Abstract
Purpose
Existing literature has demonstrated that appetite hormones are frequently dysregulated in individuals with bulimic-spectrum eating disorders (BN-EDs). Although dysregulations in appetite hormones may maintain BN-EDs, very limited research has examined the association between dysregulated appetite hormones and cognitive and behavioral bulimic symptoms. We hypothesized that greater frequency of behavioral symptoms and severity of cognitive symptoms of BN-EDs would correlate with greater dysregulation in appetite hormones.
Methods
The association between ghrelin, cortisol, leptin, GLP-1, and amylin levels and eating pathology was examined in treatment-seeking adults with BN-EDs (N = 33). Participants completed bloodwork to assess fasting blood hormone levels and bulimic symptoms were measured by the Eating Disorder Examination. Pearson partial correlations were run to examine the association between hormone levels and eating pathology, controlling for BMI.
Results
Contrary to hypotheses, none of the appetite hormones tested were significantly associated with frequency of behavioral ED symptoms (p range = 0.13–0.97, negligible to small effect sizes). Global eating pathology was positively associated with leptin (p = 0.03) and negatively associated with GLP-1 (p = 0.03) and amylin (p = 0.04), with medium effect sizes. Post hoc analyses indicated significantly stronger associations between appetite hormones and cognitive eating pathology than between appetite hormones and frequency of binge eating [GLP-1 (p = 0.02) and amylin (p = 0.02)] or compensatory behaviors [leptin (p = 0.03), GLP-1 (p = 0.02), and amylin (p = 0.04)].
Conclusion
In individuals with BN-EDs, appetite hormones may be more strongly associated with cognitive symptoms than behavioral symptoms.
Level of evidence
Level V, cross-sectional descriptive study.
Keywords: Appetite hormones, Bulimia nervosa, Binge eating, Neuroendocrinology
Introduction
Individuals with bulimia nervosa (BN) and other bulimic-spectrum eating disorders often engage in irregular eating patterns (e.g., eating large quantities of food during binge-eating episodes, going long periods of time without eating, purging) that may lead to dysregulations in metabolic hormones impacting appetite and body weight. Appetite hormones are signaling molecules and can roughly be divided into two broad categories (with some caveats, see Begg, Woods [1] for a full review of the function of appetite hormones): (1) hormones that stimulate appetite and promote increased food consumption (ghrelin and cortisol) and (2) hormones that suppress appetite and inhibit food consumption (leptin, glucagon-like peptide 1 (GLP-1), and amylin [1]). Fasting levels of appetite hormones are associated with body composition and weight, such that higher body weight is associated with lower fasting levels of appetite stimulating hormones and higher fasting levels of appetite suppressing hormones [1-3].
Dysregulated levels of appetite hormones may contribute to the maintenance of bulimic symptoms given the importance of these hormones in governing eating behaviors and preliminary evidence suggests that the severity of current behavioral disordered eating symptoms (e.g., binge eating, inappropriate compensatory behaviors) can contribute to further dysregulation of appetite hormone levels [4-6]. Identifying which bulimic symptoms are most strongly associated with appetite hormone levels is important for identifying patient groups most at risk of severely dysregulated hormone levels and medical complications. Additionally, understanding the bidirectional maintenance of dysregulated appetite hormone levels and bulimic symptoms is essential for informing treatment development, psychoeducation, and dietetic recommendations for individuals with eating disorders (EDs). Examining the symptom-level impact of appetite hormones in individuals with bulimic-spectrum EDs will grant insight into how physiological pressures contribute to the maintenance of disordered eating and, thus, is a clear research priority.
Existing research on appetite hormones in BN has largely focused on understanding how individuals with BN differ from healthy controls. While most studies have found that individuals with BN experience dysregulated fasting appetite hormones, the previous research has not always produced consistent patterns. For example, most studies to date have found that fasting ghrelin levels in individuals with BN are comparable to levels in healthy individuals [7, 8], although some evidence suggests that ghrelin release immediately before and after eating may be increased relative to healthy controls [4]. While the research on some appetite hormones is relatively consistent (e.g., fasting cortisol levels tend to be reliably higher in individuals with BN [4] and fasting GLP-1 levels are reliably lower in BN [4, 9, 10]), others have produced mixed findings. For example, while some research suggests that leptin may be decreased in individuals with BN [4, 9], the pattern of results is far from consistent, with the other studies finding no difference or even higher leptin levels in individuals with BN [8]. For some appetite hormones like amylin, there has been very limited research examining fasting and post-meal levels in individuals with BN, though one research study found that both fasting and post-meal amylin release in individuals with BN was comparable to healthy controls [11]. As a whole, the existing body of research highlights the complexity of examining hormone levels in individuals with EDs and suggests a clear need to understand what factors specifically contribute to hormone dysregulations for individuals with BN (for a complete review of the literature, see Culbert et al. [4] and Milano, Capasso [7]).
We believe that some of the mixed results found to date in appetite hormones in BN may be due to the methodologies used in these studies. For example, the majority of extant studies have failed to consider variability in the frequency and severity of BN symptoms as a possible factor associated with hormone dysregulation, as they have primarily compared individuals who meet diagnostic criteria for BN to healthy controls [7, 8]. While a small number of studies have considered the association between ghrelin [4, 5] and leptin [6] levels and frequency of BN behaviors, no studies have tested this relationship for amylin, GLP-1, or cortisol. The limited extant research suggests that circulating ghrelin levels may be negatively associated with frequency of binge eating and purging behaviors [5], whereas leptin levels may be positive associated with frequency of binge eating [6]. Even for hormones that may not be dysregulated in individuals with BN relative to healthy controls (e.g., amylin), there may still be meaningful associations between frequency of certain behaviors that may be present in BN (e.g., binge eating, dietary restriction) and these hormone levels, based on the literature identifying dysregulations in these hormones in other populations with disordered eating behaviors (e.g., anorexia nervosa [12], food addiction [13]). The association between dysregulated appetite hormone levels and other markers of BN severity like the cognitive symptoms of an ED (e.g., weight and shape concerns, concern about eating) are even less well tested. Additional research is needed to better understand the association between appetite hormone dysregulation and bulimic symptomatology severity.
In the present study, we examined the association between fasting appetite hormones and bulimic behaviors and cognitions in individuals with bulimic-spectrum EDs. Consistent with the limited extant research, we hypothesized that higher levels of ghrelin would be associated with lower frequency of binge episodes and compensatory behaviors, and higher levels of leptin would be associated with higher frequency of binge episodes and compensatory behaviors. We hypothesized that higher levels of cortisol and lower levels of GLP-1 and amylin would be associated with higher frequency of binge episodes and compensatory behaviors. Given the dearth of extant literature examining the association between appetite hormones and cognitive ED symptoms, we did not make a directional hypothesis regarding the association between hormone levels and cognitive eating pathology.
Methods
Participants
Participants (N = 33) were treatment-seeking adults recruited from the community as part of a larger parent treatment study for bulimic-spectrum EDs. Only baseline assessment data were included in the present study, and all participants from the parent study for whom hormone data were available were included. Participants were eligible for the study if they endorsed at least 12 objectively or subjectively large binge-eating episodes and at least 12 episodes of compensatory behaviors (including purging behaviors, excessive/driven exercise, and other extreme weight control behaviors). Within our sample, 29 participants (87.9%) met behavioral criteria for Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-5; American Psychiatric Association [14]) bulimia nervosa, two participants (6.1%) met behavioral criteria for Other Specified Feeding or Eating Disorder (OSFED), bulimia nervosa of low frequency and/or limited duration, and two participants (6.1%) met behavioral criteria for bulimia nervosa with only subjectively-large binge-eating episodes. Participants were excluded from the present study if their weight was below 85% of their ideal body weight, as that may have indicated a more appropriate diagnosis of anorexia nervosa [15]. Participants mostly identified as women (N = 29, 87.9%). Participants’ self-reported race/ethnicity was: 69.7% White (N = 23), 6.1% African American/Black (N = 2), 3.0% Caribbean/Haitian (N = 1), 9.1% Asian American (N = 3), and 12.1% Latino/ Latina/Hispanic (N = 4).
Procedures
Participants were recruited from the community, screened for initial eligibility via a phone screen, and subsequently completed a baseline assessment battery in-person at a hospital and an academic psychological services center. Study procedures were approved and overseen by the Drexel University Institutional Review Board, and informed consent was obtained from all participants. Participants received monetary compensation for completing the assessment battery. To obtain fasting hormone levels, participants were instructed to fast overnight beginning at 8:00 PM and venous blood sample collection occurred between approximately 8:00 AM and 9:00 AM. Two 10 cc blood samples were collected into empty, sterile, 10 mL red-top tubes (REF 367820) and sat for 15–20 min post-blood draw, and then were centrifuged for 20 min at 3400 RPM. Serum was pipetted into six 1.5 mL cryogenic tubes and then stored at −80 °C until batch analysis upon study completion. Samples were collected at the Children’s Hospital of Philadelphia Center for Human Phenomic Science Research Nursing Core and assayed at the Children’s Hospital of Philadelphia Translational Core Laboratory. Aliquots were boxed by analytes and transported on dry ice. Tube verification and data entry were double-checked for consistency.
Measures
Participants completed a variety of measures at the baseline assessment, but only those included in the present study are described below.
Eating Disorder Examination
The Eating Disorder Examination [16] is a widely used, semi-structured interview for ED symptomatology with acceptable psychometric reliability and validity [17]. Eating Disorder Examination (EDE) Global and subscale scores (Restraint, Eating Concern, Weight Concern, and Shape Concern) were used to measure cognitive ED symptoms and assessment of behavioral bulimic symptom frequency, including binge eating (i.e., subjectively and objectively large episodes) and inappropriate compensatory behaviors (i.e., self-induced vomiting, laxative misuse, diuretic misuse, excessive/driven exercise, other extreme weight control behaviors) were included in the present study. The EDE Restraint subscale measures severity of attempted, rather than actual, dietary restraint (e.g., consciously trying to restrict overall amount eaten, whether or not successful), and as such, we classified attempted dietary restraint as a cognitive ED symptom, even though attempted dietary restraint may coincide with actual dietary restraint. All EDEs were conducted by Bachelor’s- or Master’s-level independent assessors with intensive training (e.g., 100% ED diagnostic agreement and > 0.80 interrater reliability) and were supervised by the principal investigator, a licensed psychologist.
Height, weight, and BMI
Height and weight were measured by study coordinators using a digital scale and stadiometer, and BMI (kg/m2) was calculated from these values.
Appetite hormones
Although appetite and eating behaviors are only impacted by the biologically active forms of ghrelin and GLP-1, total fasting levels of these hormones reflect the accumulated release of biologically active forms. Additionally, the previous research examining appetite hormones in BN has found that total fasting ghrelin and GLP-1 levels are dysregulated in individuals with BN [10, 18], and total fasting levels of these hormones may be associated with bulimic behaviors [5, 10]. As such, we determined that total fasting levels of ghrelin and GLP-1 constitute a sufficient proxy for levels of the biologically active forms and we elected to include fasting levels of these hormones in our analyses. Blood aliquots were measured using commercially available enzyme-linked immunosorbent assay (ELISA) kits designed for each hormone analyte. Procedures were followed as outlined in the manual for each assay, including running controls. Amylin (pM), leptin (pg/mL), cortisol (μg/dL), total ghrelin (pg/mL), and total GLP-1 (pM) were measured using the Human Amylin kit (Millipore-Sigma, EZHA-52K), Human Leptin kit (R&D, DLP00), Cortisol II kit (Roche Dx, 06687733-160) Ghrelin, total kit (Millipore-Sigma, EZGRT-89K), and GLP-1, total kit (Millipore-Sigma, EZGLP1T-36K), respectively. Amylin levels below 2.734 pM were designated as < 2.734 pM and the actual value of amylin concentration was not specified. Amylin values for these participants (N = 2) were recoded as 2.734 pM for data analysis. Intraassay percent coefficients of variability (%CV) were collected for amylin (mean = 4.49, range = 0.0–17.6), leptin (mean = 4.17, range = 0.1–16.0), total ghrelin (mean = 4.05, range = 0.0–73.6), and total GLP-1 (mean = 4.02, range = 0.2–12.5). Intraassay %CV were not collected for cortisol, as cortisol assays were run on a clinical-grade analyzer tightly controlled with controls and calibrators. The %CV for one amylin measurement was below the detectable range and the %CV for one ghrelin measurement was high (> 25%; %CV = 73.6).
Data analysis
Data analyses were conducted using SPSS software version 24 for Mac and p values of ≤ 0.05 were considered significant. Because the hormone level data were non-normally distributed, the data were log-transformed to correct for skew, and then we ran Pearson’s correlations on the transformed data. Given the strong relationship between body mass index (BMI) and appetite hormones, Pearson’s correlations were run to examine the association between BMI and hormone levels. For all hormones, we ran analyses both with and without controlling for BMI. Partial correlations were used to examine the relationship between hormone levels and total binge episodes (both subjectively and objectively large), total compensatory behaviors, and total purging (i.e., self-induced vomiting, laxative use, and diuretic use) episodes over the past 3 months and global ED pathology over the past 28 days, controlling for BMI. Effect sizes for Pearson’s correlations are reported using the conventions proposed by Cohen [19, 20]: small (0.10 ≤∣r∣< 0.30), medium (0.30 ≤∣r∣< 0.50), or large (0.50 ≤∣r∣< 1.00).
As an exploratory analysis, additional post hoc partial correlations were run for hormones that were significantly associated with global ED pathology to determine which subscale(s) were driving the relationship. Post hoc analyses were also run to compare the strength of correlations between hormone levels and behavioral and global cognitive ED symptoms for the hormones that were significantly associated with EDE Global score. To compare these correlations, the correlation coefficients (i.e., the correlation between hormone level and behavioral symptom, between hormone level and EDE Global score, and between behavioral symptom and EDE Global score) were transformed to z-scores using Fischer’s Z-transformation. Then, Steiger’s [21] Eqs. (3) and (10) were used to compute asymptotic covariance of the estimates and to run asymptotic z-tests to compare the transformed correlation coefficients. Based on power analyses conducted with N = 33 and p ≤ 0.05, we will be able to detect medium to large effect sizes at ≥ 80% power.
Results
Demographics, bulimic symptoms, and hormone level descriptive statistics are summarized in Table 1. As shown in Table 1, higher BMI was significantly and strongly associated with lower levels of ghrelin and cortisol, and very strongly associated with higher levels of leptin.
Table 1.
Sample characteristics and median fasting blood hormone levels (N = 33)
| Median | Interquartile range | Range | Association with BMI ρ (p) |
|
|---|---|---|---|---|
| Age | 28.00 | 15.00 | 18–57 | |
| BMI | 24.84 kg/m2 | 8.35 kg/m2 | 17.90–40.67 kg/m2 | |
| Binge episodes | 63.00 | 68.00 | 9–212 | |
| Compensatory behaviors | 69.00 | 102.50 | 12–326 | |
| Purging episodes | 36.00 | 67.00 | 0–219 | |
| EDE Global | 3.38 | 1.46 | 0.54–5.41 | |
| EDE Restraint | 3.80 | 1.50 | 0.00–6.00 | |
| EDE Eating Concern | 2.20 | 1.90 | 0.00–5.80 | |
| EDE Weight Concern | 3.60 | 1.70 | 0.00–6.00 | |
| EDE Shape Concern | 4.13 | 1.73 | 0.25–6.00 | |
| Ghrelin | 472.22 pg/mL | 400.32 pg/mL | 193.86–3746.03 pg/mL | −0.63 (0.00)*** |
| Cortisol | 14.95 μg/dL | 8.66 μg/dL | 5.71–40.93 μg/dL | −0.55 (0.00)*** |
| Leptin | 10,654.80 pg/mL | 16,608.30 pg/mL | 1157.70–56,120.10 pg/mL | 0.82 (0.00)*** |
| GLP-1 | 16.09 pM | 9.77 pM | 7.45–75.87 pM | 0.31 (0.08) |
| Amylin | 16.02 pM | 23.67 pM | 2.73–415.08 pM | 0.08 (0.67) |
Descriptive statistics (median, interquartile range, range) calculated from non-transformed data; associations between hormone levels and BMI calculated with log-transformed data
BMI body mass index, EDE Eating Disorder Examination, GLP-1 glucagon-like peptide 1, kg/m2 kilograms/meter2, pg/mL picograms per milliliter, μg/dL micrograms per deciliter, pM picomolar
Designates significance at the p ≤ 0.001 level
Associations between hormone levels and eating pathology while controlling for BMI are summarized in Table 2. Analyses run without controlling for BMI largely yielded similar patterns in direction and effect sizes, although some associations that met statistical significance without controlling for BMI were no longer significant when controlling for BMI (these associations were between ghrelin level and EDE Global score, cortisol and EDE Global score, leptin and EDE Shape Concern, and leptin and EDE Weight Concern).The association between leptin and total binge episodes changed direction when controlling for BMI (the association was negative without controlling for BMI and positive when controlling for BMI), although effect sizes were small (when controlling for BMI) or negligible (not controlling for BMI).
Table 2.
Associations between hormone levels and eating pathology controlling for BMI (N = 33)
| Ghrelin ρ (p) |
Cortisol ρ (p) |
Leptin ρ (p) |
GLP-1 ρ (p) |
Amylin ρ (p) |
|
|---|---|---|---|---|---|
| Binge episodes | 0.19 (0.30) | 0.28 (0.13) | 0.13 (0.49) | −0.01 (0.97) | 0.07 (0.69) |
| Compensatory behaviors | −0.15 (0.41) | 0.25 (0.17) | −0.01 (0.95) | 0.03 (0.86) | −0.01 (0.95) |
| Purging episodes | 0.05 (0.80) | 0.21 (0.26) | 0.10 (0.60) | −0.28 (0.13) | −0.14 (0.44) |
| EDE Global | −0.20 (0.27) | −0.21 (0.25) | 0.38 (0.03)* | −0.40 (0.03)* | −0.36 (0.04)* |
| EDE Restraint | 0.42 (0.02)* | −0.27 (0.13) | −0.36 (0.05)* | ||
| EDE Eating Concern | 0.32 (0.07) | −0.45 (0.01)** | −0.35 (0.05)* | ||
| EDE Weight Concern | 0.25 (0.17) | −0.36 (0.04)* | −0.26 (0.16) | ||
| EDE Shape Concern | 0.22 (0.23) | −0.22 (0.23) | −0.23 (0.20) |
Analyses conducted on log-transformed data
EDE Eating Disorder Examination, GLP-1 glucagon-like peptide 1
Designates significance at the p ≤ 0.05 level
Designates significance at the p ≤ 0.01 level
Contrary to our hypotheses, associations between appetite hormones and global ED pathology were stronger than associations with bulimic behaviors (see Table 3 for comparisons of correlation strength). None of the appetite hormones which we examined were associated with frequency of binge eating, total compensatory behaviors, or purging episodes with all effect sizes in the small or negligible ranges.
Table 3.
Comparing strength of correlations between hormone levels and behavioral and global cognitive eating pathology controlling for BMI (N = 33)
| Binge episodes vs. EDE Global z (p) |
Compensatory behaviors vs. EDE Global z (p) |
Purging episodes vs. EDE Global z (p) |
|
|---|---|---|---|
| Leptin | −1.27 (0.10) | −1.94 (0.03)* | −1.13 (0.13) |
| GLP-1 | 1.98 (0.02)* | 2.16 (0.02)* | 0.50 (0.31) |
| Amylin | 2.17 (0.02)* | 1.73 (0.04)* | 0.89 (0.19) |
Analyses conducted on log-transformed data
EDE Eating Disorder Examination, GLP-1 glucagon-like peptide 1
Designates significance at the p ≤ 0.05 level
We did observe several significant associations between hormone levels and global eating pathology as measured by the EDE. Lower levels of GLP-1 and amylin were both significantly associated with greater global eating pathology, and higher levels of leptin were associated with greater global eating pathology. All effect sizes were in the medium range. Post hoc analyses revealed no consistent pattern in which subscales drove these results, with the Restraint, Eating Concern, and Weight Concern subscales all showing a significant result for at least one hormone (Shape Concern was not significantly associated with any hormone level).
When comparing the strength of correlations between leptin, GLP-1, and amylin levels with behavioral and cognitive eating disorder symptoms, our results suggest that these hormones are more strongly associated with cognitive eating disorder symptoms (i.e., global eating pathology) than with behavioral eating disorder symptoms (i.e., binge episodes, compensatory behaviors, and purging episodes; see Table 3).
Discussion
Overall, our findings suggest that appetite hormone levels may be more strongly associated with cognitive symptoms of EDs than with the frequency of behavioral symptoms of bulimic-spectrum EDs. The previous research has presumed that appetite hormone dysregulation may maintain EDs by contributing to the maintenance of behavioral symptoms like binge eating. Contrary to our hypotheses, none of the hormones tested were significantly associated with the frequency of binge eating, purging, or total compensatory behaviors in our sample. Specifically, we found non-significant positive correlations between ghrelin levels and frequency of binge episodes and purging episodes, which contradict findings by Troisi et al. [5] that suggested negative associations between circulating ghrelin levels and frequency of binge eating and purging. The directionality of the associations between ghrelin levels and total compensatory behaviors in our sample was consistent with the previous findings and our hypothesis, although the effect size was small [5]. Our results indicating a positive association between leptin levels and frequency of binge episodes were consistent with findings by Baker et al. [6] and our hypothesis. Broadly, we found that appetite hormone levels were non-significantly associated with behavioral bulimic symptoms with negligible to small effect sizes.
Our results may differ from the findings by Troisi et al. [5] in part due to differences in population between the two studies. Troisi et al. [5] examined the association between ghrelin levels and bulimic behaviors in a sample of individuals with BN diagnosed according to DSM-IV diagnostic criteria, which requires binge-eating episodes and inappropriate compensatory behaviors to have occurred at least twice per week, on average, over the previous 3 months [22]. In contrast, our sample included individuals diagnosed in accordance with DSM-5 behavioral criteria for BN, individuals with Other Specified Feeding or Eating Disorder, bulimia nervosa of low frequency and/or limited duration, and individuals with bulimia nervosa with only subjectively large binge-eating episodes. Furthermore, we examined the association between ghrelin level and frequency of loss of control eating episodes, regardless of whether binge episodes were objectively large. These differences in behavioral frequency and severity between the two studies might explain why we did not observe a consistent pattern in the association between ghrelin level and binge-eating frequency, which may suggest that ghrelin levels are differentially related to the cognitive experience of loss of control and the consumption of objectively large amounts of food; since the previous literature comparing appetite hormones in individuals with BN and individuals with purging disorder has identified differences in fasting GLP-1 levels between individuals with BN and purging disorder [10], including individuals with only subjectively large binge-eating episodes may have prevented the identification of an association between objectively large eating episodes and GLP-1 dysregulation. However, some appetite hormones (i.e., ghrelin [23] and leptin [24]) demonstrate a similar pattern of dysregulation in individuals with purging disorder and individuals with BN, suggesting that the inclusion of individuals with only subjectively large binge-eating episodes may not impact the validity of our findings for these hormones.
Another possible explanation of these mixed findings is that the association between bulimic behaviors and appetite hormones may be affected by a third variable, such as duration of illness. For instance, at shorter durations of illness, bulimic behaviors may not contribute to dysregulation of appetite hormone levels, even if at longer durations of illness, these behaviors are associated with substantially dysregulated hormone levels. Given that we did not collect information about duration of illness, we compared ghrelin levels in younger participants to ghrelin levels in older participants using a mean split (mean age = 28.00 years) as an exploratory post hoc analysis, based on the assumption that age may serve as a proxy for duration of illness. We found that older participants had significantly higher ghrelin levels than younger participants (t = 2.02, p = 0.05). Aging is typically associated with decreasing ghrelin levels [25], so this finding offers very preliminary evidence that hormone equilibria may be differently affected by behavioral bulimic symptoms as age or duration of illness increases. Alternatively, hormone levels may not be associated with binge eating and compensatory behaviors, but instead may be governed primarily by weight and nutrition status for most individuals with bulimic-spectrum EDs. Consistent with this notion, we did find that BMI demonstrated strong associations with some (ghrelin, cortisol, and leptin) but not all (amylin and GLP-1) appetite hormone levels. Analogously, weight suppression, or the difference between an individual’s historical highest weight and his or her current weight, may be a third variable that explains the mixed findings in the literature. Weight suppression has been implicated in contributing to dysregulated appetite hormones in individuals with EDs [26, 27] and there is robust evidence that weight suppression is associated with cognitive ED symptoms, including weight and shape concerns, fear of weight gain, body dissatisfaction, and fear of losing control of weight [28]. Thus, the impact of weight suppression on these relationships warrants future study.
Surprisingly, our results suggest that appetite hormone levels are more strongly associated with severity of cognitive symptoms of bulimic-spectrum EDs than with behavioral ED symptoms. As we do not theoretically expect cognitive ED symptoms to impact hormone levels (except potentially through ED behaviors), these findings could suggest that altered hormone levels contribute to greater ED cognitions in individuals with bulimic-spectrum EDs. For example, it is possible that individuals with greater appetite dysregulation experience more frequent urges to binge eat (which could increase eating concerns) even if they are able to resist some of these urges by engaging in heightened dietary restraint. However, since no causal conclusions may be drawn from this cross-sectional study design, the directionality of the association between appetite hormone levels and cognitive symptoms of EDs is an area for future study.
The present study adds to the growing body of literature on hormone dysregulation in EDs. By clarifying the extent to which hormones correlate with eating pathology, we improve our understanding of the biological underpinnings of EDs. Understanding the endocrinology of EDs can inform prevention and treatment efforts. Our findings suggesting an association between hormone dysregulation and cognitive ED symptoms may set the stage for future research identifying hormonal disturbances indicative of ED onset risk. Furthermore, the inconsistency of our findings with extant research examining the association between behavioral ED symptoms and hormonal disturbances suggest that more research is needed to examine how individual psychobiological profiles change over time, including during treatment and longitudinally across the lifespan. Finally, our findings contribute important information that can be used in psychoeducation for patients with EDs about the impact of these disorders on physiological processes, which patients may find validating in regard to how their responses to food and shape and weight are altered.
Limitations and future directions
Our study faced several limitations, which are largely shared by the existing literature examining hormone levels in ED populations. Due to an oversight in protocol, protease inhibitors were not added to the samples during collection, which may have impacted sample stability of total GLP-1. Our study involved a relatively small-sample size and assessed fasting hormone levels only. Dysregulations in pre- or post-meal changes in hormone levels may be more strongly associated with bulimic symptoms than fasting levels. Although we elected to include total fasting levels of ghrelin and GLP-1 in our analyses, based on precedence in the literature for using total fasting levels of these hormones as proxies for the active forms, our findings may not fully capture the dysregulations in these hormone levels associated with BN as only the active forms are implicated in impacting eating behavior. Some evidence also suggests that inactive and active ghrelin may have opposite effects on food intake and gastric emptying [29], which may undermine the validity of our findings based on total ghrelin concentration. Measurement of active ghrelin and GLP-1 may more precisely identify the relative effects of these hormones on appetite and eating behavior, and, thus, is a clear area for additional research in individuals with eating disorders.
The study utilized a cross-sectional design, inhibiting our ability to make conclusions about the extent to which hormone dysregulation contributes to the development of disordered eating or is secondary to the impact of disordered eating on nutrition status. The variability in our sample, in hormone levels, and ED symptoms, was both a strength and limitation. Our sample was likely more representative of the population of individuals with bulimic-spectrum EDs relative to studies that have examined only individuals with specific types of bulimic symptoms (e.g., purging), increasing generalizability. However, the high variability when combined with a small-sample size can make results more difficult to interpret. By controlling for BMI in analyses, we were able to elucidate the relationships between hormone levels and BN symptoms without the confound of BMI. However, this prevents any capacity to determine whether BMI moderates these associations. Our study was strengthened by the use of a diagnostic interview to assess the frequency of bulimic behaviors and severity of cognitive ED symptoms.
Future research should continue to examine the association between hormone levels and behavioral and cognitive ED symptoms in larger samples, including individuals in multiple levels of care. Future investigation should also examine whether BMI is a moderator of these associations in a larger sample with a diverse range of BMIs. Additionally, future research should utilize longitudinal designs to examine changes in hormone levels over the course of illness, including identifying the extent to which hormone levels normalize with normalization of eating patterns due to treatment or remission. Identifying the temporal relationship between changes in hormone levels and changes in ED behaviors will facilitate the field’s understanding of the etiologic or maintenance role that dysregulated hormone levels play in eating pathology.
What is already known on this subject?
Appetite hormones are impacted in EDs, but associations between ED symptoms and hormone levels are poorly understood. We examine associations between specific ED symptoms and appetite hormones.
What your study adds?
Appetite hormones levels may be associated with cognitive ED symptom severity. Our findings may inform treatment efforts and contribute to psychoeducation about the biological underpinnings of EDs.
Acknowledgements
This research was developed under a parent study funded by a grant from the National Institute of Mental Health (Award Number: K23MH105680). We thank Audrey Kamrin, MSN, and colleagues at the Children’s Hospital of Philadelphia Center for Human Phenomic Science Research Nursing Core and Xiangdong (Sean) Ren, MD, PhD and colleagues at the Children’s Hospital of Philadelphia Translational Core Laboratory for their contributions to this study.
Footnotes
Conflict of interest The authors declare that they have no conflicts of interest.
Code availability Not applicable.
Ethics approval Study procedures were approved and overseen by the Drexel University Institutional Review Board (IRB number 1501003366). All procedures conducted in this study were performed in accordance with ethical standards as established in the 1963 Declaration of Helsinki and its later amendments.
Consent to participate Informed consent was obtained from all individual participants included in this study.
Consent for publication All participants consented to publication of findings based on data collected in this study.
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Data availability
The datasets analyzed during the current study are not publicly available as the primary outcomes of the parent study have not yet been published, but are available from the corresponding author on reasonable request.
References
- 1.Begg DP, Woods SC (2013) The endocrinology of food intake. Nat Rev Endocrinol 9(10):584–597. 10.1038/nrendo.2013.136 [DOI] [PubMed] [Google Scholar]
- 2.Lean ME, Malkova D (2016) Altered gut and adipose tissue hormones in overweight and obese individuals: cause or consequence? Int J Obes (Lond) 40(4):622–632. 10.1038/ijo.2015.220 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Travison TG, O’Donnell AB, Araujo AB, Matsumoto AM, McKinlay JB (2007) Cortisol levels and measures of body composition in middle-aged and older men. Clin Endocrinol (Oxf) 67(1):71–77. 10.1111/j.1365-2265.2007.02837.x [DOI] [PubMed] [Google Scholar]
- 4.Culbert KM, Racine SE, Klump KL (2016) Hormonal factors and disturbances in eating disorders. Curr Psychiatry Rep 18(7):65. 10.1007/s11920-016-0701-6 [DOI] [PubMed] [Google Scholar]
- 5.Troisi A, Di Lorenzo G, Lega I, Tesauro M, Bertoli A, Leo R, Iantorno M, Pecchioli C, Rizza S, Turriziani M, Lauro R, Siracusano A (2005) Plasma ghrelin in anorexia, bulimia, and binge-eating disorder: relations with eating patterns and circulating concentrations of cortisol and thyroid hormones. Neuroendocrinology 81(4):259–266. 10.1159/000087923 [DOI] [PubMed] [Google Scholar]
- 6.Baker JH, Peterson CM, Thornton LM, Brownley KA, Bulik CM, Girdler SS, Marcus MD, Bromberger JT (2017) Reproductive and appetite hormones and bulimic symptoms during midlife. Eur Eat Disord Rev 25(3):188–194. 10.1002/erv.2510 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Milano W, Capasso A (2018) Neuroendocrine and metabolic disorders in bulimia nervosa. Endocr Metab Immune Disord Drug Targets 18:297–305 [DOI] [PubMed] [Google Scholar]
- 8.Tortorella A, Brambilla F, Fabrazzo M, Volpe U, Monteleone AM, Mastromo D, Monteleone P (2014) Central and peripheral peptides regulating eating behaviour and energy homeostasis in anorexia nervosa and bulimia nervosa: a literature review. Eur Eat Disord Rev 22:307–320 [DOI] [PubMed] [Google Scholar]
- 9.Cuesto G, Everaerts C, Leon LG, Acebes A (2017) Molecular bases of anorexia nervosa, bulimia nervosa and binge eating disorder: shedding light on the darkness. J Neurogenet 31(4):266–287. 10.1080/01677063.2017.1353092 [DOI] [PubMed] [Google Scholar]
- 10.Dossat AM, Bodell LP, Williams DL, Eckel LA, Keel PK (2015) Preliminary examination of glucagon-like peptide-1 levels in women with purging disorder and bulimia nervosa. Int J Eat Disord 48(2):199–205. 10.1002/eat.22264 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Maske CB, Williams DL, Keel PK (2020) Preliminary examination of insulin and amylin levels in women with purging disorder. Int J Eat Disord. 10.1002/eat.23230 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Singhal V, Misra M, Klibanski A (2014) Endocrinology of anorexia nervosa in young people: recent insights. Curr Opin Endocrinol Diabetes Obes 21(1):64–70. 10.1097/MED.0000000000000026 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Pedram P, Sun G (2015) Hormonal and dietary characteristics in obese human subjects with and without food addiction. Nutrients 7(1):223–238. 10.3390/nu7010223 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.American Psychiatric Association (2013) Diagnostic and statistical manual of mental disorders: DSM-5, 5th edn. Arlington, VA [Google Scholar]
- 15.Brown TA, Holland LA, Keel PK (2014) Comparing operational definitions of DSM-5 anorexia nervosa for research contexts. Int J Eat Disord 47(1):76–84. 10.1002/eat.22184 [DOI] [PubMed] [Google Scholar]
- 16.Cooper Z, Fairburn C (1987) The eating disorder examination: a semi-structured interview for the assessment of the specific psychopathology of eating disorders. Int J Eat Disord 6(1):1–8 [Google Scholar]
- 17.Berg KC, Peterson CB, Frazier P, Crow SJ (2012) Psychometric evaluation of the eating disorder examination and eating disorder examination-questionnaire: a systematic review of the literature. Int J Eat Disord 45(3):428–438. 10.1002/eat.20931 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Tanaka M, Naruo T, Muranaga T, Yasuhara D, Shiiya T, Nakazato M, Matsukura S, Nozoe S (2002) Increased fasting plasma ghrelin levels in patients with bulimia nervosa. Eur J Endocrinol 146(6):R1–3. 10.1530/eje.0.146r001 [DOI] [PubMed] [Google Scholar]
- 19.Cohen J (1988) Statistical power analysis for the behavioral sciences, 2nd edn. Routledge, Abingdon [Google Scholar]
- 20.Cohen J (1992) A power primer. Psychol Bull 112(1):155–159. 10.1037//0033-2909.112.1.155 [DOI] [PubMed] [Google Scholar]
- 21.Steiger JH (1980) Tests for comparing elements of a correlation matrix. Psychol Bull 87(2):245–251. 10.1037/0033-2909.87.2.245 [DOI] [Google Scholar]
- 22.American Psychiatric Association (2000) Diagnostic and statistical manual of mental disorders. 4th edn., Washington, DC [Google Scholar]
- 23.Keel PK, Eckel LA, Hildebrandt BA, Haedt-Matt AA, Appelbaum J, Jimerson DC (2018) Disturbance of gut satiety peptide in purging disorder. Int J Eat Disord 51(1):53–61. 10.1002/eat.22806 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Jimerson DC, Wolfe BE, Carroll DP, Keel PK (2010) Psychobiology of purging disorder: reduction in circulating leptin levels in purging disorder in comparison with controls. Int J Eat Disord 43(7):584–588. 10.1002/eat.20738 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Amitani M, Amitani H, Cheng KC, Kairupan TS, Sameshima N, Shimoshikiryo I, Mizuma K, Rokot NT, Nerome Y, Owaki T, Asakawa A, Inui A (2017) The role of ghrelin and ghrelin signaling in aging. Int J Mol Sci. 10.3390/ijms18071511 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Bodell LP, Keel PK (2015) Weight suppression in bulimia nervosa: associations with biology and behavior. J Abnorm Psychol 124(4):994–1002. 10.1037/abn0000077 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Keel PK, Bodell LP, Haedt-Matt AA, Williams DL, Appelbaum J (2017) Weight suppression and bulimic syndrome maintenance: preliminary findings for the mediating role of leptin. Int J Eat Disord 50(12):1432–1436. 10.1002/eat.22788 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Lowe MR, Piers AD, Benson L (2018) Weight suppression in eating disorders: a research and conceptual update. Curr Psychiatry Rep 20(10):80. 10.1007/s11920-018-0955-2 [DOI] [PubMed] [Google Scholar]
- 29.Asakawa A, Inui A, Fujimiya M, Sakamaki R, Shinfuku N, Ueta Y, Meguid MM, Kasuga M (2005) Stomach regulates energy balance via acylated ghrelin and desacyl ghrelin. Gut 54(1):18–24. 10.1136/gut.2004.038737 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets analyzed during the current study are not publicly available as the primary outcomes of the parent study have not yet been published, but are available from the corresponding author on reasonable request.