See

discussions, stats, and author profiles for this publication at:

http://www.researchgate.net/publication/5683018

The Night Eating Questionnaire (NEQ):

Psychometric properties of a measure of
severity of the Night Eating Syndrome
ARTICLE in EATING BEHAVIORS FEBRUARY 2008
Impact Factor: 1.58 DOI: 10.1016/j.eatbeh.2007.03.007 Source: PubMed

CITATIONS

READS

94

1,267

9 AUTHORS, INCLUDING:
Kelly C Allison

Jennifer D Lundgren

University of Pennsylvania

University of Missouri - Kansas City

94 PUBLICATIONS 2,422 CITATIONS

59 PUBLICATIONS 1,078 CITATIONS

SEE PROFILE

SEE PROFILE

David B Sarwer

Thomas A Wadden

University of Pennsylvania

University of Pennsylvania

167 PUBLICATIONS 4,815 CITATIONS

382 PUBLICATIONS 21,778 CITATIONS

SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate,

letting you access and read them immediately.

SEE PROFILE

Available from: Kelly C Allison

Retrieved on: 15 December 2015

Available online at www.sciencedirect.com

Eating Behaviors 9 (2008) 62 72

The Night Eating Questionnaire (NEQ): Psychometric properties of a

measure of severity of the Night Eating Syndrome
Kelly C. Allison a,, Jennifer D. Lundgren a,1 , John P. O'Reardon a , Nicole S. Martino a ,
David B. Sarwer a , Thomas A. Wadden a , Ross D. Crosby b ,
Scott G. Engel b , Albert J. Stunkard a
a

University of Pennsylvania School of Medicine Center for Weight and Eating Disorders, 3535 Market Street, Ste. 3021,
Philadelphia, PA 19104-3309, United States
University of North Dakota Neuropsychiatry Research Institute, 120 South 8th Street, Box 1415 Fargo, ND 58107-1415, United States
Received 17 November 2006; received in revised form 26 January 2007; accepted 16 March 2007

Abstract
The purpose of this study was to evaluate the Night Eating Questionnaire (NEQ) as a measure of severity of the Night Eating
Syndrome (NES). The 14-item NEQ assesses the behavioral and psychological symptoms of NES. The NEQ was evaluated in three
samples: 1980 persons who completed the NEQ on the Internet; 81 persons diagnosed with NES; and 194 bariatric surgery candidates.
Study 1, using principal components analysis, generated four factors (nocturnal ingestions, evening hyperphagia, morning anorexia,
and mood/sleep) and an acceptable alpha (.70). Confirmatory factor analysis suggested that 99% of covariation among factors is
accounted for by a higher-order construct. Study 2 found convergent validity of the NEQ with additional measures of night eating,
disordered eating, sleep, mood, and stress. Study 3 compared scores from obese bariatric surgery candidates with and without NES
and found appropriate discriminant validity of the NEQ. The NEQ appears to be an efficient, valid measure of severity for NES.
2007 Elsevier Ltd. All rights reserved.
Keywords: Night Eating Syndrome; Nocturnal ingestions; Evening hyperphagia; Assessment; Internal validity; Discriminant validity

The Night Eating Syndrome (NES) was first described in 1955 as a stress-related eating disorder consisting of
morning anorexia, evening hyperphagia, and insomnia (Stunkard, Grace, & Wolf, 1955). The presence of nocturnal
ingestions (awakening to eat) was added to these criteria later (Birketvedt et al., 1999). NES has also been associated
with depressed mood; it is generally lower in persons with NES, compared to controls and often worsens in the evening
and nighttime (Birketvedt et al., 1999).
Studies regarding the relationship between NES and weight status are mixed. As NES research has advanced, nonobese night eaters have been identified (Birketvedt et al., 1999; Marshall, Allison, O'Reardon, Birketvedt, & Stunkard,
2004). Epidemiological studies suggest that there is no relationship between NES and obesity (Rand, Macgregor, &
Stunkard, 1997; Striegel-Moore, Franko, Thompson, Affenito, & Kraemer, 2006), but studies of clinical populations
Corresponding author. University of Pennsylvania School of Medicine, Department of Psychiatry, Center for Weight and Eating Disorders, 3535
Market Street, Suite 3021, Philadelphia, PA 19104-3309, United States. Tel.: +1 215 898 2823; fax: +1 215 898 2878.
E-mail address: kca@mail.med.upenn.edu (K.C. Allison).
1
Present address: University of MissouriKansas City, 4825 Troost Building, Ste. 124, Kansas City, MO 64113.
1471-0153/$ - see front matter 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.eatbeh.2007.03.007

K.C. Allison et al. / Eating Behaviors 9 (2008) 6272

63

suggest that the prevalence is higher among overweight and obese patients (Aranoff, Geliebter, & Zammit, 2001;
Lundgren et al., 2006; Stunkard et al., 1955). The relationship between night eating and weight requires more attention.
Estimates of the prevalence of NES have ranged from 6% (Cer-Bjrk, Andersson, & Rssner, 1983) to 64%
(Stunkard et al., 1955) among patients seeking weight loss, and from 8% (Adami, Meneghelli, & Scopinaro, 1999;
Allison et al., 2006) to 42% (Hsu, Betancourt, & Sullivan, 1996) for persons seeking bariatric surgery. Prevalence
estimates of NES have also been reported in the following groups: 1.5% among the general population (Rand et al.,
1997), 12.3% among an outpatient psychiatric population (Lundgren et al., 2006), and 3.8% among a type 2 diabetic
population (Allison et al., 2007). Data from clinical samples have not established a relationship between NES and gender
or NES and ethnicity. Similarly, in an epidemiological sample, night eating behavior was not consistently associated with
gender or ethnicity (Striegel-Moore et al., 2006). The wide range of prevalence rates can be attributed to differing
diagnostic criteria and different assessment techniques (Allison & Stunkard, 2005; Striegel-Moore et al., 2006).
Night Eating Questionnaire (NEQ) development and description.
Differing diagnostic criteria have been used to identify NES, including the requirements for the amount of evening
hyperphagia (ranging from 25 to 50% of daily caloric intake after dinner) and the presence of nocturnal ingestions (for
review see Allison & Stunkard, 2005). Striegel-Moore and colleagues (2006) have evaluated different definitions of
night eating behavior in a community sample and found that increasingly rigorous definitions (50% of food intake after
7 pm vs. 25% of food intake after 7 pm) led to decreasing prevalence rates. Thus, these differences likely influenced the
widely varying reports of prevalence. Reliable and valid criteria, for prevalence estimates and clinical purposes, would
be improved with a validated NES assessment measure.
The original, unpublished, version of the Night Eating Questionnaire (NEQ) was a nine item measure with a 4-point
Likert scale. Its items assessed: morning anorexia (2 items), evening hyperphagia (1 item), initial insomnia (1 item),
mid-phase insomnia (1 item), nocturnal ingestions (1 item), and mood (3 items). As research continued to expand our
understanding of NES (Cer-Bjrk et al., 1983; Gluck, Geliebter, & Satov, 2001; Marshall et al., 2004; Napolitano,
Head, Babyak, & Blumenthal, 2001; O'Reardon et al., 2004), the NEQ was revised, including six new items added
over time, and the conversion of items to a 5-point scale. One of these intermediate versions was included in the weight
and lifestyles inventory (WALI; Wadden & Foster, 2001) and contained visual analog scales to assess levels of morning
hunger and percentage of caloric intake consumed after dinner. Another version was published by Marshall et al.
(2004) as the NEQ continued to be refined.
Our research and clinical experiences with NES led to the 14-item questionnaire that we sought to test in the current
studies (Appendix). Five of the new items were added to assess psychological aspects of NES, such as cravings and
feelings of control over late evening and nocturnal eating and whether there is a compulsion to eat to fall back to sleep.
The sixth new item was added to assess the level of awareness of nocturnal eating episodes to differentiate between
NES and Sleep Related Eating Disorder (Schenck & Mahowald, 1994), in which nocturnal ingestions occur with little
to no awareness or later recollection. Finally, one of the original three mood questions that assessed when mood was
best during the day was excluded to diminish the weight that mood questions contributed to the total score.
This current version of the NEQ assesses morning hunger and timing of first food consumption (2 items), food
cravings and control over eating behavior both before bedtime (2 items) and during nighttime awakenings (2 items),
percentage of food consumed after dinner (1 item), initial insomnia (1 item), frequency of nocturnal awakenings and
ingestion of food (3 items), and mood disturbance (2 items), and awareness of nocturnal eating episodes (1 item). The
NEQ and its scoring instructions are included in the Appendix.
The current items appeared in a monograph (Allison, Stunkard, & Thier, 2004) and the most recent version of the
WALI (Wadden & Foster, 2006). Nine items of the NEQ presented in the WALI (excluding the mood and awareness of
nighttime eating items) were recently evaluated by Van der Wal, Waller, Klurfeld, McBurney, and Dhurandhar (2005).
They found that including more NES symptoms at each step of an interview was positively correlated with the NEQ
score. Vander Wal and colleagues concluded that questions assessing the full range of behaviors associated with NES
were needed in order to identify a psychological disorder, not just poor eating habits in the evening.
The psychometric properties of the NEQ were evaluated in three studies. The first study examined the factor
structure and internal consistency of the NEQ, as administered via the Internet to 1980 persons who inquired about
NES. The second study examined convergent validity for the NEQ among a sample of 81 participants in a study
designed to characterize the behavioral and psychological features of NES. The third study examined discriminant
validity among a sample of 194 bariatric surgery candidates. The website used in the first study was approved by the
University of Pennsylvania's Institutional Review Board (IRB), and participants were informed that their responses

64

K.C. Allison et al. / Eating Behaviors 9 (2008) 6272

would be used in research. Written informed consent was granted by participants in Studies 2 and 3, which were also
approved by the university's IRB.
1. Study 1: factor structure and internal consistency of the NEQ with an Internet sample
The purpose of Study 1 was to examine the factor structure and internal consistency of the NEQ in a large sample of
geographically diverse participants with varying degrees of reported NES symptoms.
1.1. Method
1.1.1. Participants and procedure
Participants included 1980 adults (demographics presented in Table 1) who completed the NEQ online (http://www.
med.upenn.edu/weight/nighteating.shtml). Surveys were collected between May 2002 and April 2005. All 50 states
and Washington, D.C., 10 Canadian provinces, and 75 additional countries were represented; only 12% were from
Philadelphia.
1.1.2. Measures
Participants completed the NEQ and provided self-reported gender, height, and weight; the latter two were used to
calculate body mass index (BMI; kg/m2). Ethnicity was not assessed in this sample because of concern that participants
may have been less likely to complete the survey if asked to provide it.
The NEQ total score was calculated by reverse coding items 1, 4, and 14, and summing all items, except item 13.
Item 13 does not assess a degree of NES symptomatology; rather, it is included in the questionnaire as a rule out for
sleep-related eating disorder (Schenck & Mahowald, 1994). As such, it is not appropriate to include item 13 in the total
score. The NEQ total score provides a range from 0 to 52 points.
1.2. Results
The group's mean NEQ score was 33.1 7.5, with a range from 2 to 49. A participant may not have answered all of
the items, because the NEQ was designed with the stop criteria described in the scoring instructions (e.g., if no
awakenings are present, the respondent stops and does not answer questions regarding nocturnal ingestions; see
Appendix 1). Therefore, scores of zero were imputed for these unanswered items. Item 13, assessing awareness of
nocturnal ingestions, was excluded from analyses because it is excluded from the total score.
The factor structure was initially examined using principal components analysis with promax rotation; eigenvalues
greater than one were extracted. Item loadings of 0.40 were considered to load on a factor. A four-factor structure
explaining 67.3% of the variance was obtained. Item loadings are shown in Table 2. The first factor (eigenvalue = 4.2)
was comprised of five questions assessing nocturnal ingestions, the second factor (eigenvalue = 2.1) included three
questions assessing evening hyperphagia, the third factor (eigenvalue = 1.4) was comprised of three questions assessing
morning anorexia and the percent of total daily food intake consumed after dinner, and the final factor (eigenvalue = 1.1)
included three questions assessing mood and sleep disturbance. Item five, assessing the percent total daily food intake
consumed after dinner, cross-loaded on Factors 2 and 3. Cronbach's alpha for the total scale was .70. Factor one
Table 1
Demographic characteristics for participants in all studies
Variables

Age (years) M (SD)

BMI (kg/m2)
% Female
Ethnicity
% Caucasian
% African American
% Other/did not specify

Internet sample

NES outpatient sample

Bariatric surgery sample

N = 1980

N = 81

N = 194

39.1 (12.3)
29.6 (7.9)
82.8
Unknown

42.7 (12.4)
31.8 (8.3)
75.6

44.0 (10.7)
50.4 (8.0)
82.5

61.7
35.8
2.5

68.8
25.0
6.2

K.C. Allison et al. / Eating Behaviors 9 (2008) 6272

65

Table 2
Study 1: factor structure, item loadings, and item-total correlations for items 114 of the NEQ using principal components analysis (not including
item 13; N = 1980)
Item

Nocturnal ingestions

Evening hyperphagia

Morning anorexia

Mood/sleep

Correlation (r) with total score

1
2
3
4
5
6
7
8
9
10
11
12
14
Total variance explained

.04
.08
.08
.02
.11
.03
.13
.10
.87
.94
.88
.94
.86
32.3%

.09
.12
.87
.88
.41
.09
.16
.04
.04
.04
.00
.05
.05
16.0%

.80
.85
.12
.07
.46
.07
.20
.11
.05
.04
.00
.02
.02
10.8%

.08
.03
.04
.00
.10
.62
.73
.60
.00
.02
.02
.03
.02
8.2%

.30
.31
.17
.23
.43
.36
.22
.43
.67
.73
.73
.72
.72

Note. p b 0.001. Items highlighted in bold load on each factor.

(nocturnal ingestions) was the strongest, with an alpha of .94; It was followed by factor two (evening hyperphagia;
alpha = .65), factor 3 (morning anorexia; alpha = .57), and factor 4 (mood/sleep; alpha = .30).
A series of confirmatory factor analyses using the entire sample was conducted to assess the appropriateness of a total
score. To answer this question, three models were compared: 1) a single factor model for which all items were considered
indicators of a single global factor, 2) a correlated-factors model in which items were assigned to the Nocturnal Ingestion,
Evening Hyperphagia, Morning Anorexia, or Mood/Sleep factor and allowed to correlate freely, and 3) a second-order
model in which items were assigned to the four factors, and factors were considered to be part of a higher-order Night
Eating Syndrome construct. The Comparative Fit Index (CFI), TuckerLewis Index (TLI), Root Mean Square Error of
Approximation (RMSEA), and the Standardized Root Mean Residual (SRMR) were used to evaluate the models.
Model one, for which all items were considered indicators of a global factor, was not a good fit: CFI = .724,
TLI = .669, RMSEA = .148, and SRMR = .116. The second model, for which items were assigned to the four factors and
allowed to correlate freely, was an adequate fit: CFI = .945, TLI = .939, RMSEA = .064, and SRMR = .058. The third,
higher-order, model was also an adequate fit (CFI = .954, TLI = .940, RMSEA = .063, and SRMR = .060). The ratio of
the Chi Squares for Model 2 to Model 3 was 98.93. This suggests that nearly 99% of the covariation among factors can
be accounted for by the higher-order factor, Night Eating Syndrome. This finding suggests that the use of a total score
from the NEQ is appropriate.
2. Study 2: validity of the NEQ in a sample of outpatients identified with NES
The convergent validity of the NEQ was assessed in a sample of outpatients diagnosed with NES. Because of the
multifaceted nature of NES, the NEQ was hypothesized to exhibit convergent validity with measures of caloric intake,
eating patterns and attitudes, hunger patterns, mood, sleep, and stress. The convergent validity for each of the four
factors was also examined. In addition, normative NEQ scores for persons diagnosed with NES were established.
2.1. Method
2.1.1. Participants and procedures
Participants were 81 adults enrolled in an outpatient study examining the eating, sleeping, and psychological factors
characterizing NES (Table 1). Participants were recruited via television and newspaper advertisements seeking persons
suffering from NES. Exclusion criteria included: pregnancy, current night shift work, current participation in a weight
reduction program, diabetes, sleep apnea, lifetime history of bipolar disorder or psychotic disorder, substance abuse or
dependence within the past three months, or current major depression of greater than moderate severity as diagnosed on
the Structured Clinical Interview for DSM Diagnosis (SCID I/P; First, Spitzer, Gibbon, Williams, & Benjamin, 1996).
Persons with current anorexia nervosa or bulimia nervosa were also excluded, but those with binge eating disorder

66

K.C. Allison et al. / Eating Behaviors 9 (2008) 6272

(BED) were enrolled; 17.3% had a lifetime diagnosis of BED (9.9%) or subthreshold BED (7.4%) as diagnosed by the
Eating Disorder Examination (Fairburn & Cooper, 1993) and SCID.
2.1.2. Measures
The Night Eating Syndrome History and Inventory (NESHI; unpublished semi-structured interview) was used in
conjunction with food and sleep records to identify persons with NES. The NESHI included questions about the
schedule and amount of food intake throughout the 24-hour day, history of NES symptoms, sleeping routine, mood
symptoms and life stressors, weight and diet history, and previous treatment strategies for NES. Detailed 24-hour food
records were completed by participants for 10 consecutive days; only days 39 were included in the analyses (days 12
were practice days and day 10 lacked data for nighttime ingestions) (O'Reardon et al., 2004). Food data were analyzed
by a research dietitian using ESHA Food Processor, version 8.0. The average daily percentage of food intake consumed
after 6:00 p.m., not including dinner, was calculated (percentage of calories consumed after dinner). Participants
recorded morning hunger ratings on 100 mm visual analog scales (0 = not at all hungry to 100 = extremely hungry)
before their first meal of the day.
Participants were included in the NES group if their diaries revealed 1) evening hyperphagia (i.e., consuming 25%
of total daily calories after dinner) and/or 2) nocturnal awakenings with ingestion of food 3 times per week. The
average percentage of caloric intake consumed after the evening meal for this group was 35.7 10.9% and they
averaged 7.3 6.1 nocturnal ingestions per week.
The evening hyperphagia criterion was determined based on the convergence of two factors. First, diary records of
overweight and obese persons without NES (controls) revealed that they consumed 10 7% of their intake after the
evening meal (O'Reardon et al., 2004). Two standard deviations above their mean, therefore, reached 24%. Second,
one standard deviation below the mean percentage of evening intake for the NES group was 24.8%. Thus, the
consumption of 25% of the daily caloric intake after dinner was judged to be abnormally large. There was no
precedent for an exact number for the nocturnal ingestion criterion. Our experience with patients led us to believe that
the cut-off of 3 episodes per week represented a clinically significant number of events.
In addition to the NEQ, participants' eating attitudes, eating behavior, and body image concerns were assessed using
two measures: the Eating Inventory (EI; Stunkard & Messick, 1985) and the Eating Disorders Examination, 12th Ed.
(EDE; Fairburn & Cooper, 1993). The EI is a 51-item self-report scale with a three factor structure and good internal
consistency. Alpha values for the three factors are: Cognitive Restraint (.93), Disinhibition (.91), and Perceived Hunger
(.85). The EDE is a semi-structured clinical interview that assesses eating attitudes and behavior, binge eating and
compensatory behavior (e.g., laxative use), and body image disturbance for the previous 28-days. It yields four
subscales: Dietary Restraint, Eating Concern, Shape Concern, and Weight Concern. Discriminant validity between
eating disordered patients and normal controls has been established (Fairburn & Cooper, 1993).
Sleep behavior was assessed with the Pittsburgh Sleep Quality Index (PSQI; Buysse, Reyonlds, Monk, Berman, &
Kupfer, 1989). The PSQI measures sleep quality, latency, duration, efficiency, disturbances, daytime dysfunction, and
sleep medication use. A global score is obtained by summing the subscales and is used to classify good versus bad
sleepers. The internal consistency is good (alpha = .83).
Symptoms of depressed mood were assessed with the Beck Depression Inventory II (BDI; Beck, 1996), a widelyused 21-item self-report measure. The average alpha is .86 for psychiatric patients and .81 for non-psychiatric patients.
The BDI is significantly correlated with other measures of depressed mood and it has been shown to discriminate
depressed mood and anxiety (Beck, Steer, & Garbin, 1988).
The 10-item Perceived Stress Scale (PSS; Cohen, Kamarck, & Mermelstein, 1983) was used to measure the degree
to which situations in life are perceived as stressful. Items are rated on a 04 Likert scale and focus on appraisals of
thoughts and feelings over the past month.
Participants' height and weight were measured and BMI calculated. Participants were classified as normal weight
(BMI 17.524.9 kg/m2), overweight (BMI 25.029.9 kg/m2), and obese (BMI 30.0 kg/m2).
2.2. Results
2.2.1. Convergent validity
Significant Pearson product moment correlations were found between the NEQ total score and (1) percentage of
daily calories consumed after dinner, (2) frequency of nocturnal ingestions, (3) EDE subscales, including Dietary

K.C. Allison et al. / Eating Behaviors 9 (2008) 6272

67

Restraint, Eating Concern, Shape Concern and the EDE global score, (4) PSQI, (5) BDI-II, and (6) PSS (Table 3). The
NEQ total score did not significantly correlate with (1) morning hunger ratings, (2) EI Cognitive Restraint,
Disinhibition, and Hunger subscales and (3) EDE Weight Concern subscale.
As reported in Table 3, the Nocturnal Ingestion factor was positively correlated with frequency of nocturnal
ingestions and morning hunger ratings that were recorded in the food diaries. This factor was not related to the total
PSQI score, but was related to two of its subscales: Subjective Sleep Quality (r = .26, p b 0.05) and Habitual Sleep
Efficiency (r = .25, p b 0.05). The Evening Hyperphagia factor was positively correlated with most measures, including,
the EDE, the EI subscales, PSQI, BDI, and PSS. The Morning Anorexia factor was positively correlated with the EI
subscales, the EDE, and the percentage of calories consumed after dinner. This factor was significantly negatively
correlated with measures of morning hunger and frequency of nocturnal ingestions. Finally, the Sleep/Mood Factor was
positively correlated with the percentage of calories consumed after dinner, subscales of the EDE, PSQI, BDI, and PSS.
2.2.2. Normative data
The mean NEQ total score was 32.4 6.8 (range 12 to 45). Mean scores for the four factors are presented in
Table 4. Analysis of variance with post hoc analyses using Bonferroni corrections were used to compare NEQ scores
across the BMI categories of normal weight (BMI 24.9 kg/m2), overweight (BMI 25.029.9 kg/m2), and obese
(BMI 30.0 kg/m2). For the total NEQ score, an omnibus significant difference was found for weight class (F(2,78) =
3.5, p = 0.05), with normal weight participants scoring higher than obese participants (p = 0.05). Overweight
participants did not differ from either group. When examining the factor scores, omnibus significant differences were
found only for the Nocturnal Ingestion factor (F(2,76) = 5.4, p = 0.01), with obese participants scoring significantly
lower than normal weight participants (p = 0.02) and overweight participants (p = 0.03; Table 4).
3. Study 3: validity and normative data in a sample of bariatric surgery candidates
Positive predictive value and discriminant validity were examined by comparing NEQ scores for bariatric surgery
candidates with NES to those of candidates without NES.
3.1. Method
Participants included 194 consecutively assessed men and women undergoing routine preoperative psychological/
behavioral evaluations for bariatric surgery (sample described previously in Allison et al., 2006; Table 1). As part of
Table 3
Study 2: mean scores for all measures; correlations between NEQ and measures of eating attitudes and behavior, sleep, mood, stress, and quality of
life in outpatients diagnosed with NES
Measure

Mean (SD) NEQ total Factor 1

Factor 2

Factor 3

Factor 4

Nocturnal ingestion Evening hyperphagia Morning anorexia Mood/sleep

Mean % calories consumed after
6:00 pm, not including dinner
Total # nocturnal ingestions/week
Mean hunger rating before the 1st meal
EI Cognitive Restraint
EI Disinhibition
EI Hunger
EDE Dietary Restraint
EDE Eating Concern
EDE Weight Concern
EDE Shape Concern
EDE global score
PSQI total score
BDI
PSS

35.7 (10.9)

.31

.13

.17

.33

.24

7.3 (6.1)
37.3 (24.5)
8.6 (5.2)
10.4 (3.7)
8.2 (3.4)
1.7 (1.3)
1.1 (1.2)
2.2 (1.3)
2.7 (1.6)
1.9 (1.1)
8.3 (4.1)
16.0 (10.7)
19.2 (6.6)

.34
.02
.08
.07
.19
.35
.36
.20
.27
.37
.32
.35
.33

.50
.25
.03
.15
.03
.12
.09
.12
.07
.00
.12
.09
.07

.03
.07
.08
.23
.35
.28
.36
.24
.38
.41
.26
.32
.33

.26
.54
.25
.23
.02
.42
.18
.10
.14
.28
.10
.05
.06

.21
.08
.05
.09
.07
.08
.27
.33
.28
.31
.50
.56
.52

Note. p b 0.05; p b 0.01; EI = Eating Inventory; EDE = Eating Disorders Examination; PSQI = Pittsburgh Sleep Quality Index; BDI = Beck
Depression Inventory-II; PSS = Perceived Stress Scale.

68

K.C. Allison et al. / Eating Behaviors 9 (2008) 6272

Table 4
Mean (SD) of the NEQ total score and factor scores for all samples
Study

NEQ total

Factor 1

Factor 2

Factor 3

Factor 4

Nocturnal ingestion

Evening hyperphagia

Morning anorexia

Mood/sleep

Internet sample
N = 1980

33.1 (7.5)

14.9 (5.9)

7.9 (2.4)

6.7 (2.0)

4.8 (2.8)

NES outpatient sample

Total sample (N = 81)
Normal weight (N = 14)
Overweight (N = 26)
Obese (N = 41)

32.4 (6.8)
36.0 (6.8)a
33.1 (4.3)
31.1 (7.2)b

15.2 (4.6)
17.4 (3.0)a
16.6 (2.8)a
13.9 (5.1)b

7.8
7.6
7.9
7.9

6.7 (2.2)
7.1 (2.4)
6.4 (2.0)
6.9 (2.1)

5.4 (2.8)
6.1 (3.0)
5.4 (2.5)
5.1 (3.0)

Bariatric surgery sample

Obese controls (N = 175)
Night eaters (N = 19)

16.0 (6.3)a
26.2 (8.1)b

3.0 (3.6)a
9.3 (5.2)b

4.5 (1.9)a
5.7 (2.2)b

4.2 (2.6)
5.2 (2.9)

(2.0)
(2.4)
(1.9)
(2.0)

5.4 (2.4)a
7.8 (2.5)b

Note. Statistically significant comparisons are indicated by different superscripts.

this comprehensive evaluation (Wadden & Sarwer, 2006), all surgery candidates completed the WALI (Wadden &
Foster, 2006), a comprehensive self-report instrument (weight and dieting histories, eating and activity habits, and
psychosocial status) which contains a copy of NEQ. Participants endorsing symptoms of night eating on the NEQ were
interviewed with the NESHI to establish a diagnosis of NES.
3.2. Results
Of the 194 surgery candidates, 19 were diagnosed with NES. The 175 candidates without NES were compared to
those with NES using analysis of variance. NEQ scores for patients without NES were significantly lower than those of
night eaters (F(1,192) = 41.7, p b 0.001; Table 4). The PPV of the NEQ at a score of 25 or greater was low at 40.7%. The
PPV increased to 72.7% at a score of 30 or greater. Negative predictive value was high for cut scores of both 25 (95.2%)
and 30 (94.0%).
Scores on the four NEQ factors were compared between surgery candidates with and without NES. Those without
NES, had significantly lower scores on the Nocturnal Ingestion factor (F(1,192) = 47.3, p b 0.001), Evening Hyperphagia
factor (F(1,192) = 17.5, p b 0.001), and Morning Anorexia factor (F(1,192) = 6.3, p = 0.01). Groups did not differ on the
Mood/Sleep factor (F(1,192) = 2.5, p = 0.12). Means are presented in Table 4.
4. General discussion
The NEQ was designed as a measure of severity for symptoms related to NES. Although it has been used in NES
research, there are limited data on its psychometric properties (Van der Wal et al., 2005). The current studies examined
the factor structure, internal consistency, and validity of the NEQ, as well as established normative data for persons
with NES and those with extreme obesity without NES.
Study 1 provides support for both the proposed four-factor scale structure and for the existence of a higher-order
factor, presumably night eating symptomatology; this is consistent with the current conceptualization of the disorder
(Allison & Stunkard, 2005; Birketvedt et al., 1999; Gluck et al., 2001; O'Reardon et al., 2004). The total scale
exhibited adequate reliability (alpha = .70). The Nocturnal Ingestion factor accounted for the most variance (32.3%),
followed by the Evening Hyperphagia factor (16.0%). These factors represent the core features of NES. Not
surprisingly, item 5, which assesses the percent of daily food intake consumed after dinner, cross-loaded on both the
Evening Hyperphagia and Morning Anorexia factors. This item has clear face validity on the Evening Hyperphagia
factor. It is also behaviorally related to Morning Anorexia, for which two hypotheses can be proposed: (1) evening
overeating contributes to lack of morning appetite, and (2) lack of morning appetite shifts food intake later in the day
resulting in greater evening food consumption. It thus makes conceptual sense to include it on both factors.
Future research is required to understand the relationship among overeating at night, morning hunger levels, and
subsequent circadian patterns of food intake more fully. At present, one can only speculate that the temporal

K.C. Allison et al. / Eating Behaviors 9 (2008) 6272

69

relationship between evening overeating and morning hunger is dependent on individual factors, such as premorbid
dietary restraint or poor sleep hygiene.
The average total NEQ score for participants in the Internet study (Study 1) was notably high for an undiagnosed
sample and similar to the outpatient night eaters (Study 2). These high scores are not surprising because our web site is
one of the few resources for persons suffering from NES. Consequently, it continues to draw persons who are distressed
about their night eating and are seeking resources or treatment for NES.
In Study two, the NEQ total score was positively related to measures of food intake, eating attitudes and behaviors,
sleep, mood, and stress, but not to morning hunger ratings. The NEQ was related to EDE scales but not the EI scales.
From our clinical experience, this difference may have occurred because many persons with NES have good control
and little hunger during the day. It is possible that they answered the EI questions with their daytime behaviors in mind.
The EDE measures core thoughts and beliefs averaged over the previous month, so it is likely that the eating attitudes
and behaviors associated with night eating were detected through this method. The diversity of behaviors assessed by
the measure may have contributed to the lack of a relationship between morning hunger and the total score. This
finding also supports the results of an item response theory analysis for NES that showed morning anorexia to be a
common, but not a significantly informative diagnostic symptom of NES (Engel et al., 2004).
There were differences in the self-reported severity of NES symptomatology between normal weight and obese
night eaters, which was particularly true for the Nocturnal Ingestions factor. A study comparing normal weight and
obese night eaters has demonstrated that normal weight night eaters endorse more problems with nocturnal ingestions
(Marshall et al., 2004), which would yield more points on the scale and account for the higher scores. The reason for
this difference deserves further investigation. It may represent higher levels of daytime dietary restraint for the normal
weight night eaters, leading to more frequent nocturnal ingestions.
As expected, Study 3 showed that NEQ scores were higher for extremely obese night eaters seeking bariatric
surgery compared to surgery candidates without NES.
4.1. Issues related to use
The NEQ is a short assessment instrument that can be efficiently administered in a variety of settings. This instrument
represents a first step in helping researchers acquire more data on NES to inform future decisions about its diagnostic status.
Specifically, it is useful in recognizing problems with evening overeating, nocturnal ingestions, absence of morning
hunger, and mood and sleep disturbance. Interview experience, however, suggests that items 1 (morning anorexia) and 5
(percentage of total energy consumed after dinner) are difficult for persons to answer accurately. Morning anorexia may be
more reliable when assessed prospectively in diaries or through ecological momentary assessment, a technique in which
participants record ratings in the moment on a device such as a personal digital assistant, rather than retrospectively. This
methodology, however, has been applied to the area of night eating only once (Boseck et al., 2007). Evening hyperphagia
may be unreliable because 1) people may have difficulty estimating the amount of food consumed after the evening meal
separately from that consumed during the evening meal and 2) persons, particularly those who are obese, generally are not
very accurate in their self-reported food intake (Lindroos, Lissner, & Sjostrom, 1993). Thus, an interview and momentary
assessment of food intake, or 24-hour food recall, is needed in most cases to assess these constructs more accurately.
Based on our findings, we believe that the NEQ is an acceptable tool for assessing NES symptomatology. The use of the
NEQ may be similar to that of the Beck Depression Inventory (Beck, 1996) which has proven useful in assessing the degree
of depressed mood overall, or specifically for somatic or cognitive symptoms, but cannot on its own diagnose a major
depressive episode. Similarly, the NEQ total score, which was confirmed to be a valid indicator of a higher-order construct
of night eating in this study, may be used for an index of severity, or the factor scales may be examined separately if
particular aspects of NES are of interest. We do not believe, however, that it should be used alone to establish the diagnosis.
When using the NEQ as a screening instrument, a cut score of 30 was more appropriate than a cut score of 25 for the
bariatric surgery population with PPV's of 72.2% and 40.7%, respectively. Prevalence rates of NES for bariatric surgery
candidates have ranged from moderate (Allison et al., 2006) to high (Hsu et al., 1996), but it remains unknown if night eating
among this group differs from NES in the general population. A previous study which used the NEQ (including item 13) to
estimate the prevalence of NES among 399 psychiatric outpatients (diagnoses were confirmed by interview) reported a PPV
of 52% using a cut score of 25 and 68% at a cut score of 30 (Lundgren et al., 2006). We recalculated the PPVs excluding item
13, and the utility of the NEQ was greatly increased: 62% at a cut score of 25 and 77% at a cut score of 30. Cutoff scores for
the NEQ should also be normed for different populations in future studies. Based on the current results, if the purpose is to

70

K.C. Allison et al. / Eating Behaviors 9 (2008) 6272

cast a wide net to screen for possible NES cases, 25 may be appropriate, but if the purpose is to reduce the number of false
positives, 30 would be more appropriate. Persons with only evening hyperphagia and no nocturnal awakenings or ingestions
will not make either of these cut scores. So if that is a subgroup of interest, it would be best to examine the Evening
Hyperphagia factor separately.
Though its use as a measure of severity has been established, the NEQ does not assess the degree of perceived
distress about night eating symptomatology or the degree of daily impairment in functioning as a result of the
syndrome. We would like to propose two items to be added to the end of the NEQ to assess this (Appendix 2). These
items have not been evaluated, but providing them as items in need of further research would further help to
characterize this disorder. If a positive NEQ screen is obtained, history of distress and/or impairment of functioning
should also be evaluated as part of a clinical interview. However, those who have had NES for a long time may
habituate to the behavior and not report high levels of acute distress. Others who exhibit the night eating symptoms may
not ever have experienced associated distress or functional impairment; in those cases, they may be exhibiting an
unusual pattern of eating, but do not have the necessary features of a psychiatric disorder.
In summary, our findings suggest that the NEQ is an efficient, valid measure of severity of NES. Although the
testing of the NEQ did not occur within an iterative process, this study's strengths include the use of three, diverse
samples for testing the NEQ and the breadth of measures available for inclusion in the analyses of convergent
validity. Finally, the NEQ's assets lie in identifying persons who experience the underlying, often secretive
symptoms of this disorder, and it provides an assessment of symptoms that capture the unique, multifaceted nature
of NES.
Acknowledgements
Support for this paper was provided by NIH/NIDDK grant RO1 DK 056735 and in part by K23 DK60023 (DBS).
Thanks go to Lisa Basel-Brown, R.D. of the University of Pennsylvania's General Clinical Research Center, supported
by grant RR00040, for processing energy intake information from participants' food diaries. Thanks are also extended
to J. Kevin Thompson, PhD of the University of South Florida for his valuable feedback regarding this manuscript.
Parts of this manuscript were presented at the Annual Meeting of the Association for Behavioral and Cognitive
Therapies, Washington, D.C., November 2005.
Appendix A
Night Eating Questionnaire
Directions: Please circle ONE answer for each question.
1. How hungry are you usually in the morning?
0
1
2
3
4
Not at all
A little
Somewhat
Moderately
Very
2. When do you usually eat for the first time?
0
1
2
3
4
Before 9 am 9:01 to 12 pm
12:01 to 3 pm
3:01 to 6 pm
6:01 or later
3. Do you have cravings or urges to eat snacks after supper, but before bedtime?
0
1
2
3
4
Not at all
A little
Somewhat
Very much so
Extremely so
4. How much control do you have over your eating between supper and bedtime?
0
1
2
3
4
None at all A little
Some
Very much
Complete
5. How much of your daily food intake do you consume after suppertime?
0
1
2
3
4
0%
125%
2650%
5175%
76100%
(none)
(up to a quarter)
(about half)
(more than half)
(almost all)
6. Are you currently feeling blue or down in the dumps?
0
1
2
3
4
Not at all
A little
Somewhat
Very much so
Extremely

K.C. Allison et al. / Eating Behaviors 9 (2008) 6272

71

Appendix A (continued )
7. When you are feeling blue, is your mood lower in the:
0
1
2
3

____check if your mood does not change during the

day

Early
Late morning
Afternoon
Early evening
Late evening/
morning
nighttime
8. How often do you have trouble getting to sleep?
0
1
2
3
l4
Never
Sometimes
About half the Usually
Always
time
9. Other than only to use the bathroom, how often do you get up at least once in the middle of the night?
0
1
2
3
4
Never
Less than once a About once a More than once a Every night
week
week
week
IF 0 on #9, PLEASE STOP HERE
10. Do you have cravings or urges to eat snacks when you wake up at night?
0
1
2
3
4
Not at all
A little
Somewhat
Very much so
Extremely so
11. Do you need to eat in order to get back to sleep when you awake at night?
0
1
2
3
4
Not at all
A little
Somewhat
Very much so
Extremely so
12. When you get up in the middle of the night, how often do you snack?
0
1
2
3
4
Never
Sometimes
About half the Usually
Always
time
IF 0 on #12, PLEASE STOP HERE
13. When you snack in the middle of the night, how aware are you of your eating?
0
1
2
3
4
Not at all
A little
Somewhat
Very much so
Completely
14. How much control do you have over your eating while you are up at night?
0
1
2
3
4
None at all A little
Some
Very much
Complete
How long have your current difficulties with night eating been going on?
_____ mos. _____ years
2006 by K.C. Allison and A.J. Stunkard.

Scoring directions: Questions 19, which focus on variables occurring before sleep onset, are answered by all
participants. Stop criteria were built into the remaining items. Questions 1012 are answered by participants who wake
up in the middle of the night and score N 0 on question 9. Similarly, questions 13 and 14 are answered by participants
who eat upon awakening and score N 0 on item 12.
Items are scored on a 04 Likert scale, with the exception of item 7. It includes the option check here if your mood
does not change during the day, which is scored as a zero. Items 1, 4, and 14 are reverse-scored so that higher values
reflect greater symptomatology. All items except the awareness of nocturnal ingestions question (item 13) are summed
to obtain a global score. Item 13 is included solely as a screening tool to rule out the presence of the parasomnia SleepRelated Eating Disorder (SRED; Schenck & Mahowald, 1994). In contrast to SRED, NES sufferers may be groggy, but
they are aware of their night eating behaviors (Allison et al., 2004).
Appendix B
Proposed additional questions for the NEQ
15. How upsetting is your night eating to you?
0
1
Not at all
A little
16. How much has your night eating affected your life?
0
1
Not at all
A little

2
Somewhat

3
Very much so

4
Extremely

2
Somewhat

3
Very much so

4
Extremely

72

K.C. Allison et al. / Eating Behaviors 9 (2008) 6272

References
Adami, G. F., Meneghelli, A., & Scopinaro, N. (1999). Night eating and binge eating disorder in obese patients. International Journal of Eating
Disorders, 25, 335338.
Allison, K. C., Crow, S. J., Reeves, R. R., West, D. S., Foreyt, J. P., DiLillo, V. G., et al. (2007). The prevalence of binge eating disorder and night
eating syndrome in adults with type 2 diabetes mellitus. Obesity, 15, 12851291.
Allison, K. C., & Stunkard, A. J. (2005). Obesity and eating disorders. Psychiatric Clinics of North America, 28, 5567.
Allison, K. C., Stunkard, A. J., & Their, S. L. (2004). Overcoming night eating syndrome. Oakland, CA: New Harbinger.
Allison, K. C., Wadden, T. A., Sarwer, D. B., Fabricatore, A. N., Crerand, C. E., Gibbons, L. M., et al. (2006). Night eating syndrome and binge eating
disorder among persons seeking bariatric surgery: prevalence and related features. Obesity, 14(Suppl 2), 77S82S.
Aranoff, N. J., Gelibter, A., & Zammit, G. (2001). Gender and body mass index as related to the night eating syndrome in obese outpatients. Journal
of the American Dietetic Association, 101, 102104.
Beck, A. T. (1996). The beck depression inventory II. San Antonio: Harcort Brace.
Beck, A. T., Steer, R. A., & Garbin, M. G. (1988). Psychometric properties of the Beck Depression Inventory: Twenty-five years of evaluation.
Clinical Psychology Review, 8, 77100.
Birketvedt, G. S., Florholmen, J., Sundsfjord, J., Osterud, G., Dinges, D., Bilker, W., et al. (1999). Behavioral and neuroendocrine characteristics of
the night-eating syndrome. Journal of the American Medical Association, 282, 657663.
Boseck, J. J., Engel, S. G., de Zwaan, M., Allison, K. C., Crosby, R. D., & Mitchell, J. E. (2007). The application of ecological momentary assessment
to the study of night eating. International Journal of Eating Disorders, 40, 271276.
Buysse, D. J., Reynolds, C. F., III, Monk, T. H., Berman, S. R., & Kupfer, D. J. (1989). The Pittsburgh sleep quality index: A new instrument for
psychiatric practice and research. Psychiatry Research, 28, 193213.
Cer-Bjrk, C., Andersson, I., & Rssner, S. (1983). Night eating and nocturnal eating two different or similar syndromes among obese patients?
International Journal of Obesity, 25, 365372.
Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A global measure of perceived stress. Journal of Health and Social Behavior, 24, 385396.
Engel, S. G., Allison, K. C., Crosby, R. D., de Zwaan, M., O'Reardon, J. P., Wonderlich, S. A, et al. (2004, October). Evaluation of diagnostic criteria
for night eating syndrome using item response theory analysis. Presented at the 10th Annual Meeting of the Eating Disorders Research Society,
Amsterdam, Netherlands.
Fairburn, C. G., & Cooper, Z. (1993). The eating disorder examination. In C. G. Fairburn & T. Wilson (Eds.), Binge eating: Nature, assessment, and
treatment (pp. 317360). (12th ed.) New York: Guilford.
First, M. B., Spitzer, R. L., Gibbon, M., Williams, J. B. W., & Benjamin, L. (1996). Structured Clinical Interview for DSM-IVPatients edition (with
psychotic screen, version 2.0. New York: Biometrics Research Department, New York State Psychiatric Institute.
Gluck, M. E., Geliebter, A., & Satov, T. (2001). Night eating syndrome is associated with depression, low self-esteem, reduced daytime hunger, and
less weight loss in obese outpatients. Obesity Research, 9, 264267.
Hsu, L. K. G., Betancourt, S., & Sullivan, S. P. (1996). Eating disturbances before and after vertical banded gastroplasty: A pilot study. International
Journal of Eating Disorders, 19, 2334.
Lindroos, A. -K., Lissner, L., & Sjostrom, L. (1993). Validity and reproducibility of a self-administered dietary questionnaire in obese and non-obese
subjects. European Journal of Clinical Nutrition, 47, 461481.
Lundgren, J. D., Allison, K. C., Crow, S., Berg, K. C., Galbraith, J., Martino, N. S., et al. (2006). Prevalence of the night eating syndrome in a
psychiatric population. American Journal of Psychiatry, 163, 156158.
Marshall, H. M., Allison, K. C., O'Reardon, J. P., Birketvedt, G., & Stunkard, A. J. (2004). Night eating syndrome among nonobese persons.
International Journal of Eating Disorders, 35, 217222.
Napolitano, M. A., Head, S., Babyak, M. A., & Blumenthal, J. A. (2001). Binge eating disorder and night eating syndrome: psychological and
behavioral characteristics. International Journal of Eating Disorders, 30, 193203.
O'Reardon, J. P., Ringel, B. L., Dinges, D. F., Allison, K. C., Rogers, N. L., Martino, N. S., et al. (2004). Circadian eating and sleeping patterns in the
night eating syndrome. Obesity Research, 12, 17891796.
Rand, C. S. W., Macgregor, M. D., & Stunkard, A. J. (1997). The night eating syndrome in the general population and among post-operative obesity
surgery patients. International Journal of Eating Disorders, 22, 6569.
Schenck, C. H., & Mahowald, M. W. (1994). Review of nocturnal sleep-related eating disorders. International Journal of Eating Disorders, 15, 343356.
Striegel-Moore, R. H., Franko, D. L., Thompson, D., Affenito, S., & Kraemer, H. (2006). Night eating: Prevalence and demographic correlates.
Obesity, 14, 139147.
Stunkard, A. J., Grace, W. J., & Wolf, H. G. (1955). The night-eating syndrome: A pattern of food intake among certain obese patients. American
Journal of Medicine, 19, 7886.
Stunkard, A. J., & Messick, S. (1985). The three-factor eating questionnaire to measure dietary restraint, disinhibition and hunger. Journal of
Psychosomatic Research, 29, 7183.
Van der Wal, J. S., Waller, S. M., Klurfeld, D. M., McBurney, M. I., & Dhurandhar, N. V. (2005). Night eating syndrome: Evaluation of two screening
instruments. Eating Behaviors, 6, 6373.
Wadden, T. A., & Foster, G. D. (2001). Weight and lifestyle inventory. Philadelphia: University of Pennsylvania.
Wadden, T. A., & Foster, G. D. (2006). Weight and lifestyle inventory. Obesity, 14(Suppl 2), 99S118S.
Wadden, T. A., & Sarwer, D. B. (2006). Behavioral assessment of candidates for bariatric surgery: a patient-oriented approach. Obesity, 14(Suppl 2),
53S62S.