SEARCH

SEARCH BY CITATION

Keywords:

  • weight preoccupation;
  • menstrual cycle;
  • ovarian hormones;
  • estradiol;
  • progesterone;
  • binge eating;
  • emotional eating;
  • eating disorders;
  • bulimia nervosa

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Objective

Previous research has shown that fluctuations in ovarian hormones (i.e., estradiol and progesterone) predict the changes in binge eating and emotional eating across the menstrual cycle. However, the extent to which other eating disorder symptoms fluctuate across the menstrual cycle and are influenced by ovarian hormones remains largely unknown. This study sought to examine whether the levels of weight preoccupation vary across the menstrual cycle and whether the changes in ovarian hormones and/or other factors (i.e., emotional eating and negative affect) account for menstrual cycle fluctuations in this eating disorder phenotype.

Method

For 45 consecutive days, 352 women (age, 15–25 years) provided daily ratings of weight preoccupation, negative affect, and emotional eating. Saliva samples were also collected on a daily basis and assayed for levels of estradiol and progesterone using enzyme immunoassay techniques.

Results

Weight preoccupation varied significantly across the menstrual cycle, with the highest levels in the premenstrual and menstrual phases. However, ovarian hormones did not account for within-person changes in weight preoccupation across the menstrual cycle. Instead, the most significant predictor of menstrual cycle changes in weight preoccupation was the change in emotional eating.

Discussion

Fluctuations in weight preoccupation across the menstrual cycle appear to be influenced primarily by emotional eating rather than ovarian hormones. Future research should continue to examine the relationships among ovarian hormones, weight preoccupation, emotional eating, and other core eating disorder symptoms (e.g., body dissatisfaction, compensatory behaviors) in an effort to more fully understand the role of these biological and behavioral factors for the full spectrum of eating pathology. © 2014 Wiley Periodicals, Inc. (Int J Eat Disord 2014)


Introduction

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Ovarian hormones (i.e., estradiol and progesterone) have recently been implicated in the etiology of eating disorders in women.[1-4] Most of the research, to date, has focused on the role of ovarian hormones in risk for binge eating. This is not surprising, given the findings from animal studies, demonstrating that ovarian hormones have direct, causal effects on food intake.[5, 6] Specifically, the removal of the source of ovarian hormones through bilateral ovariectomy in rats causes increased food intake, and adminstration of estradiol reverses this effect. In contrast, progesterone causes increased food intake, in part, by antagonizing the inhibitory effects of estradiol.[6-8]

In humans, food intake, binge eating, and emotional eating (i.e., the tendency to eat when experiencing negative emotions[9]) have been found to be significantly higher during the midluteal and premenstrual phases of the menstrual cycle as compared to the follicular/ovulatory phases.[1, 3, 10-13] The studies that have directly examined the levels of estradiol and progesterone confirm that within-person changes in ovarian hormones account for these menstrual cycle fluctuations.[1, 3, 13] Specifically, although initial pilot data suggested that lower estradiol and higher progesterone levels were associated with increases in binge eating and emotional eating,[1, 3] a recent study indicated that the interactions between estradiol and progesterone (i.e., high levels of both) contribute to midluteal increases in emotional eating as well.[14] In addition, ovarian hormone/dysregulated eating associations have been shown to be stronger in women with clinically significant levels of binge eating compared to women without binge episodes.[13] Importantly, in all previous studies, hormone effects on binge eating and emotional eating were independent of important covariates that also change across the menstrual cycle, including negative affect and body mass index (BMI).[1, 3, 13, 14]

Far fewer studies, however, have examined menstrual cycle changes in other disordered eating variables. Broadening the phenotypes examined is important for developing more complete models of the role of ovarian hormones in the full spectrum of eating pathology. One particularly important set of variables to investigate are those related to weight concerns/preoccupation (i.e., intense preoccupation with weight, dieting, and the pursuit of thinness[15]). Weight concerns have been identified as one of the most robust prospective risk factors for the development of clinically significant eating disorders,[16] and weight concerns are directly related to core symptoms (e.g., undue influence of body weight/shape on self-evaluation) of anorexia nervosa and bulimia nervosa.[17]

In the only previous report of its kind, Racine et al.[4] examined the association between menstrual cycle fluctuations in ovarian hormones and changes in weight concerns in two independent samples of women. In the first sample, robust fluctuations in weight preoccupation were observed across the menstrual cycle, where the levels of weight preoccupation were highest in the midluteal phase. Menstrual cycle changes in weight preoccupation were primarily accounted for by within-person increases in progesterone and, to a lesser extent, decreases in estradiol. However, in a second sample, more modest, nonsignificant changes in weight preoccupation were observed across the menstrual cycle, and weight preoccupation was highest during the premenstrual phase. Reasons for the discrepant results across samples are unclear although the very small sample sizes of these studies (N = 8 and 10, respectively) may have contributed to instability in effects.

Clearly, additional research using larger samples of women is needed to clarify the presence/absence of within-person menstrual-cycle fluctuations in weight preoccupation. In addition, it will be important for these studies to determine whether menstrual-cycle changes in weight preoccupation are due to changes in ovarian hormones, changes in psychological factors (e.g., increased negative affect), and/or the changes in emotional eating that have been shown to fluctuate across the menstrual cycle in past research.[1, 3, 13, 14],[18, 19] For example, emotional eating has been previously linked to ovarian hormones,[1, 3, 13, 14] and we might expect weight concerns to increase during certain menstrual cycle phases as a result of emotional eating. Specifically, increased emotional eating during the midluteal phase could cause women to be more concerned and/or conscious about their body shape/weight; in this case, weight concerns may be owing to eating in the presence of negative emotions rather than to changes in ovarian hormones. Thus, it is important to determine the factors that might account for menstrual cycle changes in weight preoccupation: emotional eating, psychological factors (e.g., negative affect), and/or ovarian hormones.

Given the above, the aim of the current study was to investigate within-person changes in weight preoccupation across the menstrual cycle utilizing a large, community-based sample of women. First, we were interested in examining whether the levels of weight preoccupation significantly vary across the menstrual cycle. Second, we wanted to investigate whether within-person fluctuations in weight preoccupation across the menstrual cycle are best accounted for by within-person changes in ovarian hormones, negative affect, emotional eating, or a combination of these factors.

Method

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Participants

Participants included 352 same-sex female twins (194 monozygotic twins; 158 dizygotic twins) between the ages of 15 and 25 years drawn from the Twin Study of Hormones and Behavior Across the Menstrual Cycle[14] within the Michigan State University Twin Registry (MSUTR[20, 21]). All participants completed written informed consent before enrolling in the study. Importantly, a subset of these participants (N = 174; 49% of the current sample) was examined in the study by Klump et al.[14] in which significant hormone effects for emotional eating were detected. Our analyses extend these initial results by investigating whether similar hormone effects are present for weight preoccupation scores, after controlling for emotional eating and other important covariates.

Twins from the MSUTR are recruited using birth record methods described previously.[20, 21] Twins included in this study were demographically representative of the recruitment region (81.7% Caucasian, 16.2% African American, 1.0% Asian/Pacific Islander, and 1.0% Native American; http://www.michigan.gov/mdch).

To ensure that we captured natural hormonal variations across the menstrual cycle, we developed a variety of participant inclusion/exclusion criteria: (1) regular menstrual cycles (i.e., every 22–32 days) for last 6 months; (2) no hormonal contraceptive use for last 3 months; (3) no psychotropic or steroid medication use for last 4 weeks; (4) no pregnancy or lactation in last 6 months; and (5) no history of genetic or medical conditions that may influence hormones or appetite/weight. Despite these exclusion criteria, participants from this study and those from the previous MSUTR studies without such criteria did not meaningfully differ on the levels of disordered eating (average Cohen's d = .12, range = 0.01–0.20).

Procedures

All study procedures, methods, and materials were reviewed and approved by the Michigan State University Institutional Review Board. Behavioral and hormone data were provided on a daily basis across the 45 days of the study. Salivary samples were collected within the first 30 min of waking using the previously established methods.[3] Questionnaires were completed each evening (after 5:00 PM) using an online data system or preprinted scantrons. This pattern of morning saliva samples and evening behavioral data collection was to ensure that hormone measurements preceded behavioral ratings each day.

In addition to daily data collection, all participants completed three in-person visits occurring at the start of the study, halfway through the study (∼day 23), and at the end of data collection (∼day 45). During these in-person assessments, eligibility was reassessed, height and weight were measured, and completed materials were collected from participants. Between visits, staff contacted the participants 1×/week to answer questions and confirm continued protocol adherence. These procedures were effective at identifying the individuals who were no longer eligible to participate due to missed periods, medication use, and/or pregnancy during the study (<3%) as well as minimizing drop-outs. The percentage of participants who were ultimately dropped from the study or whose data were not analyzed owing to the failure to collect a sufficient number of samples was minimal (<3 and <6%, respectively).

Measures

The sample characteristics for all study measures are summarized in Table 1. Although daily, longitudinal data were examined for most variables (i.e., weight preoccupation, emotional eating, negative affect, and ovarian hormones), we present averages across data collection in Table 1 to characterize the sample for comparisons with future studies.

Table 1. Sample characteristics
VariableMean (SD)Range
  1. Note: The values for weight preoccupation, estradiol, progesterone, negative affect, and emotional eating are average values across the 45-day data collection period. The values for BMI are average values across the three study visits.

Age18.10 (1.76)16–22
Weight preoccupation2.19 (1.98)0.02–7.98
Estradiol (pg/mL)2.87 (1.39)0.70–12.43
Progesterone (pg/mL)124.76 (67.08)18.73–397.04
Negative affect15.21 (3.78)10–29
Emotional eating0.34 (.41)0–3
BMI24.07 (5.67)15.81–47.59
Weight Preoccupation

Weight preoccupation was assessed on a daily basis for 45 days using the Weight Preoccupation Scale of the Minnesota Eating Behaviors Survey (MEBS).[22] 1 The MEBS Weight Preoccupation Scale consists of eight true/false questions that ask about a variety of cognitions and behaviors related to weight concerns (e.g., “I am really afraid of gaining weight” and “I often weigh myself to see if I am gaining weight”). Internal consistency for the MEBS Weight Preoccupation Scale has ranged from acceptable to excellent in the previous studies (α = .71–.85)[22] and in this study (α = .82). In addition, in support of its criterion validity, women with eating disorders score higher on this scale compared to women without eating disorders.[22]

Emotional Eating

Emotional eating was assessed on a daily basis for 45 days using the Emotional Eating Scale of the Dutch Eating Behavior Questionnaire (DEBQ).[9] The Emotional Eating Scale assesses eating in response to negative emotions (e.g., “Did you have desire to eat when you were discouraged?”) on a five-point scale ranging from not at all to very often. Internal consistencies for the DEBQ Emotional Eating Scale are excellent in the previous research (α = .93)[3, 4, 9] and in this study (average, α = .90). Importantly, eating in response to negative emotions is believed to be a core feature of binge eating, and the Emotional Eating Scale of the DEBQ has demonstrated validity in differentiating between individuals with bulimia nervosa and/or binge eating, overweight individuals, and college students. Furthermore, the DEBQ Emotional Eating Scale is correlated with the established measures of binge eating (r = .55–.69)[23, 24] as well as with palatable food intake (i.e., ice cream) in a laboratory setting.[24]

Negative Affect

The Positive and Negative Affect Schedule (PANAS) Negative Affect Scale[25] was used to assess negative affect on a daily basis for 45 days. This scale consists of 10 items that assess the full range of daily negative emotions (e.g., distress, nervousness, irritability, and fear). The degree to which each emotion was experienced was rated on a five-point scale ranging from very slightly/not at all to extremely. The PANAS Negative Affect Scale has exhibited excellent internal consistency as well as good convergent and discriminant validity.[25] Internal consistency in this study was excellent (average, α = .85).

Ovarian Hormones

Estradiol and progesterone were assayed from daily saliva samples. Saliva samples are preferred over other methods (e.g., blood spots) because they represent a less invasive collection method, particularly when repeated samples are needed. Previous research has found that saliva samples are associated with higher compliance and more robust hormone–behavior associations than blood spot sampling.[1]

Saliva samples were processed by Salimetrics, LLC (State College, PA) using enzyme immunoassay kits designed specifically for analyzing saliva. These assays show excellent intra- and interassay coefficients of variation (estradiol = 7.1 and 7.5%; progesterone = 6.2 and 7.6%), as well as assay sensitivity (measured by interpolating the mean optical density minus 2 SDs of 10–20 replicates at the 0 pg/mL level; estradiol = 0.10 pg/mL; progesterone = 5 pg/mL) and method accuracy (determined by spike recovery and linearity, estradiol = 104.2 and 99.4%; progesterone = 99.6 and 91.8%). To conserve resources, the samples were only assayed every other day during menstrual bleeding and early follicular phase when hormones are expected to be low and stable. This process ensured that we captured the periods of maximum hormonal change across the menstrual cycle (e.g., midlate follicular though premenstrual phase), whereas in turn maximizing the number of participant samples assayed.

Body Mass Index

BMI was included as a covariate in the analyses, given the previous research showing its association with weight preoccupation,[26] as well as with ovarian hormone levels.[27, 28] To determine BMI, participants' height and weight were measured during the three in-person study visits using a wall-mounted ruler and digital scale, respectively. BMI was calculated using the following formula: (BMI = weight [kg]/height [m][2]).

Statistical Analyses

Data Preparation

Data preparation followed that used in the previous studies examining the relationship between ovarian hormones and binge eating/emotional eating across the menstrual cycle.[1, 3, 13, 14] For our repeated measures (i.e., hormones, weight preoccupation, negative affect, emotional eating, and BMI), 5-day rolling averages were calculated and standardized within-person. Five-day rolling averages were calculated by averaging the scores of the variable (e.g., weight preoccupation) across a chosen day, 2 days prior, and 2 days after the chosen day. For example, the levels of weight preoccupation on day 8 were calculated as the average from day 6 to day 10.[11] Previous research has used rolling averages, and they are preferred because of their ability to minimize the random variation that is present in behavioral data owing to the environmental circumstances.[10] To accommodate the fact that BMI was assessed at only three time points across the study, rolling averages were calculated using visit 1 BMI for days in-between the first and second in-person assessments, visit 2 BMI for days in-between the second and third in-person assessment, and the visit 3 BMI for the last day of the study. These rolling averages were then converted to within-person standardized scores that were based on each individual participant's overall standard deviation across the study. This standardization allowed for the examination of the degree to which the changes in a woman's ovarian hormones, relative to her equilibrium, predict the changes away from the woman's equilibrium in weight preoccupation.

Statistical Models

Similar to the previous analyses of this data set,[14, 29] mixed linear models (MLMs) were used to examine the possible influence of menstrual cycle phase, ovarian hormones, negative affect, and emotional eating on within-person changes in weight preoccupation. These models were well suited for testing our hypotheses as they could examine the effects of predictors while controlling for the non-independence of the repeated measures and twin data. Specifically, we allowed residual errors for weight preoccupation to correlate between the members of a twin pair, and we estimated a time-specific dyadic correlation that allowed twin's residual errors to correlate from day to day. Each time-varying predictor was included as a random effect to model random slopes and the relationship between these slopes within twin pairs. However, there was no evidence that random slopes were correlated across twins, and hence an identity covariance matrix, which estimates a single variance for both twins in a pair, was used.

We first examined whether weight preoccupation significantly varied across the menstrual cycle to then investigate whether hormones and/or other factors (i.e., negative affect and emotional eating) accounted for any observed fluctuations. Similar to the previous studies, menstrual cycle phase was coded based on the dates of menstrual bleeding and increases/decreases in ovarian hormone levels for each participant.[1, 3, 14] Specifically, each woman's menstrual cycle(s) was categorized into five primary cycle phases (i.e., follicular, ovulatory, midluteal, premenstrual, and menstrual) and three transition phases (i.e., follicular to ovulatory, ovulatory to mid-luteal, and mid-luteal to premenstrual). In the analyses examining weight preoccupation changes across menstrual cycle phase, the follicular phase included the days immediately after the end of menstrual bleeding, during which time progesterone is low and estradiol slowly begins to increase; the ovulatory phase included the days during which estradiol has its peak, as well as the transition phases preceding and following ovulation, during which time estradiol is rising and falling, respectively; the mid-luteal phase included the midluteal phase and the mid-luteal to premenstrual transition phase, a time characterized by the highest levels of progesterone and a secondary peak in estradiol; the perimenstrual phase included both the premenstrual and the menstrual phases when both hormone levels are low. Using MLM, fluctuations in weight preoccupation across the menstrual cycle were tested by including phase as a predictor of weight preoccupation.

Next, we investigated whether significant fluctuations in weight preoccupation across the menstrual cycle were accounted for by within-person changes in ovarian hormones, negative affect, emotional eating, or a combination of these factors. We first fit a model (“Model 1”) that tested the main effects of estradiol, progesterone, as well as the estradiol × progesterone interaction while controlling for any effects of BMI. In Models 2 and 3, negative affect and emotional eating, respectively, were individually added as predictors to examine whether they were significantly associated with within-person changes in weight preoccupation across the menstrual cycle and whether they accounted for any effects of ovarian hormones on weight preoccupation. Finally, Model 4 included both negative affect and emotional eating as the predictors of weight preoccupation to examine which of these psychological factors might be most strongly associated with the changes in weight preoccupation.

Results

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The results from the MLM examining menstrual phase as a predictor of within-person changes in weight preoccupation indicated that the levels of weight preoccupation varied significantly across the menstrual cycle (F(3,12,674) = 3.36, p = .02). The highest levels of weight preoccupation were observed during the peri-menstrual phase (i.e., premenstrual and menstrual phases; M(SE) = .05 (.02)). Importantly, peri-menstrual weight preoccupation scores were significantly higher than the scores in all other menstrual cycle phases (follicular, (M(SE) = .003 (.02); ovulatory, M(SE) = −.02 (.02); mid-luteal, M(SE) = .005 (.01); t's = 2.02–3.22; p's =.002-.04). There were no significant differences among the follicular, ovulatory, and mid-luteal menstrual phases (t = 0.08–1.28; p = .20–.94), but weight preoccupation scores appeared to be lowest in the ovulatory phase (Fig. 1).

image

Figure 1. Changes in weight preoccupation, estradiol, and progesterone across the menstrual cycle. T = transition days that are in-between phases. Mean z-score = the mean of the 5-day rolling averages calculated within participants, then averaged across participants. Mean values within each phase are included for descriptive purposes only as the daily z-scores were included in the hierarchical linear models for each phase contrast. The number of days included in each phase varied by participant based on their cycle length, but the days roughly corresponded to the following (first day of menstrual bleeding = +1; previous day = −1): follicular = +6 to +11; transition from follicular to ovulatory = +12 to +13; ovulatory = −15 to −12; transition from ovulatory to mid-luteal = −11 to −10; mid-luteal = −9 to −5; transition from mid-luteal to premenstrual = −4; premenstrual = −3 to +1; and menstrual = +2 to +5.

Download figure to PowerPoint

The second set of MLMs considered whether the changes in ovarian hormones, negative affect, and/or emotional eating account for observed menstrual cycle fluctuations in weight preoccupation (Table 2). Across all models, estradiol, progesterone, and the estradiol–progesterone interaction did not significantly predict within-person changes in weight preoccupation. In contrast, negative affect and emotional eating were both significant predictors of weight preoccupation when entered individually in the MLMs, suggesting that these psychological/behavioral factors may have a stronger effect on menstrual cycle changes in weight preoccupation than ovarian hormones. Interestingly, when negative affect and emotional eating were included together in Model 4, only emotional eating (and not negative affect) emerged as a significant predictor of weight preoccupation. Thus, of the factors examined in this study, the strongest predictor of within-person changes in weight preoccupation across the menstrual cycle was emotional eating. 2

Table 2. Ovarian hormones, negative affect, and emotional eating predicting within-person changes in weight preoccupation
Modelb (SE)tdfp-Value
  1. Abbreviations: BMI, body mass index; NA, negative affect; and EE, emotional eating.

Model 1: Hormones and BMI only    
Estradiol−.006 (.02)−0.30315.76
Progesterone.007 (.02)0.30326.77
Estradiol × progesterone.002 (.01)0.12314.90
BMI−.02 (.03)−0.74302.46
Model 2: Controlling for NA    
Estradiol−.01 (.02)−0.52309.60
Progesterone.01 (.02)0.42327.67
Estrogen × Progesterone−.001 (.01)−0.01317.99
BMI−.03 (.03)−0.95302.34
Negative affect.05 (.02)2.42313.02
Model 3: Controlling for EE    
Estradiol−.01 (.02)−0.62309.53
Progesterone.01 (.02)0.63313.53
Estrogen × Progesterone−.005 (.01)−0.38306.70
BMI−.02 (.03)−0.84301.40
Emotional Eating.12 (.02)5.07314<.001
Model 4: Controlling for NA and EE    
Estradiol−.01 (.02)−0.70305.48
Progesterone.01 (.02)0.67315.50
Estrogen × Progesterone−.004 (.01)−0.29312.77
BMI−.02 (.03)−0.93301.35
Negative Affect.02 (.02)1.23307.22
Emotional Eating.11 (.03)4.67307<.001

Given this pattern of results, we were interested in examining whether emotional eating episodes earlier in the menstrual cycle may drive later increases in weight preoccupation. Figure 2 shows mean emotional eating and weight preoccupation scores across the menstrual cycle, and the examination of menstrual cycle fluctuations in these symptoms suggests that the changes in emotional eating are followed by changes in weight preoccupation. To further investigate this hypothesis, we conducted time-lagged prospective analyses that investigated whether emotional eating scores on 1 day predict weight preoccupation scores during later days in the menstrual cycle. We considered both proximal (i.e., 1-, 2-, and 3-day time-lagged associations) and more distal (i.e., 5- and 10-day time-lagged associations) relationships between emotional eating and weight preoccupation. To account for within-person stability over time, same-day weight preoccupation was entered as a covariate in analyses. The results indicated that emotional eating significantly predicted within-person changes in weight preoccupation 1, 2, and 3 days later, and associations in 5- and 10-day time-lagged analyses were marginally significant (Table 3). Importantly, the analyses examining the reverse relationship (i.e., whether weight preoccupation predicted emotional eating later in the cycle) supported the hypothesis that the direction of effects is from emotional eating to weight preoccupation as weight preoccupation did not significantly predict lagged emotional eating scores, with the exception of 1-day time-lagged analyses (Table 3).

image

Figure 2. Changes in weight preoccupation and emotional eating across the menstrual cycle. T = transition days that are in-between phases. Mean z-score = the mean of the 5-day rolling averages calculated within participants, then averaged across participants. Mean values within each phase are included for descriptive purposes only as the daily z-scores were included in the hierarchical linear models for each phase contrast. The number of days included in each phase varied by participant based on their cycle length, but the days roughly corresponded to the following (first day of menstrual bleeding = +1; previous day = −1): follicular = +6 to +11; transition from follicular to ovulatory = +12 to +13; ovulatory = −15 to −12; transition from ovulatory to mid-luteal = −11 to −10; mid-luteal = −9 to −5; transition from mid-luteal to premenstrual = −4; premenstrual = −3 to +1; and menstrual = +2 to +5.

Download figure to PowerPoint

Table 3. Time-lagged analyses examining the direction of the association between emotional eating and weight preoccupation
 Emotional Eating to Weight PreoccupationWeight Preoccupation to Emotional Eating
Time-Lagb (SE)t (df)p-Valueb (SE)t (df)p-Value
1 Day.02 (.005)2.94 (340).003.03 (.01)2.27 (338).02
2 Days.02 (.009)2.82 (322).005.02 (.02)1.35 (346).18
3 Days.03 (.01)2.13 (332).03−.02 (.02)−1.42 (333).16
5 Days.03 (.02)1.81 (334).07−.02 (.02)−1.09 (331).27
10 Days.03 (.02)1.77 (322).08−.01 (.02)−0.51 (308).61

Discussion

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The current study used a large, community-based sample of women to examine whether weight preoccupation, a core eating disorder symptom, significantly varies across the menstrual cycle and whether ovarian hormones, psychological factors (i.e., increased negative affect), and/or emotional eating might account for these within-person changes. The importance of this study is highlighted by the fact that most of the previous studies examining the influence of the menstrual cycle and ovarian hormones on disordered eating symptoms have focused solely on binge eating and emotional eating. The only previous report to investigate menstrual-cycle effects for weight preoccupation examined two very small samples (N = 8–10) and generated discrepant results.[4]

Results from the current study suggest that levels of weight preoccupation vary significantly across the menstrual cycle. Specifically, weight preoccupation scores were highest in the peri-menstrual (i.e., premenstrual and menstrual) phase of the menstrual cycle, and weight preoccupation scores in this phase were significantly higher than weight preoccupation scores in all other menstrual cycle phases (i.e., follicular, ovulatory, and mid-luteal). Our findings are in line with almost all the previous studies examining disordered eating symptoms (i.e., binge eating, emotional eating, and body dissatisfaction), suggesting that the risk for disordered eating is higher in the post-ovulatory versus the pre-ovulatory half of the menstrual cycle.[1, 3, 10-12] Most importantly, in the one previous report to investigate menstrual cycle changes in weight preoccupation,[4] the findings pointed to both mid-luteal and premenstrual increases in weight preoccupation although only the mid-luteal peaks were significant. These results are consistent with the significant peaks in weight preoccupation during the premenstrual and menstrual phases and the second highest levels of weight preoccupation during the mid-luteal phase observed in this study. Given that our large sample provided ample power to detect fluctuations in weight preoccupation across the menstrual cycle, the current findings significantly contribute to our understanding of peak periods of risk for weight concerns across the menstrual cycle.

In considering whether the changes in ovarian hormones, emotional eating, or negative affect accounted for menstrual-cycle fluctuations in weight preoccupation, we found that within-person changes in negative affect and emotional eating significantly predicted the changes in weight preoccupation. However, when both emotional eating and negative affect were included in the model as predictors of weight preoccupation, only the changes in emotional eating were significantly associated with the changes in weight preoccupation. Interestingly, the analyses examining the longitudinal relationship between weight preoccupation and emotional eating confirmed that emotional eating significantly predicts later increases in weight preoccupation and that the direction of the effect is from emotional eating to weight preoccupation and not the reverse (i.e., from weight preoccupation to emotional eating). In contrast, there were no significant effects of estradiol, progesterone, or the estradiol–progesterone interaction on weight preoccupation changes. Therefore, different from the previous research that has demonstrated direct effects of ovarian hormones on other disordered eating symptoms (e.g., binge eating and emotional eating),[1, 3] weight preoccupation appears to be more strongly related to emotional eating than to ovarian hormone changes. Taken together, within the broader construct of disordered eating, specific symptoms may be differentially related to ovarian hormones. In particular, although ovarian hormones significantly interact to predict menstrual cycle changes in emotional eating, ovarian hormones were not associated with weight preoccupation in our large, community sample.

Although it remains unclear why emotional eating is the strongest predictor of weight preoccupation, a few tentative hypotheses can be put forth. The mid-luteal phase is associated with the highest levels of emotional eating, and the previous research has shown that elevated emotional eating is due to higher levels of progesterone and estradiol during this phase.[14] Increased emotional eating due to the changes in ovarian hormones may lead to subsequent increases in concerns about the effect of these eating behaviors on body weight and shape. Indeed, in our data, the highest peaks in weight preoccupation scores occurred during the perimenstrual phase after the high levels of emotional eating during the mid-luteal phase (Fig. 2), and additional analyses confirmed that emotional eating significantly predicts later increases in weight preoccupation. These findings map on well to the previous research using daily diary and ecological momentary assessment designs, demonstrating that binge episodes are often followed by an increase in guilt and dietary restraint, constructs that are significantly related to weight preoccupation.[30-32] Taken together, our findings suggest that mid-luteal increases in emotional eating (as a result of ovarian hormone effects) may trigger weight concerns and motivate women to more closely monitor their body weight/shape during the subsequent peri-menstrual phase.

Alternatively, an additional set of factors that might explain the peaks in weight preoccupation during the peri-menstrual phase are the physical changes that occur prior to and during menstruation (e.g., water retention) as these physical changes could lead to greater concern and preoccupation with body weight. Indeed, research has shown that body dissatisfaction and appearance-related anxiety are associated with water retention and other symptoms of menstrual distress during the peri-menstrual phase.[33] Although we controlled for changes in BMI across the cycle, we did not assess peri-menstrual physical symptoms and could therefore not investigate whether increases in weight preoccupation during the peri-menstrual phase were associated with these physical symptoms. Furthermore, body weight/BMI was not measured during the peri-menstrual phase for all participants as the three study visits during which height and weight were measured did not correspond to specific menstrual cycle phases. Although changes in body weight across the study period were minimal in our sample (M = 0.40 lb change; SD = 3.80), daily or weekly measurements of body weight may provide additional information with regard to mechanisms for increased weight preoccupation during the peri-menstrual phase. Therefore, future research should combine more frequent weight measurements with an examination of the effects of water retention and other physical changes on fluctuations in weight preoccupation across the menstrual cycle to help identify additional factors that may account for these changes.

Finally, it is possible that other biological factors that fluctuate across the menstrual cycle may contribute to the changes in weight preoccupation. For example, leptin levels significantly fluctuate across the menstrual cycle,[34, 35] and longitudinal data have shown that leptin levels are highest during the luteal phase, which is also associated with increases in BMI.[34] Furthermore, other studies have suggested a positive correlation between circulating levels of leptin and BMI in women with eating disorders.[36] Taken together, it is possible that luteal phase increases in leptin, which correspond to increases in BMI, may result in greater weight preoccupation in the subsequent peri-menstrual phase. Although the results from this study did not suggest that ovarian hormones were a significant predictor of weight preoccupation, other biological factors, like leptin, may be important biological candidates to investigate with regard to weight preoccupation changes across the menstrual cycle.

Notably, it is also important for future studies to examine other key symptoms of eating disorders (e.g., body dissatisfaction, dietary restriction, and compensatory behaviors) that may fluctuate across the menstrual cycle and that may (or may not) demonstrate associations with ovarian hormones. Information about these symptoms could contribute to a more comprehensive understanding of the role of ovarian hormones in the full spectrum of eating pathology and may point to further differential associations between ovarian hormones and dimensions of disordered eating. In addition, we might expect peri-menstrual increases in weight preoccupation to confer risk for the subsequent occurrence of eating disorder behaviors aimed at preventing weight gain (e.g., dietary restriction, compensatory behaviors). While data from the current study are not sufficient to investigate these behaviors across the menstrual cycle (i.e., dietary restraint was measured only three times across the 45-day study; the frequency of compensatory behaviors in our community sample is too low), future studies examining the relationships between changes in weight preoccupation and changes in these other eating disorder symptoms across the menstrual cycle could help address this possibility. Findings may have important clinical significance for identifying peak periods of risk for specific eating disorder symptoms across a woman's menstrual cycle.

Primary strengths of this study include the analysis of longitudinal behavioral and hormone data collected across 45 consecutive days in a large sample of women. Despite these strengths, there are limitations of our study that should be acknowledged. First, given the age range of our sample (15–25 years), many participants were not through the peak period of risk for eating disorders, which extends up until at least 25 years.[37] It is important to note that eating disorder attitudinal symptoms, such as weight preoccupation, have been studied in children[38, 39] and have been implicated as risk factors for the later development of full eating disorders.[16, 40] Therefore, the age range of our sample is likely appropriate for studying weight preoccupation, a core cognitive correlate of eating disorders. Second, it is unknown if findings for weight preoccupation in our community sample generalize to clinical samples of eating disorder patients, and whether the relationship between objective binge episodes and weight preoccupation would map on to what we observed for emotional eating and weight preoccupation. Although these will be important questions for future research, the examination of a community sample of women allowed us to investigate the full spectrum of disordered eating severity and to further explore possible etiologic factors contributing to the development of weight preoccupation.

Acknowledgments

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The content is solely the responsibility of the authors and does not necessarily represent the official views of Michigan State University or the National Institute of Mental Health. Parts of this manuscript were presented at the Eating Disorders Research Society meeting, Bethesda, Maryland, September 19–21, 2013. None of the authors have biomedical financial conflicts of interest or other potential conflicts of interest to disclose.

  • 1

    The Minnesota Eating Behavior Survey (MEBS; previously known as the Minnesota Eating Disorder Inventory [M-EDI]) was adapted and reproduced by special permission of Psychological Assessment Resources, Inc., 16204 North Florida Avenue, Lutz, Florida 33549, from the Eating Disorder Inventory (collectively, EDI and EDI-2) by Garner, Olmsted, Polivy, Copyright 1983 by Psychological Assessment Resources. Further reproduction of the MEBS is prohibited without prior permission from Psychological Assessment Resources.

  • 2

    Given that weight preoccupation varies by BMI (e.g., Baker and Galambos, 2003), the analyses were also run to examine the possible differences in the predictors of weight preoccupation in participants in the uppermost (BMI > 27.22; N = 71) and lowest quintiles (BMI < 20.02; N = 70) of BMI. The results indicated that neither ovarian hormones nor psychological factors predicted weight preoccupation changes in the lowest BMI quintile group. In the uppermost quintile group, the results suggested that, similar to the full sample, only emotional eating scores were significant predictors of weight preoccupation (b(SE) = .13 (.06); p = .04).

References

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  • 1
    Edler C, Lipson SF, Keel PK. Ovarian hormones and binge eating in bulimia nervosa. Psychol Med 2007;37:131141.
  • 2
    Hildebrandt T, Alfano L, Tricamo M, Pfaff DW. Conceptualizing the role of estrogens and serotonin in the development and maintenance of bulimia nervosa. Clin Psychol Rev 2010;30:655668.
  • 3
    Klump K, Keel P, Culbert K, Edler C. Ovarian hormones and binge eating: Exploring associations in community samples. Psychol Med 2008;38:17491757.
  • 4
    Racine SE, Culbert KM, Keel PK, Sisk CL, Burt SA, Klump KL. Differential associations between ovarian hormones and disordered eating symptoms across the menstrual cycle in women. Int J Eat Disord 2012;45:333.
  • 5
    Asarian L, Geary N. Modulation of appetite by gonadal steroid hormones. Philos Trans R Soc Lond B Biol Sci 2006;361:12511263.
  • 6
    Kemnitz JW, Gibber JR, Lindsay KA, Eisele SG. Effects of ovarian hormones on eating behaviors, body weight, and glucoregulation in rhesus monkeys. Horm Behav 1989;23:235250.
  • 7
    Blaustein JD, Wade GN. Ovarian influences on the meal patterns of female rats. Physiol Behav 1976;17:201208.
  • 8
    Klump KL, Suisman JL, Culbert KM, Kashy DA. The effects of ovariectomy on binge eating proneness in adult female rats. Horm Behav 2011;59:585593.
  • 9
    Van Strien T, Frijters JE, Bergers GP, Defares PB. The Dutch Eating Behavior Questionnaire (DEBQ) for assessment of restrained, emotional, and external eating behavior. Int J Eat Disord 1986;5:295315.
  • 10
    Gladis MM, Walsh BT. Premenstrual exacerbation of binge eating in bulimia. Am J Psychiatry 1987;144:15921595.
  • 11
    Lester NA, Keel P, Lipson S. Symptom fluctuation in bulimia nervosa: Relation to menstrual-cycle phase and cortisol levels. Psychol Med 2003;33:5160.
  • 12
    Price WA, Torem MS, Dimarzio LR. Premenstrual exacerbation of bulimia: 60% increase in the mean number of binge episodes. Psychosomatics 1987;28:378379.
  • 13
    Klump KL, Racine SE, Hildebrandt B, Burt SA, Neale M, Sisk CL, et al. Ovarian hormone influences on dysregulated eating: A comparison of associations in women with versus without binge episodes. Clin Psychol Sci, in press.
  • 14
    Klump K, Keel P, Racine S, Burt A, Neale M, Sisk C, et al. The interactive effects of estrogen and progesterone on changes in emotional eating across the menstrual cycle. J Abnorm Psychol 2013;122:131137.
  • 15
    Klump KL, McGue M, Iacono WG. Age differences in genetic and environmental influences on eating attitudes and behaviors in preadolescent and adolescent female twins. J Abnorm Psychol 2000;109:239251.
  • 16
    Jacobi C, Hayward C, de Zwaan M, Kraemer HC, Agras WS. Coming to terms with risk factors for eating disorders: Application of risk terminology and suggestions for a general taxonomy. Psychol Bull 2004;130:1965.
  • 17
    Wade TD, Zhu G, Martin NG. Undue influence of weight and shape: Is it distinct from body dissatisfaction and concern about weight and shape? Psychol Med 2011;41:819828.
  • 18
    Dennerstein L, Burrows GD. Affect and the menstrual cycle. J Affect Disord 1979;1:7792.
  • 19
    Ivey ME, Bardwick JM. Patterns of affective fluctuation in the menstrual cycle. Psychosom Med 1968;30:336345.
  • 20
    Burt SA, Klump KL. The Michigan State University Twin Registry (MSUTR): An update. Twin Res Hum Genet 2012;1:17.
  • 21
    Klump KL, Burt SA. The Michigan State University Twin Registry (MSUTR): Genetic, environmental and neurobiological influences on behavior across development. Twin Res Hum Genet 2006;9:971977.
  • 22
    von Ranson KM, Klump KL, Iacono WG, McGue M. The Minnesota Eating Behavior Survey: A brief measure of disordered eating attitudes and behaviors. Eat Behav 2005;6:373392.
  • 23
    Racine SE, Culbert KM, Larson CL, Klump KL. The possible influence of impulsivity and dietary restraint on associations between serotonin genes and binge eating. J Psychiatr Res 2009;43:12781286.
  • 24
    Van Strien T, Cleven A, Schippers G. Restraint, tendency toward overeating and ice cream consumption. Int J Eat Disord 2000;28:333338.
  • 25
    Watson D, Clark LA, Tellegen A. Development and validation of brief measures of positive and negative affect: The PANAS scales. J Pers Soc Psychol 1988;54:1063.
  • 26
    Davis C. Body image and weight preoccupation: A comparison between exercising and non-exercising women. Appetite 1990;15:1321.
  • 27
    Ukkola O, Gagnon J, Rankinen T, Thompson PA, Hong Y, Leon AS, et al. Age, body mass index, race and other determinants of steroid hormone variability: The HERITAGE Family Study. Eur J Endocrinol 2001;145:19.
  • 28
    Yoo KY, Kim H, Shin HR, Kang D, Ha M, Park SK, et al. Female sex hormones and body mass in adolescent and postmenopausal Korean women. J Korean Med Sci 1998;13:241246.
  • 29
    Racine SE, Keel PK, Burt SA, Sisk CL, Neale M, Boker S, et al. Individual differences in the relationship between ovarian hormones and emotional eating across the menstrual cycle: A role for personality? Eat Behav 2013;14:161166.
  • 30
    Steiger H, Gauvin L, Engelberg MJ, Ying Kin NMKN, Israel M, Wonderlich SA, et al. Mood-and restraint-based antecedents to binge episodes in bulimia nervosa: Possible influences of the serotonin system. Psychol Med 2005;35:15531562.
  • 31
    Thomas JG. Toward a better understanding of the development of overweight: A study of eating behavior in the natural environment using ecological momentary assessment. A Thesis. Philadelphia: Drexel University, 2009.
  • 32
    Wegner KE, Smyth JM, Crosby RD, Wittrock D, Wonderlich SA, Mitchell JE. An evaluation of the relationship between mood and binge eating in the natural environment using ecological momentary assessment. Int J Eat Disord 2002;32:352361.
  • 33
    Carr-Nangle RE, Johnson WG, Bergeron KC, Nangle DW. Body image changes over the menstrual cycle in normal women. Int J Eat Disord 1994;16:267273.
  • 34
    Hardie L, Trayhurn P, Abramovich D, Fowler P. Circulating leptin in women: A longitudinal study in the menstrual cycle and during pregnancy. Clin Endocrinol 1997;47:101106.
  • 35
    Shimizu H, Shimomura Y, Nakanishi Y, Futawatari T, Ohtani K, Sato N, et al. Estrogen increases in vivo leptin production in rats and human subjects. J Endocrinol 1997;154:285292.
  • 36
    Monteleone P, Di Lieto A, Tortorella A, Longobardi N, Maj M. Circulating leptin in patients with anorexia nervosa, bulimia nervosa or binge-eating disorder: Relationship to body weight, eating patterns, psychopathology and endocrine changes. Psychiatry Res 2000;94:121129.
  • 37
    Lewinsohn PM, Striegel-Moore RH, Seeley JR. Epidemiology and natural course of eating disorders in young women from adolescence to young adulthood. J Am Acad Child Adolesc Psychiatry 2000;39:12841292.
  • 38
    Lucero LD, Hill FA, Ferraro FR. Body dissatisfaction in young children. Psychology 1999;36:3642.
  • 39
    Schur EA, Sanders M, Steiner H. Body dissatisfaction and dieting in young children. Int J Eat Disord 1999;27:7482.
  • 40
    Killen JD, Taylor CB, Hayward C, Wilson DM, Haydel KF, Hammer LD, et al. Pursuit of thinness and onset of eating disorder symptoms in a community sample of adolescent girls: A three-year prospective analysis. Int J Eat Disord 1994;16:227238.