Sleep quality in eating disorders: A systematic review and meta-analysis.

This systematic review and meta-analysis investigated several aspects of sleep in individuals diagnosed with eating disorders (EDs) compared to healthy controls (HCs). Specifically, the researchers aimed to:

1. **Compare sleep parameters and circadian preferences:** They looked at both objective (physiological) and subjective (self-reported) measures of sleep quality and an individual's natural preference for morning or evening activity.

2. **Assess changes in sleep quality and circadian preferences following ED treatment:** They explored whether specialized treatments for eating disorders had an impact on patients' sleep.

3. **Determine the prevalence of specific sleep disorders:** They examined if ED patients had a higher incidence of conditions like sleep apnea.

The eating disorders included in the analysis were Anorexia Nervosa (AN), Bulimia Nervosa (BN), and Binge Eating Disorder (BED).

The meta-analysis synthesized data from 27 individual studies. For the primary random effect analyses comparing sleep quality, the combined sample included:

**711 individuals with eating disorders (EDs)**

**653 healthy controls (HCs)**

The specific breakdown of ED subgroups (Anorexia Nervosa, Bulimia Nervosa, Binge Eating Disorder) contributing to the 711 ED patients was not detailed in the abstract, but subgroup analyses were performed. The introduction notes that eating disorders primarily affect "young female adolescents (male to female ratio between 0.3 and 0.6)," suggesting that the majority of participants across the included studies would likely fall into this demographic, although specific age ranges or gender distributions for the combined sample were not provided. Participants with EDs were diagnosed by clinicians based on standardized criteria (e.g., DSM-5), while healthy controls were defined as individuals with no self-reported diagnoses of psychological or neurological diseases. The studies were case-control designs, meaning they compared groups of individuals with a condition (ED) to those without (HC).

The researchers categorized sleep measurements into physiological (objective) and subjective (self-reported) outcomes:

**Physiological Sleep Measures:**

These were primarily collected using:

**Polysomnography (PSG):** Considered the "gold standard" for sleep measurement, PSG involves monitoring multiple physiological parameters during sleep in a lab setting. When both PSG and actigraphy data were available in a study, PSG was preferred.

**Actigraphy:** A non-invasive method using a wrist-worn device to estimate sleep-wake cycles based on movement.

From these objective measures, several specific sleep parameters were extracted and grouped:

**"Sleep continuity" measures:**

* **Sleep Efficiency Index (SE):** The percentage of time spent asleep while in bed (calculated as Total Sleep Time / Time in Bed x 100). A higher percentage indicates better sleep continuity.

* **Sleep Onset Latency (SOL):** The time (in minutes) it takes to fall asleep after getting into bed. Lower values are better.

* **Total Sleep Time (TST):** The total duration (in minutes) an individual spends asleep during the night. Higher values generally indicate more adequate sleep.

* **Wake After Sleep Onset (WASO):** The total time (in minutes) spent awake after initially falling asleep but before the final awakening. Lower values are better.

**"Sleep depth" measures:** These refer to the different stages of non-REM sleep, reported as a percentage of total sleep time:

* **Stage 1 (N1):** Lightest sleep stage.

* **Stage 2 (N2):** Deeper than N1, but still relatively light sleep.

* **Slow Wave Sleep (SWS):** Also known as N3 or deep sleep, crucial for physical restoration. Higher percentages are generally considered better.

**"REM pressure" measures:** These relate to Rapid Eye Movement sleep, important for cognitive function and emotional processing:

* **Rapid Eye Movement Sleep (REM):** The percentage of total sleep time spent in REM sleep.

* **REM Latency (REML):** The interval (in minutes) between falling asleep and the onset of the first REM sleep episode.

* **REM Density (REMD):** The frequency of rapid eye movements during REM sleep.

**Subjective Sleep Measures:**

These were collected through self-report instruments:

**Pittsburgh Sleep Quality Index (PSQI):** A widely used questionnaire assessing subjective sleep quality over the past month. It generates a global score (ranging from 0 to 21), where higher scores indicate poorer sleep quality.

**Morningness-Eveningness Questionnaire (MEQ) and MEQ reduced version (MEQ-r):** Questionnaires designed to assess an individual's circadian preference, classifying them as "morning types" (larks), "evening types" (owls), or "intermediate types." Higher scores typically indicate a morning preference.

**Sleep Diaries:** Daily self-reported logs where individuals record aspects of their sleep, such as bedtime, wake-up time, sleep onset latency, number of awakenings, and subjective sleep quality. When both questionnaires and diaries were available, diaries were preferred.

For sleep stages (N1, N2, N3, REM), only papers reporting data as percentages were included in the analysis.

This study was a **systematic review and meta-analysis**, which means the researchers systematically identified, evaluated, and synthesized findings from multiple existing studies on a specific topic. This approach allows for a more robust conclusion than any single study alone, as it pools data and increases statistical power.

**Study Design:**

The meta-analysis included **case-control studies**. In a case-control study, researchers identify a group of individuals with a particular condition (the "cases," in this instance, individuals with an eating disorder) and compare them to a similar group of individuals without the condition (the "controls," healthy individuals). They then look back in time to see if there are differences in exposures or characteristics (like sleep patterns) between the two groups.

**Search Strategy and Study Selection:**

**Databases:** The researchers conducted a comprehensive literature search in December 2023 across four major scientific databases: Web of Science, Pubmed, PsycINFO, and Medline. This broad search aimed to capture as many relevant studies as possible.

**Keywords:** A wide range of keywords related to eating disorders (e.g., "anorexia nervosa," "binge eating disorder") and sleep (e.g., "sleep," "insomnia," "PSQI," "REM") were used to ensure thoroughness.

**Inclusion Criteria (PICOs approach):**

* **Population:** Participants had a clinician-diagnosed eating disorder (AN, BN, or BED) and were compared to a healthy control group (no self-reported psychological or neurological diseases).

* **Intervention/Exposure:** The studies assessed subjective and/or physiological sleep parameters and/or circadian preference.

* **Comparison/Outcome:** Physiological sleep was measured by polysomnography (PSG) or actigraphy (PSG preferred if both available). Subjective data came from diaries or questionnaires (diaries preferred).

* **Study Design:** Only case-control studies were included.

* **Language:** Manuscripts in English, Italian, German, or Spanish.

**Exclusion Criteria:** Narrative reviews, other meta-analyses, case reports, and studies with overlapping samples (to avoid counting the same data multiple times).

**Screening Process:** Two independent authors (G.D. and D.M.) screened titles and abstracts, then full texts. Disagreements were resolved through discussion, and authors were contacted for missing data. This independent review process helps minimize bias in study selection.

**Risk of Bias (RoB) Assessment:**

The quality of the included studies was assessed using section A of the **Critical Appraisal Skills Programme (CASP) tool for case-control studies**. This tool helps evaluate the methodological rigor and potential for bias in individual studies, which is crucial for interpreting the overall meta-analysis results. A consensus procedure was used among the authors to resolve any disagreements in bias assessment.

**Statistical Approach:**

**Random Effect Analyses:** The meta-analysis used random effect models. This statistical approach assumes that the true effect size (e.g., the difference in sleep quality between ED patients and controls) might vary across studies due to differences in populations, interventions, or study designs. This is generally a more conservative and appropriate approach when combining studies that are likely to be heterogeneous.

**Subgroup Analyses:** To investigate potential differences among specific ED diagnoses, subgroup analyses were performed for Anorexia Nervosa (AN), Bulimia Nervosa (BN), and Binge Eating Disorder (BED). This helps to see if the overall findings apply equally to all types of eating disorders.

**Standardized Mean Difference (SMD):** While not explicitly stated in the abstract for the main results, meta-analyses typically use SMD to combine results from studies that measure the same outcome using different scales. SMD expresses the difference between groups in standard deviation units, allowing for comparison across studies.

**What this design can and cannot prove:**

**What it CAN prove:**

* **Stronger Evidence for Association:** By combining data from multiple studies, a meta-analysis provides a more precise estimate of the association between eating disorders and sleep quality than any single study. It can detect effects that might be missed in smaller individual studies.

* **Generalizability:** If the included studies are diverse enough, the findings of a meta-analysis can be more general