Oxytocin and its role in caloric intake and appetite: A preregistered living systematic review and meta-analysis

This preregistered living systematic review and meta-analysis provides evidence that exogenous oxytocin reduces appetite and caloric intake in specific patient groups, such as those with obesity or type II diabetes, but finds no statistically significant effects in healthy individuals, suggesting the need for further research to draw definitive conclusions.

The Gist

Imagine your body has a tiny, internal thermostat that constantly checks how much fuel (food) you have and how much you're burning. This thermostat is called energy balance. When you have too much fuel, the thermostat should tell you to stop eating. When you're running on empty, it tells you to find food.

For a long time, scientists have been fascinated by a chemical messenger in our brains called oxytocin. You might know it as the "love hormone" because it helps with childbirth, breastfeeding, and making us feel connected to others. But recently, scientists started wondering: Could oxytocin also be a key player in turning off that "I'm hungry" signal?

The problem is that previous studies gave mixed answers. Some said oxytocin stops hunger; others said it does nothing. It was like trying to assemble a puzzle with pieces from different boxes.

So, a team of researchers decided to build a "Living Systematic Review." Think of this not as a static report, but as a living, breathing library. Instead of just looking at the books available today, they plan to keep adding new books as they are published over the next few years, constantly updating their conclusions.

Here is what they found, broken down into simple concepts:

1. The "Healthy" vs. "Clinical" Difference

The researchers split the data into two main groups: Healthy People and People with Specific Health Conditions (like obesity, Type 2 diabetes, or eating disorders).

• The Healthy Group: When they gave oxytocin to healthy people with normal weight, it was like trying to stop a car that wasn't moving. The oxytocin didn't really change their appetite or how much they ate. The body's thermostat was already working fine, so the extra oxytocin didn't seem to make a difference.
• The Clinical Group: This is where things got interesting. For people with conditions like obesity or Type 2 diabetes (where the body's fuel system is often stuck in "overdrive"), oxytocin acted like a brake pedal. It significantly reduced their appetite and made them feel fuller.

2. The "Goldilocks" Dose

The researchers also looked at how much oxytocin was used. They found that the dose matters, but it's complicated.

• In people with obesity, a specific dose (24 International Units) seemed to work best to lower calorie intake.
• However, if the dose was too high, the effect sometimes got weaker or even flipped. It's like seasoning a soup: a pinch makes it delicious, but a whole shaker ruins it.

3. Why the Results Were "Fuzzy"

You might wonder, "If it works for some, why didn't it work for everyone?"
The researchers found that many of the original studies were underpowered. Imagine trying to hear a whisper in a noisy stadium with a tiny microphone. The studies were too small to hear the "whisper" of the effect clearly.

• The "Inconclusive" Finding: For healthy people, the data wasn't just "no effect"; it was "we can't tell yet." The data was too insensitive to say for sure that oxytocin does nothing, but it certainly didn't prove it does something.
• The "Noise": The studies varied wildly. Some measured hunger right after a meal, some measured it hours later. Some used nasal sprays, others used IVs. It was like trying to compare apples, oranges, and bananas to figure out which fruit is the sweetest.

4. The Big Picture: A Context-Sensitive Thermostat

The main takeaway is that oxytocin isn't a magic "stop-eating" pill for everyone. Instead, it seems to be a context-sensitive helper.

Think of oxytocin as a smart home assistant.

• If your house (body) is already at a comfortable temperature (healthy weight), the assistant doesn't need to do much.
• But if your house is overheating (obesity or metabolic issues), the assistant steps in and turns down the heat (reduces appetite) to bring things back to a comfortable balance.

What's Next?

Because this is a "Living" review, the story isn't over. The researchers plan to keep adding new studies to their library. As more data comes in, they hope to:

1. Confirm exactly who benefits from oxytocin treatment.
2. Figure out the perfect dose for different people.
3. Understand why it works for some conditions (like obesity) but not others (like Prader-Willi syndrome, a genetic condition where people feel constantly hungry).

In short: Oxytocin shows promise as a tool to help people with specific metabolic struggles regain control over their appetite, acting like a reset button for a broken thermostat. But for healthy people, it's not a magic bullet, and we need bigger, better studies to know for sure how it works.

Technical Summary

1. Problem Statement

Oxytocin, a hypothalamic neuropeptide, is well-established for its roles in parturition and social behavior, but its involvement in energy balance, metabolism, and appetite regulation remains inconsistent in human literature. While animal models suggest oxytocin reduces food intake (hyperphagia reduction) and body weight, human studies have yielded conflicting results regarding its efficacy on caloric intake and appetite across different populations (healthy vs. clinical). Previous systematic reviews suffered from methodological limitations, including the exclusion of psychiatric populations, failure to account for effect size dependency (multiple outcomes from single studies), lack of publication bias assessment, and outdated data. Furthermore, existing evidence often fails to distinguish between healthy individuals and those with metabolic or eating disorders, obscuring potential context-dependent effects.

2. Methodology

The authors conducted a preregistered, living systematic review and meta-analysis following PRISMA-LSR guidelines. The "living" aspect implies the analysis is designed to be updated regularly (planned twice over four years) to incorporate new evidence.

• Search Strategy: Databases searched included PubMed/Medline, Embase, APA PsycInfo, Cochrane Library, Scopus, and Web of Science (search dates: April 2025).
• Inclusion Criteria: Randomized placebo-controlled trials (RCTs) involving exogenous oxytocin administration (intranasal, IV, oral, or IM) in humans. Outcomes had to relate to caloric intake or appetite (hunger, satiety, fullness, cravings).
• Exclusion Criteria: Animal studies, endogenous oxytocin studies, non-RCT designs, and non-English publications.
• Data Extraction & Management:
  • Initial screening identified 21 studies (691 participants) for the final analysis after exclusions due to data unavailability or irrelevance.
  • Data extraction was performed by the first author, with validation by a co-author.
  • Study Groups: Analyses were stratified into four distinct meta-analyses (MAs):
    1. Caloric intake in healthy participants (MA1).
    2. Caloric intake in clinical participants (MA2).
    3. Appetite in healthy participants (MA3).
    4. Appetite in clinical participants (MA4).
• Statistical Analysis:
  • Effect Size: Hedges' g was calculated (correcting for small sample bias). For within-participant designs, Standardized Mean Change with Change Score Standardization (SMCC) was used.
  • Model: A frequentist three-level random-effects meta-analysis (using metafor in R) accounted for dependency when multiple effect sizes were extracted from a single study.
  • Bayesian Approach: Robust hierarchical Bayesian model-averaged meta-analysis (hBMA) was used to quantify evidence for null vs. alternative hypotheses.
  • Bias & Power: Publication bias was assessed via Egger's regression and contour-enhanced funnel plots. Equivalence testing (TOST) was performed to determine if non-significant results were due to true equivalence or low power.
  • Moderators: Pre-specified and post-hoc moderators included sex, age, dosage, administration route, frequency, and clinical condition type.
  • Significance Threshold: A Bonferroni-adjusted $\alpha$ of 0.017 was applied to correct for multiple comparisons.

3. Key Contributions

• Living Meta-Analysis Framework: This is one of the first "living" meta-analyses in the field of oxytocin research, designed to evolve with new data to prevent rapid obsolescence.
• Granular Stratification: Unlike previous reviews, this study rigorously separates healthy and clinical populations and distinguishes between caloric intake (objective behavior) and appetite (subjective sensation).
• Methodological Rigor: The study addresses common pitfalls in oxytocin research by:
  • Using three-level meta-analyses to handle dependent effect sizes.
  • Applying equivalence testing to distinguish between "no effect" and "insensitive data."
  • Conducting robust Bayesian analyses alongside frequentist approaches.
  • Explicitly testing for publication bias and small-study effects.
• Moderator Analysis: The study provides detailed insights into how clinical condition, dosage, and sex influence oxytocin's efficacy, moving beyond a "one-size-fits-all" conclusion.

4. Results

The analysis included 21 studies (691 participants) yielding 63 effect sizes.

• Appetite in Clinical Populations (MA4):
  • Significant Finding: Oxytocin administration significantly reduced appetite in clinical populations (Hedges' g = -0.208, p = 0.008).
  • Bayesian Support: The hBMA yielded a Bayes Factor of 1.823, indicating anecdotal evidence for the alternative hypothesis.
  • Subgroups: The effect was driven by populations with obesity, Type II Diabetes (T2D), schizophrenia (SCZ), and bulimia/binge eating disorder (BED).
• Caloric Intake in Clinical Populations (MA2):
  • Result: No statistically significant pooled effect (Hedges' g = -0.049, p = 0.607).
  • Equivalence: Equivalence testing showed the effect was statistically equivalent to zero under relaxed bounds (±0.23), suggesting the data is currently too insensitive to draw definitive conclusions on intake reduction in this broad group, though a trend exists.
• Healthy Populations (MA1 & MA3):
  • Caloric Intake (MA1): Non-significant effect (Hedges' g = -0.225, p = 0.036, but failed to meet the strict $\alpha$ = 0.017 threshold).
  • Appetite (MA3): Non-significant effect (Hedges' g = 0.074, p = 0.492).
  • Equivalence: Equivalence tests were non-significant for both, indicating the data is inconclusive rather than proving a lack of effect (likely due to low statistical power).
• Moderator Analyses:
  • Condition Type: In clinical caloric intake (MA2), the specific diagnosis significantly moderated the effect. The obesity group showed a significantly more negative effect (reduced intake) compared to the reference group (Prader-Willi Syndrome).
  • Dosage: In MA2, a 24 IU dose was associated with a significant reduction in caloric intake, whereas doses >24 IU were associated with a less negative (or positive) effect.
  • Sex: In MA2, a higher percentage of female participants was associated with less anorexigenic (appetite-reducing) effects. However, the authors note this may be a proxy for the underlying condition (e.g., obesity groups were often male-only in the dataset, while eating disorder groups were female-only).
• Bias and Heterogeneity:
  • MA3 (Appetite in healthy) showed significant funnel plot asymmetry and publication bias.
  • High heterogeneity was observed, primarily driven by within-study variance in MA1/MA2 and between-study variance in MA3.
  • Power: All meta-analyses were severely underpowered (ranging from ~5% to 17% power to detect observed effects), limiting the ability to rule out small but meaningful effects.

5. Significance and Conclusion

The study concludes that exogenous oxytocin acts as a conditional, context-sensitive anorexigenic agent.

• Clinical Relevance: Oxytocin shows promise in reducing appetite and potentially caloric intake in clinical populations characterized by excessive body weight and caloric consumption (e.g., obesity, T2D, SCZ, BED). It does not appear to have a pronounced effect in healthy individuals or those with conditions characterized by low intake (e.g., anorexia nervosa, Prader-Willi Syndrome).
• Mechanism: The authors propose a homeostatic mechanism where oxytocin facilitates a return to energy balance in states of positive energy surplus but is less effective or ineffective in states of negative energy balance or genetic dysregulation (like PWS).
• Future Directions: The authors emphasize the need for:
  • Larger, well-powered studies to resolve inconclusive findings in healthy populations.
  • Standardized protocols regarding meal timing, dosage (specifically the 24 IU threshold), and formulation.
  • Reporting of body composition (BMI, lean mass) to better understand moderators.
  • Continued updates to this "living" meta-analysis to incorporate new data on biomarkers and body composition.

In summary, while oxytocin is not a universal appetite suppressant, it represents a potential therapeutic target for specific metabolic and psychiatric conditions involving hyperphagia and obesity, provided future research addresses current methodological limitations regarding power and standardization.