, & Fernandez‐
• FA was associated with eating‐disorder
Aranda, 2017). FA appears more common in women,
psychopathology, depression, and lower
whereas studies examining the association of FA and
quality of life.
racial/ethnic identity have been mixed (Burrows et al.,
2018). Overall, the majority of studies examining FA have
been conducted in the United States with relatively homo-
geneous participant groups (e.g., White women with obe- significantly associated with increased likelihood of an
sity). Further, FA is most commonly studied among FA “diagnosis,” and moreover, this effect was driven pre-
individuals seeking weight loss treatment, individuals with dominately by participants identifying as South Asian
disordered eating, and undergraduate students (Burrows (Meadows, Nolan, & Higgs, 2017). Compared with White
et al., 2018; Pursey et al., 2014). These parameters limit participants, South Asian participants endorsed signifi-
generalizability of findings to populations of diverse eth- cantly more YFAS symptoms and were significantly more
nicity, race, and socio‐economic and weight status, as well likely to self‐classify as having FA (Meadows et al., 2017).
as to nonclinical groups. These findings parallel research citing significantly more
As so few studies have examined FA in culturally heter- disordered eating behaviours (such as binge eating) and
ogenous groups, little is known about how FA presents in atypical cultural presentations of eating disorders in indi-
these understudied populations. As noted above, the viduals from the Indian subcontinent when compared
method predominately used to measure and operationalize with White individuals (Gupta, Chaturvedi, Chandarana,
FA (i.e., the YFAS; Gearhardt et al., 2009) is a self‐report & Johnson, 2001; Lal, Abraham, Parikh, & Chhibber,
questionnaire based on the seven substance use disorder 2015). These findings may also be of importance within
diagnostic criteria in the Diagnostic and Statistical Manual the context of the rising obesity prevalence in India,
of Mental Disorders (American Psychiatric Association which has increased from 12.1%, documented in 2000,
[DSV‐IV‐TR], 2000). Two recent meta‐analytic studies to 19.7% in 2016 (World Health Organization, 2016).
reported the weighted mean percentage of participants A recent review suggests that increases in obesity and
that met the YFAS clinical threshold for FA was 16.2% problematic eating behaviours have occurred concurrent
and 19.9%, respectively (Burrows et al., 2018; Pursey with the rise of globalization and industrialization in South
et al., 2014). Burrows et al. (2018) observed that the YFAS Asian countries, including India (Pike & Dunne, 2015).
has been translated and validated into several languages Moreover, emerging evidence suggests that meeting the
including Arabic, Italian, French, German, Spanish, Chi- clinical threshold of FA may be even more common among
nese, and Turkish. Such work takes important steps Indian individuals compared with other ethnicities
towards understanding how FA translates across cultures (Meadows et al., 2017). Coupled together, these findings
by evaluating the validity of the FA construct. Research is provide further clinical relevance for assessing FA in this
still needed, however, to understand cultural influences population. This study aims to address a cultural gap in
and correlates of FA across cultural constructs. Addressing the FA literature and examine FA, as measured by the
this gap may be particularly pertinent for countries under- YFAS, and associated clinical features among an online
going rapid cultural and economic changes, as eating community group residing in India.
behaviours appear to be shaped by sociocultural phenom-
ena and thus may respond to cultural transition (Kelly,
2 | METHODS
2003; Pike, Hoek, & Dunne, 2014).
India is one such country undergoing rapid global
2.1 | Participants
integration characterized by economic, nutritional, and
sociocultural changes (Raskind, Patil, Haardörfer, & Cun- Participants (N = 415) were recruited from the Mechani-
ningham, 2018; Thomas, Lee, & Becker, 2016). FA and cal Turk website to complete an online survey about
associated clinical features remain relatively weight and eating. Individuals were eligible to participate
understudied among individuals from India. Importantly, and included in study analyses if they were over 21 years
a study conducted with an ethnically diverse online inter- old, spoke English, and were born and currently resided
national sample found that non‐White ethnicity was in India. Mechanical Turk is a recruitment platform that
WIEDEMANN ET AL. 599

yields high‐quality and convenient participant data lifestyle behaviours. Participants reported their cur-
(Buhrmester, Kwang, & Gosling, 2011; Buhrmester, rent weight and height, which were used to calculate
Talaifar, & Gosling, 2018; McCredie & Morey, 2018), in BMI (kg/m2). All measures were completed in
part by including checks for attentiveness during survey English.
completion. Samples have greater demographic (e.g., age)
and geographical diversity than traditionally recruited
convenience samples, such as undergraduate students 2.2.1 | Yale Food Addiction Scale
and other internet samples (Behrend, Sharek, Meade, &
The YFAS (Gearhardt et al., 2009) is a 25‐item self‐report
Wiebe, 2011) but otherwise have similar characteristics to
measure of FA. Items were developed in correspondence
samples recruited by other methodologies (McCredie &
with substance dependence criteria from DSM‐IV‐TR,
Morey, 2018). Mechanical Turk also allows for the specific
and symptoms are assessed with respect to the consump-
recruitment of participants by qualifications, including
tion of certain foods such as sweets, starches, salty
country of residence (i.e., India). Comparisons of the
snacks, fatty foods, and sugary drinks. Symptoms are
psychometric properties of measures completed by
assessed with respect to the past 12 months. The YFAS
Mechanical Turk participants have similar reliability and
offers two scoring methods: (a) dimensional total of
validity as those completed by participants recruited using
endorsed symptoms and (b) a dichotomous assessment
traditional sources (Hauser & Schwarz, 2016; Shapiro,
of FA “diagnosis,” which is met when at least three symp-
Chandler, & Mueller, 2013). Mechanical Turk has increas-
toms and clinically significant impairment or distress are
ing popularity in the social sciences (Buhrmester et al.,
endorsed (Gearhardt et al., 2009). Psychometrically, the
2018) and has been used in psychological and psychiatric
YFAS has a one‐factor structure with adequate internal
research (Mathes, Norr, Allan, Albanese, & Schmidt,
consistency (α = 0.86), convergent validity, and incre-
2018; McCredie & Morey, 2018), including eating research
mental validity in predicting binge eating (Gearhardt
(Burrows et al., 2017; Powell, Frankel, Umemura, &
et al., 2009; Gearhardt, Boswell, & White, 2014a).
Hazen, 2017). For the present investigation, participants
Cronbach's α of 0.93 was obtained in the present study.
were paid 50 cents, consistent with most Mechanical Turk
research (Buhrmester et al., 2011). This payment exceeds
the payment threshold recommended in a recent study to 2.2.2 | Eating Disorder Examination‐
increase the attentional aspects and quality of data pro- Questionnaire
vided by Mechanical Turk participants recruited in India
(Litman, Robinson, & Rosenzweig, 2015). The Eating Disorder Examination Questionnaire (EDE‐
Among participants, 79.5% (n = 330) reported English Q) provides a measure of the range and severity of
as their primary language. Participants were predomi- eating‐disorder psychopathology during the past 28 days.
nantly male (63.4%, n = 263), and the racial–ethnic distri- An abbreviated version (behavioural symptoms plus
bution of the sample was 98.6% Asian (n = 409), 1% seven items comprising three confirmatory factor analy-
White (n = 4), 0.2% Black (n = 1), and 0.2% multiracial sis‐established factors or subscales) used in this study
(n = 1). The mean age and body mass index (BMI) of par- has demonstrated strong psychometric properties that
ticipants were 32.01 (SD = 7.93) years and 24.25 are significantly superior to those for the original EDE‐
(SD = 4.99) kg/m2, respectively. Of the total sample, Q in both nonclinical (Grilo, Reas, Hopwood, & Crosby,
53.5% of participants were within the “healthy” weight 2015; Machado, Grilo, & Crosby, 2018) and clinical
range, 8.6% were in the underweight range, 26.2% were groups (Machado et al., 2018). Subscales include
within the overweight range, and 11.7% were in the obe- restraint, overvaluation of shape and weight, and dissat-
sity range. The majority of participants (n = 238; 55.7%) isfaction with shape and weight. Frequency of objective
had attended college, 33.3% attended more than college, binge eating episodes, characterized as eating an unusu-
7% attended some college or obtained an associate degree, ally large amount of food while feeling a subjective
and 4.1% attended high school or less. This study received sense of loss of control, was also assessed. Cronbach's
approval from the University Institutional Review Board, α for the EDE‐Q subscales in this study ranged from
and all participants provided electronic informed consent 0.84 to 0.87.
prior to participation.
2.2.3 | Medical Outcomes Study Short
2.2 | Measures Form Health Survey
Participants completed self‐report measures assessing The Medical Outcomes Study Short Form Health Survey
demographic information, eating patterns, and —12‐item version (SF‐12) is a brief version of the SF‐36
600 WIEDEMANN ET AL.

(Ware Jr. & Sherbourne, 1992) and a widely used mea- SD = 1.27), t(117) = −2.23, p = 0.03. Independent sam-
sure of health‐related quality of life, including two sum- ples t tests were conducted to assess for group differences
mary scores: physical and mental health. The SF‐12 has between participants who met criteria for FA and those
been psychometrically assessed and validated across eth- who did not. There were no significant differences found
nicities (Gandek et al., 1998). Internal consistency statis- between the FA and no‐FA groups for age, BMI, or years
tics reported among a sample of individuals from of education (all p values >0.05). YFAS symptom scores
Northern India (α = 0.68–0.80) suggest it is appropriate were also not significantly associated with age
for use in Indian samples (Wind, Joshi, Kleber, & (r = −0.04, p = 0.44), BMI (r = 0.03, p = 0.58), or years
Komproe, 2013). of education (r = 0.003, p = 0.95). Chi‐squared tests
revealed no significant differences for FA group by sex,
χ2 (2, N = 369) = 4.01, p = 0.14.
2.2.4 | Patient Health Questionnaire Frequencies of the seven symptoms on the YFAS were
The Patient Health Questionnaire‐2 (PHQ‐2; Kroenke, examined. Across all participants, YFAS symptoms were
Spitzer, & Williams, 2003) is a brief version of the PHQ‐ endorsed with the following frequencies: 28.2%
9 (Spitzer, Kroenke, & Williams, 1999) that assesses the (n = 104) reported consuming more than intended;
frequency of depressive mood and anhedonia over the 88.3% (n = 326) reported an inability to cut down or stop
past 2 weeks. Composed of the first two items of the eating; 60.7% (n = 224) reported reduced participation in
PHQ‐9 (“feeling down, depressed, or hopeless” and “little important activities; 59.3% (n = 219) reported tolerance
interest or pleasure in doing things”), the PHQ‐2 symptoms; 42.3% (n = 156) reported continued use
decreases screening burden for depression due to its brev- despite consequences; 39.0% (n = 144) reported with-
ity and ease of scoring. The PHQ‐2 total score ranges drawal symptoms; 34.7% (n = 128) reported they spent a
from 0 to 6; a cut‐point of 3 indicates the need for further great deal of time obtaining food, eating throughout the
depression screening (Kroenke et al., 2003). day, or recovering from overeating; and 32.8% (n = 121)
reported clinical impairment. Of the participants who
met the clinical threshold of FA, 10 participants (9.0%)
2.3 | Statistical analyses had a BMI in the underweight category, 64 participants
We used SPSS 24.0 version to analyse these data. Bivari- (57.7%) had a BMI in the “healthy” weight category, 24
ate correlations were used to examine the association participants (21.6%) had a BMI in the overweight cate-
between the YFAS symptom score and demographics gory, and 13 participants (11.7%) had a BMI in the obesity
(e.g., age), BMI, eating‐disorder psychopathology (i.e., category.
EDE‐Q: brief version), and measures of health‐related The YFAS symptom score was also examined in rela-
quality of life and depression (i.e., SF‐12 and PHQ‐2). tion to clinical features. Table 1 summarizes descriptives
Independent samples t tests were conducted to compare and statistical tests on the YFAS. When analysed dimen-
participants who met the clinical threshold of FA on the sionally, the YFAS was significantly correlated with
YFAS (FA group) to those without (no‐FA group) for all depression as assessed by the PHQ‐2 and both mental
continuous variables (e.g., age and BMI), and chi‐squared and physical subscales on the SF‐12. Participants who
analyses were used to compare the groups on sex. met the clinical threshold for FA on the YFAS had signif-
Cohen's d was computed for effect sizes. icantly higher depression scores as measured by the
PHQ‐2 than participants who did not meet the FA
threshold. The effect size for this analysis (d = 0.93)
3 | RESULTS was found to exceed Cohen's parameter for a large effect
(d ≥ 0.80). Similarly, participants in the FA group
The YFAS FA symptom mean was 3.53 (SD = 1.90), and reported significantly poorer SF‐12 mental and physical
32.5% (n = 120) of the sample met the clinical threshold scores compared with the no‐FA group. The effect sizes
for FA. YFAS symptom scores were analysed dimension- for these analyses as measured by Cohen's d were
ally and categorically with respect to demographic vari- medium.
ables. Of the participants who met criteria for FA, 70 The YFAS was also significantly correlated with the
(58.3%) identified as male, 49 (40.8%) identified as female, following features of eating‐disorder psychopathology as
and 1 (.8%) identified as “other.” There were significant measured by the brief version of the EDE‐Q: restraint,
sex differences in the number of FA symptoms endorsed overvaluation of weight and shape, dissatisfaction with
when comparing men with women, with women weight and shape, and binge eating frequency (Table 1).
reporting a greater total number of symptoms on the When comparing participants with and without FA, cate-
YFAS (M = 5.51, SD = 1.26) than men (M = 4.99, gorically, participants with FA endorsed significantly
WIEDEMANN ET AL. 601

TABLE 1 Means and standard deviations of BMI, eating disorder psychopathology, and psychosocial functioning by group and bivariate
correlations with YFAS

Total sample Food addiction No food addiction Test statistic Effect size
YFAS symptom
Variables M (SD) score (r) M (SD) M (SD) t test (t) Cohen's d

BMI 24.31 (5.03) 0.03 24.23 (5.16) 24.28 (5.04) 0.09 0.01
EDE‐Q restraint 2.12 (1.74) 0.13* 2.27 (1.51) 2.04 (1.86) −1.30 0.14
EDE‐Q overvaluation 2.72 (1.56) 0.14** 2.93 (1.42) 2.58 (1.61) −2.10* 0.23
EDE‐Q dissatisfaction 2.44 (1.69) 0.27*** 2.82 (1.56) 2.22 (1.70) −3.29*** 0.37
Binge eating frequency 3.72 (6.21) 0.24*** 5.02 (5.98) 3.01 (6.24) −2.94** 0.33
SF‐12 physical 45.78 (8.90) −0.40*** 40.07 (6.96) 45.05 (8.25) 6.04*** −0.65
SF‐12 mental 47.07 (10.44) −0.40*** 41.05 (7.47) 47.78 (9.83) 7.27*** −0.77
PHQ‐2 2.17 (1.63) 0.50*** 3.07 (1.43) 1.70 (1.51) −8.44*** 0.93

Note. BMI: body mass index; YFAS: Yale Food Addiction Scale; EDE‐Q: Eating Disorder Examination‐Questionnaire; SF‐12: Short Form Health Survey—12‐
item version; PHQ‐2: Patient Health Questionnaire‐2.
*p < 0.05. **p < 0.01. ***p ≤ 0.001.

greater overvaluation of weight and shape, binge eating presence of FA and the relatively higher experience of
frequency, and dissatisfaction with weight and shape. FA among Indians.
Effect sizes measured by Cohen's d indicated small effects Although the percentage of participants meeting the
for these analyses. The two groups did not significantly clinical threshold of FA in this study was higher than
differ in EDE‐Q restraint or BMI. has been observed in previous studies, the overall mean
of symptoms reported on the YFAS was similar to past
findings (Burrows et al., 2018; Pursey et al., 2014). On
4 | DISC USS I ON average, participants in this study endorsed 3.53 out of a
possible total of seven symptoms on the YFAS, similar
This, to our knowledge, is the first study to examine FA to the mean reported in a recent meta‐analysis
in a study group of adults residing in India. The findings (M = 3.30), which included a broader population and
indicated that 32.5% of participants met the YFAS‐ studies with a greater range of international locations
defined clinical level of FA, which is higher than (Burrows et al., 2018) than a previous meta‐analysis
observed in recent meta‐analytic findings (16.2% to (M = 2.80) examining the YFAS (Pursey et al., 2014).
19.9%) that compiled studies with samples predominantly The overall mean on the YFAS was also similar to the
from the United States (Burrows et al., 2018; Pursey et al., mean obtained in a prior sample of primarily South Asian
2014). Our findings, however, were consistent with a participants from India and Pakistan (M = 3.20; Meadows
recent study examining South Asian participants using et al., 2017). South Asian participants were more likely to
similar research methodology, which cite a similar YFAS self‐identify as having “FA” and were more likely to
symptom mean in their sample (Meadows et al., 2017). endorse nearly all symptoms on the YFAS when com-
Our findings suggest that symptoms of FA are associated pared with White participants and other ethnicities, aside
with more frequent binge eating and eating‐disorder psy- from “repeated failed attempts to quit or cut down despite
chopathology, as well as poorer health‐related quality of negative consequences” (Meadows et al., 2017). Likewise,
life and depression; these results are consistent with stud- South Asian participants in their study reported higher
ies conducted primarily in the United States (Burrows scores on the Food Cravings Questionnaire (Meadows
et al., 2018; Pursey et al., 2014). In the present study, et al., 2017), and previous findings demonstrate that food
FA was not related to either sex or BMI (when examined cravings are associated with FA (Gearhardt, Rizk, &
dimensionally or categorically). On average, the FA group Treat, 2014b; Meule & Kübler, 2012). In the present
reported lower health‐related quality of life, including study, FA “diagnosis” was common (32.5%) and strongly
physical and mental health; however, both groups were associated with poorer health‐related quality of life and
within the normal range. These findings add to the poorer depression. The observed frequency of FA was in
relatively unexplored but growing research on eating part due to the number of participants endorsing clinical
pathology in non‐Western societies by confirming the impairment (32.8%). Taken together, results of the
602 WIEDEMANN ET AL.

present study and previous research (Meadows et al., weight spectrum (Schulte et al., 2016), and many studies
2017) suggest that individuals in India with FA experi- have not identified an association between BMI and
ence clinically significant distress and associated psycho- YFAS symptoms or diagnoses (Meule & Gearhardt,
pathology. These findings warrant the need for 2014) leading to the consideration that obesity should
continued cross‐cultural examination of FA and related not be used as a proxy for FA (Schulte et al., 2016). Addi-
sequalae including prospective studies in this South tionally, FA in this study was unrelated to sex or age,
Asian population. inconsistent with previous studies of the YFAS (Pursey
Our findings suggest that FA was also related to et al., 2014); however, women categorized with FA
greater eating‐disorder psychopathology, including binge reported a greater number of symptoms on the YFAS,
eating and body image dissatisfaction, which is consistent which is a consistent finding in past research (Pursey
with previous studies of varying cultural groups (Burrows et al., 2014). Our findings might reflect cultural differ-
et al., 2018; Pursey et al., 2014). Despite the widespread ences in the presentation of FA and thus call for longitu-
belief that disordered eating behaviours are less prevalent dinal and epidemiological methodology to examine
in Asian countries, a recent review highlighted that the eating behaviours in the context of a newly industrialized
differences in disordered eating, dieting, shape, and country, such as India.
weight concerns as compared with countries in the West The Cronbach's alpha for the YFAS in the current
are diminishing (Pike & Dunne, 2015). It is hypothesized study suggests excellent internal consistency of the
that this may be in part due to Asian countries becoming YFAS in this Indian sample. The construct of FA has
more industrialized and globalized (Pike & Dunne, 2015). received little research attention with Indian partici-
For instance, it has been suggested that urbanization and pants, and it is not yet known how people living in
socio‐economic status are associated with increased risk India may perceive or experience the construct of FA
of weight dissatisfaction and dieting in India (Mishra & as measured by the YFAS. It is important to note, how-
Mukhopadhyay, 2011; Talukdar, 2012). Moreover, past ever, that a recent study examining “self‐perceived” FA
studies directly comparing those in the United States with found that South Asian individuals were more likely to
those in India have not found significant differences in consider they had an addiction to food when compared
eating disorder psychopathology, including body dissatis- with all other ethnicities (i.e., White, African‐American,
faction and frequency of binge eating episodes (Lal et al., Hispanic, or “other”; Meadows et al., 2017). Taken
2015; Meadows et al., 2017; Rubin, Gluck, Knoll, Lorence, together, future studies, which examine the psychomet-
& Geliebter, 2008). Nonetheless, more research that ric properties of the YFAS in this population, are
includes large‐scale epidemiological studies of cross‐cul- important.
tural comparisons is needed to reduce biases that can Our findings should be interpreted in light of the
arise in meta‐analytic comparisons. study limitations. The cross‐sectional study design and
Interestingly, some of the results of our study depart convenience sampling approach using self‐reported sur-
from earlier research. Although endorsement of FA, as vey data limit the generalizability of results. Question-
assessed by the YFAS, was quite common in this study, naires are susceptible to inattention; however, the
FA was unrelated to BMI when examined both dimen- online platform used for recruitment in the current study,
sionally and categorically. Past studies have documented Mechanical Turk, provides high‐quality data and allows
that a higher BMI is associated with more symptoms on for the inclusion of items designed to evaluate
the YFAS and greater likelihood of FA; however, it is participants' attention (Buhrmester et al., 2018). Indeed,
important to note that the majority of past studies using this platform has growing popularity and acceptance
the YFAS have been conducted in clinical groups of par- and has been used in psychiatric research (Mathes
ticipants with overweight/obesity (Pursey et al., 2014). et al., 2018; McCredie & Morey, 2018), including eating
Furthermore, approximately 12% of participants in the research (Burrows et al., 2017; Gearhardt, Corbin, &
current study were in the obesity weight category, and Brownell, 2016; Powell et al., 2017). Moreover, the study
the percentage of participants with obesity who met the design allowed us to examine FA among adults residing
clinical threshold for FA was nearly identical. The per- in India, addressing a significant gap in the literature.
centage of participants in the obesity weight category, Of note, participants in this study had higher levels of
however, is somewhat lower than rates observed from education than found in prior studies, and findings
past research—approximately 20% in 2016 (World Health pertain to those residing in India who spoke English.
Organization, 2016), and the majority of participants Findings may not generalize to non‐English‐speaking
meeting the clinical threshold for FA in this study were individuals residing in India or to those with varying
in the “healthy” or underweight category. Nonetheless, education or sociodemographic characteristics. Perhaps
it has been noted that FA is experienced across the surprisingly, roughly half of the participants were male,
WIEDEMANN ET AL. 603

although this feasibly resulted in greater generalizability Burrows, T., Kay‐Lambkin, F., Pursey, K., Skinner, J., & Dayas, C.
and allowed for more adequate exploration of potential (2018). Food addiction and associations with mental health
gender differences in the YFAS scores as previous studies symptoms: A systematic review with meta‐analysis. Journal of
Human Nutrition and Dietetics.
of the YFAS have been predominately female (Pursey
et al., 2014). Burrows, T., Skinner, J., Joyner, M., Palmieri, J., Vaughan, K., &
Gearhardt, A. (2017). Food addiction in children: Associations
with obesity, parental food addiction and feeding practices. Eat-
5 | CONCLUSIONS ing Behaviors, 26, 114–120.
Chao, A. M., Shaw, J. A., Pearl, R. L., Alamuddin, N., Hopkins, C.
The findings of the present study suggest that YFAS‐ M., Bakizada, Z. M., … Wadden, T. A. (2017). Prevalence and
defined “FA” might be common among individuals resid- psychosocial correlates of food addiction in persons with obesity
seeking weight reduction. Comprehensive Psychiatry, 73, 97–104.
ing in India. FA in this study was significantly associated
with eating‐disorder psychopathology, including binge Chen, G., Tang, Z., Guo, G., Liu, X., & Xiao, S. (2015). The Chinese
version of the Yale Food Addiction Scale: An examination of its
eating, depression, and health‐related quality of life but
validation in a sample of female adolescents. Eating Behaviors,
was not related to BMI or sex. These findings challenge
18, 97–102.
previous notions of disordered eating as a uniquely
de Vries, S. K., & Meule, A. (2016). Food addiction and bulimia
“Western” problem and highlight the importance of
nervosa: New data based on the Yale Food Addiction Scale
examining eating behaviours, such as addictive‐like eat- 2.0. European Eating Disorders Review, 24(6), 518–522.
ing, across cultures.
Gandek, B., Ware, J. E., Aaronson, N. K., Apolone, G., Bjorner, J. B.,
Brazier, J. E., et al. (1998). Cross‐validation of item selection and
A C K N O WL E D G E M E N T scoring for the SF‐12 Health Survey in nine countries: Results
from the IQOLA Project. International Quality of Life Assess-
This research was supported, in part, by the National ment. J Clin Epidemiol, 51(11), 1171–1178.
Institutes of Health (K24 DK070052). Gearhardt, A. N., Boswell, R. G., & White, M. A. (2014a). The asso-
ciation of “food addiction” with disordered eating and body
mass index. Eating Behaviors, 15(3), 427–433.
CONFLICT OF INTEREST
Gearhardt, A. N., Corbin, W. R., & Brownell, K. D. (2009). Prelimi-
The authors report no conflicts of interest. nary validation of the Yale Food Addiction Scale. Appetite, 52(2),
430–436.
Gearhardt, A. N., Corbin, W. R., & Brownell, K. D. (2016). Develop-
ORCID ment of the Yale Food Addiction Scale Version 2.0. Psychology of
Ashley A. Wiedemann http://orcid.org/0000-0002-4918- Addictive Behaviors, 30, 113–121.

7502 Gearhardt, A. N., Davis, C., Kuschner, R., & Brownell, K. D. (2011).
Janet A. Lydecker http://orcid.org/0000-0001-6425-514X The addiction potential of hyperpalatable foods. Current Drug
Abuse Reviews, 4(3), 140–145.
Carlos M. Grilo http://orcid.org/0000-0003-0245-3444
Gearhardt, A. N., Rizk, M. T., & Treat, T. A. (2014b). The associa-
tion of food characteristics and individual differences with
R EF E RE N C E S ratings of craving and liking. Appetite, 79, 166–173.

American Psychiatric Association. (2000). Diagnostic and Gearhardt, A. N., White, M. A., Masheb, R. M., Morgan, P. T.,
statistical manual of mental disorders (4th ed. text revision). Crosby, R. D., & Grilo, C. M. (2012). An examination of the food
Washington, DC. addiction construct in obese patients with binge eating disorder.
The International Journal of Eating Disorders, 45(5), 657–663.
Behrend, T. S., Sharek, D. J., Meade, A. W., & Wiebe, E. N. (2011).
Grilo, C. M., Reas, D. L., Hopwood, C. J., & Crosby, R. D. (2015).
The viability of crowdsourcing for survey research. Behavior
Factor structure and construct validity of the Eating Disorder
Research Methods, 43, 800–813.
Examination‐Questionnaire in college students: Further support
Brunault, P., Courtois, R., Gearhardt, A. N., Gaillard, P., Journiac, for a modified brief version. The International Journal of Eating
K., Cathelain, S., … Ballon, N. (2017). Validation of the French Disorders, 48(3), 284–289.
version of the DSM‐5 Yale Food Addiction Scale in a nonclinical
Gupta, M. A., Chaturvedi, S. K., Chandarana, P. C., & Johnson, A.
sample. Canadian Journal of Psychiatry, 62(3), 199–210.
M. (2001). Weight‐related body image concerns among 18–24‐
Buhrmester, M., Kwang, T., & Gosling, S. D. (2011). Amazon's year‐old women in Canada and India: An empirical comparative
Mechanical Turk: A new source of inexpensive, yet high‐quality, study. Journal of Psychosomatic Research, 50(4), 193–198.
data? Perspectives on Psychological Science, 6, 3–5.
Hauser, D. J., & Schwarz, N. (2016). Attentive Turkers: MTurk par-
Buhrmester, M., Talaifar, S., & Gosling, S. D. (2018). An evaluation ticipants perform better on online attention checks than do
of Amazon's Mechanical Turk, its rapid rise, and its effective subject pool participants. Behavior Research Methods, 48,
use. Perspectives on Psychological Science, 13(2), 149–154. 400–407.
604 WIEDEMANN ET AL.

Kelly, B. D. (2003). Globalisation and psychiatry. Advances in Psy- Pursey, K. M., Stanwell, P., Gearhardt, A. N., Collins, C. E., & Bur-
chiatric Treatment, 9(6), 464–470. rows, T. L. (2014). The prevalence of food addiction as assessed
Kroenke, K., Spitzer, R. L., & Williams, J. B. (2003). The Patient by the Yale Food Addiction Scale: A systematic review. Nutri-
Health Questionnaire‐2: Validity of a two‐item depression ents, 6(10), 4552–4590.
screener. Medical Care, 41(11), 1284–1292. Raskind, I. G., Patil, S. S., Haardörfer, R., & Cunningham, S. A.
Lal, M., Abraham, S., Parikh, S., & Chhibber, K. (2015). A compar- (2018). Unhealthy weight in Indian families: The role of the
ison of eating disorder patients in India and Australia. Indian family environment in the context of the nutrition transition.
Journal of Psychiatry, 57(1), 37–42. Population Research and Policy Review, 37(2), 157–180.

Litman, L., Robinson, J., & Rosenzweig, C. (2015). The relationship Rubin, B., Gluck, M. E., Knoll, C. M., Lorence, M., & Geliebter, A.
between motivation, monetary compensation, and data quality (2008). Comparison of eating disorders and body image distur-
among US‐ and India‐based workers on Mechnical Turk. Behav- bances between Eastern and Western countries. Eating and
ior Research Methods, 47, 519–528. Weight Disorders, 13(2), 73–80.

Long, C. G., Blundell, J. E., & Finlayson, G. (2015). A systematic Schulte, E. M., Grilo, C. M., & Gearhardt, A. N. (2016). Shared and
review of the application and correlates of YFAS‐diagnosed unique mechanisms underlying binge eating disorder and addic-
‘food addiction’ in humans: Are eating‐related ‘addictions’ a tive disorders. Clinical Psychology Review, 44, 125–139.
cause for concern or empty concepts? Obesity Facts, 8(6), Shapiro, D. N., Chandler, J., & Mueller, P. A. (2013). Using Mechan-
386–401. ical Turk to study clinical populations. Clinical Psychological
Machado, P. P. P., Grilo, C. M., & Crosby, R. D. (2018). Replication Science, 1, 213–220.
of a modified factor structure for the Eating Disorder Examina- Spitzer, R. L., Kroenke, K., & Williams, J. B. (1999). Validation and
tion‐Questionnaire: Extension to clinical eating disorder and utility of a self‐report version of PRIME‐MD: The PHQ primary
non‐clinical samples in Portugal. European Eating Disorders care study. Primary Care Evaluation of Mental Disorders.
Review, 26(1), 75–80. Patient Health Questionnaire. JAMA, 282(18), 1737–1744.
Mathes, B. M., Norr, A. M., Allan, N. P., Albanese, B. J., & Schmidt, Talukdar, J. (2012). Thin but not skinny: Women negotiating the
N. B. (2018). Cyberchondria: Overlap with health anxiety and “never too thin” body ideal in urban India. Women's Studies
unique relations with impairment, quality of life, and service International Forum, 35(2), 109–118.
utilization. Psychiatry Research, 261, 204–211. Thomas, J. J., Lee, S., & Becker, A. E. (2016). Updates in the epide-
McCredie, M. N., & Morey, L. C. (2018). Who are the Turkers? A miology of eating disorders in Asia and the Pacific. Current
characterization of MTurk workers using the Personality Assess- Opinion in Psychiatry, 29(6), 354–362.
ment Inventory. Assessment. 1073191118760709 Ware, J. E. Jr., & Sherbourne, C. D. (1992). The MOS 36‐item short‐
Meadows, A., Nolan, L. J., & Higgs, S. (2017). Self‐perceived food form health survey (SF‐36). I. Conceptual framework and item
addiction: Prevalence, predictors, and prognosis. Appetite, 114, selection. Medical Care, 30(6), 473–483.
282–298. Wind, T. R., Joshi, P. C., Kleber, R. J., & Komproe, I. H. (2013). The
Meule, A., & Gearhardt, A. N. (2014). Food addiction in the light of impact of recurrent disasters on mental health: A study on sea-
DSM‐5. Nutrients, 6(9), 3653–3671. sonal floods in Northern India. Prehospital and Disaster
Meule, A., & Kübler, A. (2012). Food cravings in food addiction: Medicine, 28(3), 279–285.
The distinct role of positive reinforcement. Eating Behaviors, Wolz, I., Granero, R., & Fernandez‐Aranda, F. (2017). A compre-
13(3), 252–255. hensive model of food addiction in patients with binge‐eating
Mishra, S. K., & Mukhopadhyay, S. (2011). Eating and weight con- symptomatology: The essential role of negative urgency. Com-
cerns among Sikkimese adolescent girls and their biocultural prehensive Psychiatry, 74, 118–124.
correlates: An exploratory study. Public Health Nutrition, 14(5), World Health Organization (2016). Global Health Observatory
853–859. (GHO) data. Retrieved from http://www.who.int/gho/ncd/
Pike, K. M., & Dunne, P. E. (2015). The rise of eating disorders in risk_factors/overweight/en/
Asia: A review. J Eat Disord, 3, 33.
Pike, K. M., Hoek, H. W., & Dunne, P. E. (2014). Cultural trends How to cite this article: Wiedemann AA,
and eating disorders. Current Opinion in Psychiatry, 27(6),
Lawson JL, Cunningham PM, et al. Food addiction
436–442.
among men and women in India. Eur Eat Disorders
Powell, E. M., Frankel, L. A., Umemura, T., & Hazen, N. (2017).
Rev. 2018;26:597–604. https://doi.org/10.1002/
The relationship between adult attachment orientation and
erv.2613
child self‐regulation in eating: The mediating role of persua-
sive‐controlling feeding practices. Eating Behaviors, 26, 121–128.

You might also like