A Pilot Test of Mood and Circadian Rhythm Mechanisms Driving Binge Eating

Mood and circadian rhythm disruptions are associated with binge eating (BE). BE is a discrete, episodic behavior characterized by 1) eating an objectively large amount of food and 2) experiencing a subjective sense of loss of control. BE episodes are often preceded by negative mood states, and a subtype of individuals with BE has high levels of negative moods. This group has more comorbid psychopathology and a poorer response to treatment. Thus, understanding the role of negative mood is a critical area for research on BE. Individuals with BE demonstrate disruptions in several circadian rhythms, including diurnal meal timing, hormone patterns (e.g., daily cortisol rhythms), and mood variations. The most potent synchronizer of circadian rhythms is light. Thus, exposure to light may explain other phenomena that fluctuate similarly, such as mood and the occurrence of BE. Mood is subject to the influence of light, and BE is also influenced by exposure to bright light. It is unknown whether regulating circadian rhythms via regular exposure to light improves BE through its effects on mood or via changes in other biological or behavioral rhythms. This knowledge can inform the development of treatments targeting biobehavioral mechanisms that maintain BE and indicate for whom this may be most effective. This project aims to test the roles of negative mood and circadian rhythms in the relationship between light exposure and BE and identify subtypes of individuals in whom this effect is strong. The investigators hypothesize that individuals exposed to less natural bright light will experience more frequent BE, more negative mood, and a blunted morning cortisol response. The investigators further hypothesize that manipulating exposure to artificial bright light will reduce the frequency of BE and negative mood and increase the morning cortisol response. Finally, the investigators hypothesize that the effects of this artificial bright light exposure on BE frequency will be mediated by changes in negative mood, which itself will be accounted for by changes in circadian rhythms as indicated by the morning cortisol response. Additionally, the investigators have two moderation hypotheses: that the effects of artificial bright light exposure on BE will be greater for those who fit the high negative mood type than those who do not and that the effects will be greater for individuals with a blunted morning cortisol response at baseline than for those without.

Women (N = 30) who are at least 18 years of age with minimally normal body weight (BMI ≥ 18.5 kg/m2) and exhibit clinically significant binge eating (BE; ≥ 2 BE episodes/week over the previous 3 months) and clinically significant impairment will be eligible to participate. Binge eating will be defined as described in the DSM-5: consuming an objectively large amount of food in a discrete period of time while experiencing a sense of loss of control over eating. The minimum frequency of 2 BE episodes/week will prevent a statistical floor effect from making decreases in BE frequency too small to reliably detect. Further, this frequency should ensure that a sufficient number of BE events occur during the sampling period. A nontreatment-seeking sample is appropriate because the vast majority of individuals with eating disorders do not seek treatment (e.g., 6% of individuals with bulimia nervosa receive mental health services per year). Participants will be recruited from the community using advertisements (e.g., flyers, newspaper ads, listservs, social media). Advertisements will inform interested individuals that they can complete a brief online or phone-based eligibility screen. Individuals who appear eligible on the online screen will be called by a trained research assistant to confirm the screening results. Thus, all prospective participants will be screened by a trained research assistant over the phone. The eligibility screen is modeled off an existing validated tool (i.e., the Eating Disorder Diagnostic Scale) that is modified to assess specific inclusion/exclusion criteria. Individuals will be excluded for several factors that impact hormone function, weight, or appetite or may result in changes in binge eating, including medical conditions and/or medication use in the past 6 weeks months and underweight status, or receiving psychotherapy or psychotropic medication for the 6 weeks prior to study entry. In addition, individuals for whom engaging in the manipulation may be dangerous (e.g., those with diabetes or bipolar disorder) will be excluded. Individuals who appear to meet study eligibility based on the telephone screening and are interested in participating will be scheduled for their baseline appointment. Baseline Assessment The baseline appointment will begin by obtaining informed consent. Participants will be informed of their rights as research participants, the foreseeable risks involved in participation, expected benefits, and other relevant issues (e.g., remuneration schedule).Body Mass Index (BMI) will be assessed via height and weight measured with a stadiometer and digital scale. Inclusion/exclusion criteria will be confirmed via clinical interviews: Eating Disorders Examination (EDE) and Structured Clinical Interview for DSM-5 Axis I disorders (SCID-I). The EDE will be used to confirm current eating disorder symptoms and diagnoses. The SCID-I will be used to assess lifetime history of an eating disorder, and current and lifetime presence of other Axis I disorders that frequently co-occur with eating disorders. The SCID interview will also be used to assess exclusion criteria (e.g., bipolar disorder). Interviews will conducted by doctoral students in clinical psychology who are equivalent to master's degree level mental health professionals. They will be trained by the PI. Their interviews will be audio recorded to assess interrater reliability. Participants will also complete baseline self-report measures at this visit. These include the Clinical Impairment Assessment, the Change in Eating Disorder Symptoms scale, the Pittsburgh Sleep Quality Index, the Center for Epidemiological Studies Depression scale - Revised, the Positive and Negative Affect Schedule, and Subjective Appetite Scales. Prior to leaving this baseline appointment, research assistants will train participants on the use of the therapy light box, the Actiwatch device, and on the proper collection of saliva samples, and they will explain what appointment reminders participants should expect, in addition to querying their preferred method of reminder delivery (e.g., phone call, text, or email). Ecological Momentary Assessment (EMA) At the first day of the 22-day sampling period, EMA data collection will begin to capture daily eating patterns and subjective loss of control during eating and negative mood. Signal-contingent ratings will be semi-randomly sampled six times per day during waking hours. These will assess the strength of negative mood. These ratings will be made when the Actiwatch device alerts participants by vibrating that a rating is being requested. These assessments take only a moment to complete using an ordered rating scale. Participants will also be asked to complete ratings following eating episodes (Event-contingent ratings). They will indicate the extent of loss of control experienced and the strength of negative mood. These data will allow the investigators to evaluate daily patterns in the timing of eating and rates of dysregulated eating episodes. All ratings provided (i.e., regardless of the type of EMA sampling schedule) are time stamped, so that reports cannot be retrospectively completed or backdated. Data are saved locally to the device and are downloaded when participants return the device to the lab. Randomization The covariate adaptive randomization method will be used to accomplish randomization of participants to conditions while assuring that participants are equally allocated to each of the two conditions (i.e., normal light first or bright light first). Normal Light (control) During the normal light phase, participants will be instructed to sit with their faces approximately 22 inches away from the therapy light for 30 minutes each day, starting 30 minutes after waking. The lights will be preset to deliver ~500lux at this distance. This phase will last 10 days. On the 10th day, participants will return the light to the lab (and if this was their first light phase, pick up a new light) and complete self-report measures. These measures include the the Change in Eating Disorders Symptoms scale, the Center for Epidemiological Studies Depression scale - Revised, and subjective appetite scales. Bright Light Manipulation (active) During the bright light phase, participants will be instructed to sit with their faces approximately 22 inches away from the therapy light for 30 minutes each day, starting 30 minutes after waking. The lights will be preset to deliver ~10,000lux at this distance. This phase will last 10 days. On the 10th day, participants will return the light to the lab (and if this was their first light phase, pick up a new light) and complete self-report measures. These measures include the the Change in Eating Disorders Symptoms scale, the Center for Epidemiological Studies Depression scale - Revised, and subjective appetite scales. Actiwatch Spectrum PRO® The Actiwatch Spectrum PRO® by Philips Respironics is a wrist-worn device that quantifies movement using 3-axis accelerometry. The device uses a cantilevered piezoelectric beam to measure the direction and intensity of movement. In combination with the accompanying software, validated algorithms compute sleep and wake times, sleep efficiency, and nighttime awakenings and provide information on daytime activity. Movement data are sampled in 1-minute epochs, which the device can store for up to 30 days of continuous monitoring without charging or needing to upload data. This epoch length allows for a largely uninterrupted view of movement, and therefore precise timing of the sleep period. In addition to the accelerometer, this device contains an ambient light monitor, which uses a photodiode element with a spectral frequency responsivity that is similar to the human eye (i.e., 330 to 720 nanometers), characterizing the brightness of light falling on the device from 0.1 to 150,000 lux (+/-2%). The data collected with the light monitor can be a useful adjunct to the sleep algorithm, as changes to ambient light are often driven by behavioral constraints on the sleep episode. Critically, for this study, the measurement of light will allow for quantifying participants' exposure to naturally occurring light during the period of manipulation of light in addition to confirming adherence to the light manipulation itself. Finally, this device is waterproof at 1 meter for 30 minutes, meaning participants can do all normal daily tasks while wearing the device, except swimming/diving. Hormone Assessment Saliva samples for cortisol will be collected on six days: the first and second days (prior to any light manipulation), the 11th and 12th days (at the end of the first light manipulation) and the 21st and 22nd days (at the end of the second light manipulation). Salivary Cortisol: Cortisol awakening response levels will be collected at home via passive drool saliva samples (0.5 ml) at three times: immediately upon waking, 15 minutes after waking, and 30 minutes after waking. Cortisol saliva-serum correlations are high (r = .91), and saliva is advantageous over serum because collection is less invasive and levels reflect the biologically active fraction of the hormone. Standard collection (e.g., no brushing their teeth, eating, chewing gum for 1 hour prior) and storage (i.e., immediately freeze at -20 °C) procedures will be followed. Samples will be analyzed using commercially available high sensitivity enzyme immunoassays kits from Salimetrics, LLC (Carlsbad, CA, USA). Salimetrics cortisol assay characteristics are excellent in terms of minimum detection, intra-assay and inter-assay coefficients, spike recovery, and linearity, and samples will be run in triplicate to ensure reliable results. Samples will be tested according to the testing protocol provided by Salimetrics, LLC.

Participants will use a therapy light that provides ~500 lux at 22" for 30 minutes each morning beginning 30 minutes after waking.

Participants will receive Bright Light Exposure using a therapy light that provides ~10,000 lux at 22" for 30 minutes each morning beginning 30 minutes after waking.

For 10 consecutive days, participants will use an Ultralux(R) V HD-LED light therapy light for 30 minutes beginning 30 minutes after waking. The light will be preset to deliver ~10,000 lux at 22".

For 10 consecutive days, participants will use an Ultralux(R) V HD-LED light therapy light for 30 minutes beginning 30 minutes after waking. The light will be preset to deliver ~500 lux at 22".

The cortisol awakening response will estimated from saliva samples collected at home via passive drool (0.5 ml) at three times: immediately upon waking, 15 minutes after waking, and 30 minutes after waking. The curve of this response can be estimated from these three measurements. The response will be averaged over two consecutive days of collection and then compared between the two intervention phases to quantify the change in the awakening response accounted for by the interventions while controlling for the response at baseline before any interventions have been employed.

Participants will report using ecological momentary assessment (EMA) when they eat the degree to which they experienced their eating as out of control. These reports are provided after each eating episode that occurs during the study period (i.e., days 1 -22 of the protocol). The ratings will be averaged within protocol phases (i.e., average loss of control for eating episodes occurring during first intervention phase and the average during the second intervention phase) and compared to quantify the change in binge eating accounted for by the difference between the two interventions.

Participants will be signaled using ecological momentary assessment (EMA) at six times per day distributed across their waking hours and will be asked to provide momentary negative mood ratings. Participants will also be asked to initiate reports after they have eaten, and they will provide ratings for mood then as well. These ratings will be made on a scale ranging from 1 to 9 using a wrist-worn electronic device. The ratings will be averaged within protocol phases (i.e., average negative mood occurring during first intervention phase and the average during the second intervention phase) and compared to quantify the change in negative mood accounted for by the difference between the two interventions.

Participants will complete a self-report measure of recent change in eating disorder symptoms (the Change in Eating Disorder Symptoms Scale; Spangler, 2010) at laboratory visits as secondary, aggregate measures of clinical change. This measure contains seven subscales (body preoccupation, body dissatisfaction, body checking, binge eating, restrictive eating, food preoccupation, and vomiting) in addition to the total score. Thirty-five items are rated on a 0 to 4 scale, with higher scores indicating more severe eating disorder symptoms. The change in these symptom scores between baseline (protocol day 0) and each light phase (protocol days 12 and 22, respectively) will serve as an index of the change in the severity of global eating disorder symptoms accounted for by each intervention.

Participants will complete a self-report measure of depressive symptoms (the Center for Epidemiological Studies Depression scale - Revised; Eaton et al., 2004) at laboratory visits as secondary, aggregate measure of clinical change in depression symptoms. Twenty items are rated on a 0 to 4 scale, with higher scores indicating more severe depressive symptoms. The change in this score between baseline (protocol day 0) and following each light phase (protocol days 12 and 22, respectively) will serve as an index of the change in the severity of clinical depressive symptoms accounted for by each intervention.

Inclusion Criteria: Female sex Age of at least 18 years Minimally normal body weight (BMI ≥ 18.5 kg/m2) Clinically significant binge eating (i.e., ≥ 2 binge-eating episodes/week over the previous 3 months Clinically significant eating disorder-related impairment (indicated by a score of > 16 on the Clinical Impairment Assessment) Exclusion Criteria: Receiving psychotherapy or psychotropic medication for the 6 weeks prior to study entry. Presence of medical conditions or medications that result in changes in hormone function, weight or appetite. Lifetime presence of bipolar disorder.