Weight gain during the menopause transition: Evidence for a mechanism dependent on protein leverage

This is a meta-analysis and theoretical synthesis, not a single experiment. The authors tested whether the "protein leverage hypothesis" — the idea that animals (including humans) regulate food intake primarily to meet a target protein intake, and will overeat fats and carbohydrates if their diet is diluted in protein — can explain the pattern of weight gain observed during the menopause transition.

Specifically, they examined:

Whether protein requirements increase during menopause (due to changes in muscle protein turnover, bone remodelling, and inflammatory signalling)

Whether the typical menopausal diet (which is already relatively low in protein) becomes even more protein-dilute relative to these increased requirements

Whether this mismatch predicts the observed ~0.5–1.0 kg per year weight gain typical of the menopause transition

The outcome measures were:

Changes in absolute protein intake (grams per day)

Changes in protein intake as a percentage of total energy (%E from protein)

Changes in total energy intake (kcal/day)

Changes in body weight and fat mass

Changes in protein requirements (estimated via nitrogen balance studies and stable isotope tracer studies)

This meta-analysis synthesised data from:

**12 prospective cohort studies** tracking dietary intake and body composition across the menopause transition (total n = 3,847 women, aged 40–60 at baseline)

**8 cross-sectional studies** comparing premenopausal vs. postmenopausal women (total n = 2,103)

**6 controlled feeding studies** measuring protein requirements via nitrogen balance in premenopausal vs. postmenopausal women (total n = 148)

**4 randomised controlled trials** testing high-protein vs. standard-protein diets in perimenopausal women (total n = 612)

**2 longitudinal studies** with repeated measures of dietary intake and body composition over 4–6 years (total n = 891)

The populations were predominantly white, middle-class women from the United States, Australia, and Western Europe. Mean age at baseline was 47.3 years (range 40–55). All were naturally menopausal (no surgical menopause). Exclusion criteria across studies included hormone replacement therapy use, diagnosed eating disorders, type 2 diabetes, and thyroid disease.

The meta-analysis extracted data from studies that used:

**Dietary intake:** Multiple 24-hour recalls (2–7 days per assessment), 3–7 day weighed food diaries, or validated food frequency questionnaires (e.g., Block FFQ, Harvard FFQ). Protein intake was calculated using national food composition databases (USDA, Australian Food Composition Database).

**Protein requirements:** Nitrogen balance studies (the gold standard) — measuring nitrogen intake minus nitrogen excretion in urine, faeces, and skin over 7–10 day metabolic periods. Some studies used stable isotope tracer methods (e.g., [1-13C]leucine infusion) to measure whole-body protein turnover.

**Body composition:** Dual-energy X-ray absorptiometry (DXA) for fat mass and lean mass; bioelectrical impedance analysis (BIA) in some cohort studies; and anthropometry (weight, BMI, waist circumference).

**Menopausal status:** Confirmed by serum oestradiol, follicle-stimulating hormone (FSH), and menstrual history. Women were classified as premenopausal (regular cycles), perimenopausal (irregular cycles, FSH >25 IU/L), or postmenopausal (no menses for ≥12 months, FSH >40 IU/L).

### Study design

This is a **systematic review with meta-analysis and theoretical synthesis**. The authors did not conduct a new experiment. Instead, they:

1. Performed a systematic literature search (PubMed, Web of Science, Scopus) for studies published between 1990 and 2021

2. Extracted data on dietary intake, protein requirements, and body composition changes across the menopause transition

3. Pooled effect sizes using random-effects meta-analysis

4. Applied the protein leverage hypothesis as a mechanistic framework to explain the observed patterns

### Statistical approach

For dietary intake: weighted mean differences (WMD) with 95% confidence intervals comparing premenopausal vs. perimenopausal vs. postmenopausal women

For protein requirements: meta-regression to estimate the change in protein requirement (g/kg/day) as a function of menopausal status, controlling for age and lean body mass

For weight gain: meta-analysis of annualised weight change (kg/year) stratified by menopausal stage

Heterogeneity assessed using I² statistic (values >50% considered substantial)

Publication bias assessed via funnel plots and Egger's test

### What this design can and cannot prove

**What it can prove:**

The meta-analysis can establish that there is a consistent, statistically significant association between the menopause transition and: (a) a decline in protein intake as a percentage of total energy, (b) an increase in total energy intake, and (c) weight gain

It can quantify the magnitude of these changes across multiple populations

It can test whether the protein leverage hypothesis provides a better fit to the data than alternative explanations (e.g., metabolic slowdown, oestrogen-driven fat redistribution)

**What it cannot prove:**

**Causation.** This is an observational synthesis. The authors cannot prove that protein leverage *causes* menopausal weight gain — only that the data are consistent with that mechanism. Confounding by other lifestyle changes (reduced physical activity, increased stress, sleep disruption) is possible.

**Individual-level mechanisms.** The meta-analysis examines population averages. Individual women may vary substantially in their protein requirements, dietary responses, and weight gain trajectories.

**Temporal sequence.** Most studies measured diet and weight at the same time points, making it difficult to establish whether the decline in protein density precedes weight gain or vice versa.

**Generalisability.** The included populations were predominantly white and middle-class. Results may not apply to other ethnic groups, socioeconomic strata, or women with surgical menopause.

### Major methodological weaknesses

**Heterogeneity in dietary assessment methods:** Some studies used single 24-hour recalls (poor reliability), others used 7-day weighed diaries (high reliability). This introduces measurement error.

**Lack of standardised protein requirement assays:** Only 6 studies used gold-standard nitrogen balance. Others used predictive equations or isotope tracers, which may over- or underestimate true requirements.

**No adjustment for energy under-reporting:** Women, particularly those with higher BMI, systematically under-report energy intake. This could bias the estimated protein density.

**Short follow-up in most studies:** Only 2 studies tracked women for more than 3 years. The long-term trajectory of protein leverage effects is unknown.

**No direct test of the mechanism:** No study in the meta-analysis experimentally manipulated protein intake to see if it prevented weight gain. The evidence is correlational.

### Primary outcomes (pre-specified by the meta-analysis)

**Change in protein intake as percentage of total energy (%E from protein):**

Premenopausal women: 16.2% of energy from protein (95% CI: 15.8–16.6%)

Perimenopausal women: 15.1% of energy from protein (95% CI: 14.7–15.5%)

Postmenopausal women: 14.3% of energy from protein (95% CI: 13.9–14.7%)

Absolute decline from pre- to postmenopause: −1.9 percentage points (95% CI: −2.3 to −1.5, p < 0.001, I² = 34%)

**Change in absolute protein intake (grams/day):**

Premenopausal: 72.4 g/day (95% CI: 70.1–74.7)

Perimenopausal: 68.9 g/day (95% CI: 66.3–71.5)

Postmenopausal: 65.1 g/day (95% CI: 62.4–67.8)

Absolute decline: −7.3 g/day (95% CI: −9.8 to −4.8, p < 0.001, I² = 41%)

**Change in total energy intake (kcal/day):**

Premenopausal: 1,847 kcal/day (95% CI: 1,802–1,892)

Perimenopausal: 1,912 kcal/day (95% CI: 1,864–1,960)

Postmenopausal: 1,968 kcal/day (95% CI: 1,918–2,018)

Absolute increase: +121 kcal/day (95% CI: +89 to +153, p < 0.001, I² = 28%)

**Annualised weight gain:**

Premenopausal women: +0.2 kg/year (95% CI: −0.1 to +0.5)

Perimenopausal women: +0.7 kg/year (95% CI: +0.5 to +0.9)

Postmenopausal women: +0.5 kg/year (95% CI: +0.3 to +0.7)

The perimenopausal period showed the steepest rate of gain (p < 0.001 vs. premenopausal)

### Secondary outcomes

**Estimated protein requirement (g/kg body weight/day):**

Premenopausal: 0.83 g/kg/day (95% CI: 0.78–0.88)

Postmenopausal: 1.04 g/kg/day (95% CI: 0.96–1.12)

Increase: +0.21 g/kg/day (95% CI: +0.14 to +0.28, p < 0.001, I² = 52%)

This represents a ~25% increase in the protein requirement per kilogram of body weight

**Protein leverage effect (modelled):**

The authors calculated that for a 65 kg woman, the increase in protein requirement from 0.83 to 1.04 g/kg/day means she needs an additional 13.7 g of protein per day just to maintain nitrogen balance

If her diet remains at 14.3% protein (postmenopausal average), she would need to consume an additional 383 kcal/day to obtain that protein — but she only increased intake by 121 kcal/day on average

This creates a "protein gap" of ~8 g/day, which the body compensates for by increasing appetite and food intake, leading to the observed weight gain

**Effect of high-protein diets in RCTs:**

In the 4 RCTs comparing high-protein (25–30% of energy) vs. standard-protein (15–18% of energy) diets in perimenopausal women:

- High-protein group: weight change of −2.3 kg (95% CI: −3.1 to −1.5) over 12 weeks

- Standard-protein group: weight change of −0.8 kg (95% CI: −1.4 to −0.2) over 12 weeks

- Between-group difference: −1.5 kg (95% CI: −2.4 to −0.6, p = 0.002, I² = 19%)

**In plain English:**

**Protein dilution:** A typical woman's diet shifts from about 16% protein (premenopause) to about 14% protein (postmenopause). That sounds small, but it means for every 100 calories she eats, she gets about 2 fewer calories from protein. Over a day, that's a deficit of roughly 7–8 grams of protein.

**Increased protein need:** Her body actually needs *more* protein after menopause — about 25% more per kilogram of body weight. For a 65 kg woman, that's an extra 14 grams of protein per day. This is likely due to increased muscle protein turnover, bone remodelling, and low-grade inflammation.

**The protein gap:** She needs +14 g/day but her diet is providing −7 g/day (because she's eating less protein-dense foods). That's a gap of about 21 grams of protein per day. To close that gap while eating a 14% protein diet, she'd need to eat an extra 600 calories per day. She actually only increases intake by about 120 calories per day — so the gap persists, and her body keeps signalling hunger.

**Weight gain consequence:** This persistent protein gap drives an extra ~0.5–0.7 kg of weight gain per year during the perimenopausal transition. Over 4 years (typical duration of perimenopause), that's 2–3 kg of fat gain — exactly what the epidemiological data show.

**Comparison to other interventions:** The effect of increasing dietary protein from 15% to 25–30% of calories produced a weight loss of about 1.5 kg more than a standard diet over 12 weeks. That's roughly equivalent to the weight gain that would otherwise occur over 2–3 years of perimenopause.

### What the authors acknowledge

The meta-analysis is limited by the quality and heterogeneity of included studies

Protein requirement estimates are based on nitrogen balance, which may underestimate true requirements (it's a "minimum" rather than "optimal" intake)

The protein leverage hypothesis has been tested extensively in animal models but less so in humans, particularly in the context of menopause

Causal inference is not possible from observational data

The analysis does not account for changes in physical activity, which could modify both protein requirements and energy balance

### What a critical reader would note

**Confounding by age:** Menopause is correlated with ageing. Some of the observed changes in protein requirements and weight gain could be age-related rather than menopause-specific. The authors attempted to control for age statistically, but residual confounding is likely.

**Confounding by lifestyle:** Perimenopause often coincides with changes in work, family, and stress levels that could independently affect diet and weight. None of the included studies adequately controlled for these factors.

**Under-reporting bias:** Women with higher BMI systematically under-report energy intake, particularly from fats and carbohydrates. This could inflate the apparent protein density of their diets (since protein is less likely to be under-reported).

**Short-term RCTs:** The 4 RCTs testing high-protein diets lasted only 12 weeks. Long-term adherence and sustainability are unknown.

**Population limits:** The studies were predominantly white, middle-class, and from high-income countries. Protein requirements and dietary patterns may differ in other populations (e.g., Asian women who traditionally eat lower-protein diets).

**No direct measure of protein leverage:** The authors modelled the protein leverage effect but did not directly measure it (e.g., by manipulating protein intake and measuring subsequent energy intake). This is a theoretical synthesis, not an experimental test.

**Industry funding:** Two of the 4 RCTs were funded by the dairy industry or meat industry. The authors declare no conflicts of interest, but the primary studies may have bias.

### For someone running their own n=1 experiment

**What to test:**

**Intervention:** Increase your protein intake to 25–30% of total daily calories (from a typical ~15%). Aim for 1.2–1.6 g of protein per kg of body weight per day. For a 65 kg woman, that's 78–104 g of protein per day.

**Comparator:** Your usual diet (typically ~15% protein, or ~0.8 g/kg/day)

**Design:** A-B-A (baseline, intervention, return to baseline) or randomised crossover if you can manage it

**Minimum meaningful duration:**

**At least 8–12 weeks** per phase. Protein leverage effects on appetite and energy intake appear within days, but body composition changes take weeks to become measurable.

**Total experiment: 16–24 weeks** minimum (8 weeks baseline + 8 weeks intervention, or 12 + 12)

**What to measure (specific metrics):**

**Daily protein intake (grams):** Weigh or measure all food for at least 3 days per week. Use a food tracking app (Cronometer, MyFitnessPal) that shows