Familial aggregation in the night eating syndrome

Familial Aggregation in the Night Eating Syndrome having an affected first-degree relative were significantly greater than that of a control proband (odds ratio ¼ 4.9, p < drome (NES) affects first-degree relatives included in the model: proband bodymass index (BMI) (kg/m2), proband gen- first-degree relative gender, relationship to proband (i.e., mother, father, or sib- 10 day sleep and food records, the Eating (night eater or control); none was statisti- tured Clinical Interview for DSM IV Axis I ence of NES among first-degree relatives.
A proband predictive model, using logis-tic regression analyses and the general- Keywords: night eating syndrome; familial correlation among observations withinfamilies was used to assess familialaggregation.
A first step in understanding potential genetic and environmental influences is a study of familial The night eating syndrome (NES) is characterized aggregation. Accordingly, the current study exam- as a delay in the circadian pattern of food intake ined the extent to which the NES phenotype affects manifested by evening hyperphagia and/or noctur- first-degree relatives of both night eaters and con- nal awakenings with ingestion of food.1,2 Night eat- ers frequently report morning anorexia, insomnia,and negative mood that worsens during the eve-ning; the onset of NES is often associated with lifestress.3 The etiology of NES is unknown, but genetic influences have been suspected. The response ofNES to sertraline4 suggests that genes regulating serotonin gene may be involved. Additionally, a Night eaters were recruited from printed advertise- recent study found that mice with a mutation in ment, radio talk shows, and television commercials the Clock gene consume significantly more food describing NES. Controls, recruited from printed and than wild-type mice during the light period (when radio advertisements, were sought to approximate the mice usually sleep more and eat less) and become age, gender, body mass, and ethnicity of the night eaters.
Advertisements for both night eaters and controlsincluded the following eligibility requirements: age 18 orolder, not currently involved in an occupation requiring nightshift work, not currently enrolled in a weight reduc- *Correspondence to: Jennifer D. Lundgren, PhD, 3535 Market tion program, and not currently taking any psychotropic Street, Suite 3123, Philadelphia, PA 19104-3309.
medications. Advertisements to recruitment night eaters University of Pennsylvania School of Medicine Weight and specified that they must have difficulties with overeating Eating Disorder Program, Philadelphia, Pennsylvania after dinner and/or with waking up at night to eat; adver- tisements for controls stated that they could not experi- (www.interscience.wiley.com). DOI: 10.1002/eat.20269 International Journal of Eating Disorders 39:4 000–000 2006—DOI 10.1002/eat Participants included 103 adults diagnosed with NES (mean age ¼ 43.0 6 11.3 years; mean BMI ¼ 33.3 6 8.1 As part of a comprehensive study of the characterization kg/m2; percentage white ¼ 62.1; percentage female ¼ of NES, NES and control participants provided informed 71.8) and 42 control participants (mean age ¼ 39.3 6 consent, completed the NEQ and the FHQ, and heights 11.3 years; mean BMI ¼ 36.0 6 6.2 kg/m2; percentage and weights were measured. NES participants were inter- white ¼ 59.5; percentage female ¼ 66.7). Night eater and viewed using the NESHI to assess their NES symptoms.
control participants differed significantly only on BMI Both NES and control participants were interviewed with (p ¼ .05); accordingly, proband BMI was included as a the EDE and kept detailed 24-hour food records for 10 con- secutive days to confirm their diagnostic status.
Participants were diagnoses as ‘NES positive’ if they met the following criteria based on food records and NESHI interview: (1) consuming 25% of total daily calo- The Night Eating Questionnaire (NEQ6) was used ric intake after dinner, and/or (2) nocturnal awakenings to assess the presence of night eating behavior. It is with ingestion of food occurring 3 times/week. Partici- comprised of 14 questions assessing hunger and craving pants must have experienced symptoms for no less than patterns throughout the day, percentage of calories 3 months based on self-report. This operational defini- ingested after suppertime, feelings of depression, insom- tion was modified from previously described criteria1 nia and awakenings, and nocturnal food cravings and because recent data with a larger sample of night eaters and controls found that night eaters consumed 34.6% 6 The Night Eating Syndrome History and Inven- 10.1% of total daily caloric intake after dinner compared tory (NESHI; unpublished semi-structured interview) with controls who consumed 10.0% 6 6.9% of total daily was used in conjunction with food records to establish caloric intake after dinner.2 Based on a figure of 2 SD a diagnosis of NES. The NESHI included questions above the mean for controls, 25% of calories consumed about the schedule and amount of food intake after the evening meal was considered abnormally large.
throughout the 24-hour day, history of NES symptoms, Participants who also met criteria for binge eating disor- sleeping routine, mood symptoms and life stressors, der (BED; NES, n ¼ 20; control, n ¼ 0) were retained in the weight and diet history, and previous treatment strat- analyses because the literature suggests that BED and NES are two distinct conditions, although they may co-occur.9,10 The Eating Disorder Examination (EDE7) was used to assess the presence of any concurrent eating disorder.
It is a semi-structured clinical interview that assesses The proband predictive model described by Hudson dietary restraint, eating concern, and weight and shape et al.11 and Laird and Cuenco12 was used to predict the concern, as well as bingeing and purging behaviors.
binary outcome (affected or not affected) of first-degree The Structured Clinical Interview for DSM IV Axis I relatives. One logistic regression analysis was conducted, Disorders (SCID I8) was used to formally assess the pres- using the generalized estimating equation (GEE) with ence of Axis I psychiatric diagnoses.
exchangeable correlation structure to control for correla-tion among the family members. Status as NES or control Night eaters and controls completed detailed 24-hour proband was used as the predictor; covariates in the food records for 10 consecutive days. The first two days model were proband BMI, proband age, proband gender, were practice days and day 10 did not include nighttime proband ethnicity (white vs. non-white), first-degree rel- data. Accordingly, days 3–9 were used to confirm a diag- ative gender, first-degree relative relationship to the pro- nosis of NES. Food intake was analyzed using ESHA Food band (mother, father, or sibling), and the interaction between proband status (night eater or control) and first- A family history questionnaire (FHQ) was completed degree relative relationship to the proband.
by NES and control probands to assess the presence ofNES among parents, siblings, and offspring. Probandresponses were compared with their self-report duringthe NESHI interview. When discrepancies were encoun-tered and when the probands were unsure about the presence of NES, the first-degree relative was coded‘unaffected.’ The first-degree relatives of NES probands had sig- BMI was calculated from height and weight measured nificantly greater odds of being affected with NES at their interview visits. Normal weight was defined as than the first-degree relatives of control probands 18.5–24.9 kg/m2, overweight was 25–29.9 kg/m2, and (odds ratio [OR] ¼ 4.9; standard error [SE] ¼ 2.9; 95% confidence interval [CI] ¼ 2.0, 12.3; p ¼ .0006).
International Journal of Eating Disorders 39:4 000–000 2006—DOI 10.1002/eat FAMILIAL AGGREGATION IN THE NIGHT EATING SYNDROME None of the covariates was statistically significant Despite these limitations, the study was able to (all p values > .05) including proband BMI, pro- show strong familial aggregation of NES; these results band age, proband gender, proband ethnicity, first- highlight the need for a classic twin study of NES.
degree relative gender, relationship to proband(mother, father, or sibling), and the interaction Support for this work was provided by NIH/NIDDK grant between relationship to proband and proband sta- RO1 DK 056735. Thanks go to Lindon Eaves, MA(Oxon), PhD, DSc, Distinguished Professor of Human Geneticsand Psychiatry, Virginia Commonwealth UniversitySchool of Medicine, for his valuable feedback regardingthe statistical analyses. Thanks are also extended to Wei Yang of the Center for Clinical Epidemiology and Biosta- This is the first study to examine the familial aggre- tistics at the University of Pennsylvania School of Medi- gation of NES. It found that NES was more likely to cine for his assistance with the statistical analyses.
aggregate among family members of night eatersthan of controls, suggesting that NES is familial.
Congruent with our findings, a recent study of German children (5–6 years of age) and their moth-ers by Lamerz et al.13 found that the children of 1. Birketvedt GS, Florholmen J, Sundsfjord J, Osterud B, Dinges D, mothers who engaged in night eating behavior Bilker W, et al. Behavioral and neuroendocrine characteristicsof the night-eating syndrome. JAMA 1999;282:657–663.
were 7.8 times more likely than children of non- 2. O’Reardon JP, Ringel BL, Dinges DF, Allison KC, Rogers NL, Mar- night-eating mothers to show signs of the syn- tino NS, et al. Circadian eating and sleeping patterns in the night eating syndrome. Obes Res 2004;12:1789–1796.
The strengths of the current study, in contrast to 3. Stunkard AJ, Grace WJ, Wolf HG. The night-eating syndrome: a pattern of food intake among certain obese patients. Am J Med the study carried out by Lamerz et al.,13 included the use of an adult sample of probands, parents, and sib- 4. O’Reardon JP, Stunkard AJ, Allison KC. Clinical trial of sertraline ling, diagnosis of probands using food and sleep in the treatment of night eating syndrome. Int J Eat Disord records in conjunction with clinical interview, and robust statistical methods using the generalized esti- 5. Turek FW, Joshu C, Kohsaka A, Lin E, Ivanova G, McDearmon E, et al. Obesity and metabolic syndrome in circadian clock mating equation to control for correlated family data.
mutant mice. Science 2005;308:1043–1045.
Although 20 of the NES probands also met criteria 6. Allison KC, Stunkard AJ, Their SL. Overcoming night eating syn- for BED, only seven of those concurrently diagnosed drome: a step-by-step guide to breaking the cycle. Oakland, reported a family history of NES. It is unlikely that 7. Fairburn CG, Cooper Z. The Eating Disorder Examination (12th familial aggregation can be attributed to BED aggre- ed). In: Fairburn CG, Wilson GT, editors. Binge eating: nature, gation. Several NES probands, however, also met cri- assessment, and treatment. New York: The Guilford Press, 1993, teria for an Axis I disorder; co-aggregation of NES and Axis I disorders (e.g., Major Depressive Disorder) 8. First MB, Spitzer RL, Gibbon M, Williams JBW. Structured Clini- should be examined in future studies.
cal Interview for DSM-IV Axis I Disorders–Patient Edition (SCID-I/P with Psychotic Screen, Although a first step in understanding the etiol- Department, New York State Psychiatric Institute; 1996.
ogy of NES, this study was limited in the following 9. Stunkard AJ, Berkowitz R, Wadden T, Tanrikut C, Reiss E, Young L.
ways. First, the family history method, rather than Binge eating disorder and the night eating syndrome. Int J Obes the family interview method, was used to assess 10. Napolitano MA, Head S, Babyak MA, Blumenthal JA. Binge eat- night eating among first-degree relatives. Because ing disorder and night eating syndrome: psychological and the limitations of the family history method are behavioral characteristics. Int J Eat Disord 2001;30:193–203.
well documented,14 we sought, but we were unsuc- 11. Hudson JI, Laird NM, Betensky RA. Multivariate logistic regres- cessful, in recruiting family members to participate sion for familial aggregation of two disorders. I. Development in verification interviews. Second, some probands of models and methods Am J Epidemiol 2001;153:500–505.
12. Laird NM, Cuenco KT. Regression methods for assessing familial were unsure if family members were affected with aggregation of disease. Statis Med 2003;22:1447–1455.
NES, and were, therefore, considered unaffected.
13. Lamerz A, Kuepper-Nybelen J, Bruning N, Wehle C, Trost-Brink- Third, control probands may have been less likely hues G, Brenner H, et al. Prevalence of obesity, binge eating, to recognize NES among relatives compared with and night eating in a cross-sectional field survey of 6-year-old NES probands, for which the disorder is more sali- children and their parents in a German urban population. JChild Psychol Psychiatry 2005;46:385–393.
ent. Fourth, the study was nonblinded in that the 14. Roy MA, Walsh D, Kendler KS. Accuracies and inaccuracies of interviewer knew the diagnosis of the proband at the family history method: a multivariate approach. Acta Psy- the time the family history was obtained.
International Journal of Eating Disorders 39:4 000–000 2006—DOI 10.1002/eat

Source: http://xn--essstrungen-aargau-h3b.ch/media/archive1/fachpersonen/diagnostik/differentialdiagnostik/nighteatingfamilial.pdf