June 2025 is the month of thresholds and architecture. Three macronutrient meta-analyses reveal that the internal composition of a dietary pattern determines outcomes more than the pattern label. Nuts lower lipids — but type matters enormously (113 RCTs). Ketogenic diets improve body composition — but only below 50g/day carbohydrate (33 RCTs). And high-protein diets work best when the fat-to-carb ratio is tuned to the clinical target (83 RCTs).
The gut-brain evidence is equally precise. Probiotics improve cognition more than mood (SMD 0.48 vs 0.29), but only at sufficient dose and duration. A psilocybin meta-analysis tempers the psychedelic hype — effects attenuate in better-controlled trials. And a natural extract NMA reveals that combination formulations outperform single compounds, with an unexpected Cistanche + Ginkgo pairing topping the executive function rankings.
Two inflammation studies complete the picture by identifying complementary pathways: dietary weight loss reduces IL-6 (above 5%) but not TNF-α, while mind-body practices reduce TNF-α and cortisol but not IL-6. Prescribed together, they would cover a broader inflammatory profile than either alone. Eight studies, three themes, one message: architecture and thresholds matter more than categories.
Studies at a Glance
Macronutrient Architecture
Three meta-analyses this month move beyond the tired "which diet is best" debate and into the finer question of macronutrient architecture — how the internal composition of a dietary pattern shapes outcomes. Nuts lower lipids across 113 trials, but type matters enormously. Ketogenic diets improve body composition, but only below 50g/day carbohydrate. And high-protein diets work best when the accompanying fat-to-carb ratio is tuned to the target outcome.
Nut consumption has been recommended for cardiovascular health for decades, but the evidence has been scattered across small trials with different nut types and populations. This updated meta-analysis pooled 113 RCTs — the largest lipid-focused nut synthesis to date — to quantify effects on seven lipid markers and identify which nut types drive the strongest improvements.
Study Design & Protocol
Design: Systematic review and meta-analysis of RCTs
Included trials: 113 RCTs, 8,060 adults
Median dose: 45.5 g/day
Outcomes: TC, LDL-C, HDL-C, triglycerides, TC:HDL ratio, LDL:HDL ratio, ApoB
Subgroup analysis: By nut type (almond, walnut, pistachio, pecan, mixed)
Nut consumption at ~45g/day significantly reduced total cholesterol, LDL-C, triglycerides, lipid ratios, and apolipoprotein B. HDL-C was unaffected. Almonds, pecans, pistachios, and walnuts showed the strongest effects, while mixed nut interventions and other single types were less consistent. The ApoB reduction is particularly clinically relevant as a marker of atherogenic particle number.
Strengths
- 113 RCTs — largest nut-lipid meta-analysis to date
- Seven lipid markers assessed including ApoB
- Subgroup analysis by nut type provides actionable specificity
- Includes dose-response analysis
Limitations
- Heterogeneous control diets across trials
- Cannot determine independent effect of nut matrix vs nutrient profile
- Publication bias possible with positive trials
- Median dose (45g) exceeds typical intake
Practitioners can confidently recommend almonds, walnuts, pistachios, or pecans (~45g/day) for lipid management. The ApoB reduction adds cardiovascular risk prediction value beyond standard lipid panels. For clients resistant to medication, nut-based dietary strategies represent a well-evidenced first-line approach with the benefit of sustained adherence potential.
The carbohydrate restriction spectrum ranges from mild (100g/day) to ketogenic (<50g/day), yet most meta-analyses treat "low-carb" as a single category. This review stratified by carbohydrate threshold to determine whether the degree of restriction determines the breadth of body composition improvement — specifically whether fat mass reduction requires the ketogenic threshold.
Study Design & Protocol
Design: Systematic review and meta-analysis of RCTs
Included trials: 33 RCTs, 2,821 participants with overweight/obesity
Threshold analysis: ≤100g/day vs ≤50g/day carbohydrate
Outcomes: Body weight, BMI, fat mass, body fat percentage
Duration subgroups: <1 month vs ≥1 month
At the ≤100g/day threshold, weight, BMI, and body fat percentage improved but fat mass did not. Only when carbohydrate intake dropped to ≤50g/day (ketogenic range) did all four body composition measures improve significantly — including fat mass. Interventions needed to be at least 1 month to show fat mass effects. This identifies a clear carbohydrate dose-response for body composition.
Strengths
- Threshold-based analysis — clinically actionable
- 33 RCTs with overweight/obese populations
- Separates fat mass from body weight and body fat %
- Duration subgroup analysis adds temporal guidance
Limitations
- Carbohydrate intake often self-reported — adherence uncertain
- Cannot control for spontaneous calorie reduction on keto
- Short-term RCTs predominate — long-term data sparse
- Does not assess lean mass preservation
For practitioners prescribing carbohydrate restriction, the 50g/day threshold appears to be the line between partial and comprehensive body composition improvement. Clients seeking fat mass reduction (not just weight loss) likely need to achieve and maintain ketosis for at least one month. Those unwilling to go that low can still expect weight and BMI improvements at 100g/day.
High-protein diets are well-established for weight management, but the question of what fills the remaining calories — fat or carbohydrate — has received less attention. This network meta-analysis ranked different high-protein macronutrient combinations against each other and standard diets to determine which configuration optimises which outcome.
Study Design & Protocol
Design: Systematic review with pairwise and network meta-analysis of RCTs
Included trials: 83 RCTs
Comparisons: High-protein + moderate-carb + high-fat vs high-protein + low-carb + high-fat vs high-protein + high-carb + low-fat vs standard
Outcomes: Body mass, BMI, waist circumference, fat mass, lean mass, lipids
High-protein diets significantly improved body composition across all measures. The NMA ranked high-protein + moderate-carb + high-fat as best for weight and fat reduction, while high-protein + low-carb + high-fat maximised triglyceride reduction and HDL improvements. Lean body mass was preserved (+0.34 kg) regardless of carb-to-fat ratio. The message: protein is the constant; the carb-fat balance should be tuned to the clinical target.
Strengths
- 83 RCTs — large evidence base
- NMA allows head-to-head ranking of macronutrient combinations
- Assesses body composition AND cardiometabolic markers
- Lean mass preservation data clinically valuable
Limitations
- Protein threshold definitions vary across studies
- Cannot account for food quality within macronutrient categories
- Short-to-medium term RCTs predominate
- Adherence data limited across included trials
This enables genuinely personalised macronutrient prescription. For weight/fat loss priority: high protein with moderate carb and higher fat. For lipid optimisation: high protein with low carb and higher fat. For lean mass preservation: protein adequacy matters more than the carb-fat ratio. Practitioners can now tailor the "what else goes on the plate" conversation around specific client goals.
Mind, Mood, and the Microbiome
Three studies explore the intersection of gut ecology, plant compounds, and mental health. A 54-trial meta-analysis confirms that probiotics genuinely improve anxiety, depression, and cognition — with dose and duration thresholds now identified. A network meta-analysis ranks 19 natural extracts for cognitive function, with an unexpected combination topping the list. And a sobering psilocybin meta-analysis tempers the psychedelic therapy hype with a dose of methodological reality.
The gut-brain axis has moved from theoretical framework to clinical evidence. This meta-analysis pooled up to 54 RCTs to quantify the effects of pro- and prebiotic supplementation on three distinct mental health outcomes — anxiety, depression, and cognitive function — while identifying the dose and duration thresholds that separate effective from ineffective interventions.
Study Design & Protocol
Design: Systematic review and meta-analysis of RCTs (PRISMA)
Anxiety analysis: 54 RCTs, 4,295 participants
Depression analysis: 40 RCTs, 3,179 participants
Cognition analysis: 13 RCTs, 915 participants
Subgroups: Duration (6+ weeks), dose (10⁹+ CFU/day), clinical status
Pro- and prebiotics significantly improved all three outcomes, with the largest effect on cognitive function (SMD 0.48). Greater benefits emerged with longer interventions (6+ weeks), higher doses (≥10⁹ CFU/day), and in clinically symptomatic populations. The cognitive improvement effect size is particularly noteworthy — larger than the mood effects — suggesting the gut-brain connection may be strongest for cognitive rather than affective outcomes.
Strengths
- Up to 54 RCTs — largest gut-brain mental health meta-analysis
- Three distinct outcomes assessed independently
- Dose and duration thresholds identified
- Clinical vs healthy population subgroup analysis
Limitations
- Heterogeneous probiotic strains and formulations
- Cannot identify which strains drive each effect
- Self-reported outcome measures in many trials
- Publication bias possible
This provides prescriptive clarity: probiotics at ≥10⁹ CFU/day for ≥6 weeks, preferably in clients with existing symptoms. The cognitive effect being larger than the mood effect is clinically interesting — practitioners working with cognitive decline may find gut-targeted strategies more productive than previously assumed. The dose threshold also helps avoid the common clinical error of underdosing probiotic interventions.
Dozens of natural extracts claim cognitive benefits, but no study had simultaneously compared them all. This network meta-analysis ranked 19 distinct natural extract interventions across 27 RCTs, using SUCRA probability ranking to determine which compounds perform best for each cognitive domain — global cognition, executive function, memory, and attention.
Study Design & Protocol
Design: Systematic review and network meta-analysis of RCTs
Included trials: 27 RCTs, 2,334 healthy adults
Interventions: 19 distinct natural extracts (single and combination)
Cognitive domains: Global cognition, executive function, cognitive flexibility, memory, attention
Ranking: Surface Under Cumulative Ranking Curve (SUCRA)
The Cistanche + Ginkgo combination ranked first for executive function (SUCRA 96.9%), cognitive flexibility, and memory (SUCRA 89.3%). Polygala tenuifolia (yuan zhi) ranked first for global cognition (SUCRA 95.9%, SMD = 1.28). No single extract significantly improved attention — a notable gap in the nootropic evidence base. The combination approach (Cistanche + Ginkgo) outperformed any single extract across multiple domains.
Strengths
- 19 interventions compared simultaneously via NMA
- SUCRA ranking provides clinically actionable hierarchy
- Assesses multiple cognitive domains separately
- Healthy adult population — relevant for preventive use
Limitations
- Small trial sizes limit statistical power
- Heterogeneous extract preparations and doses
- Limited replication for combination formulations
- Cannot determine optimal dosing from available data
For practitioners advising on cognitive support, the Cistanche + Ginkgo combination deserves attention — though replication is needed. Polygala tenuifolia (yuan zhi) is an underexplored option with strong SUCRA ranking for global cognition. The null attention finding is equally useful: no natural extract has reliable evidence for attention enhancement, which helps manage client expectations around "focus" supplements.
Psilocybin-assisted therapy has generated enormous enthusiasm for treatment-resistant depression, with early trials reporting dramatic response rates. But as the evidence base grows, methodological scrutiny reveals a more nuanced picture. This meta-analysis specifically examined whether psilocybin's apparent superiority holds up in larger, better-controlled studies — an "incremental efficacy" framework that tests whether initial promise survives methodological rigour.
Study Design & Protocol
Design: Systematic review and meta-analysis of RCTs
Included trials: 9 RCTs (10 subgroups), 602 participants
Population: Adults with MDD or treatment-resistant depression
Outcome: Depression symptom reduction vs controls
Quality assessment: Risk of bias, harm-reporting quality
Psilocybin showed a moderate effect over controls (Hedges' g = 0.62), but effects attenuated in larger and better-controlled studies. Five of nine trials had low or very low harm-reporting quality, and most demonstrated high risk of bias. The authors note that even modest incremental effects could be meaningful if a single dose produces durable reduction compared to daily medication — but the field needs better trials before clinical adoption.
Strengths
- "Incremental efficacy" framework — tests whether promise survives rigour
- Includes quality-of-evidence assessment
- Transparent about effect attenuation with better methods
- Evaluates harm-reporting quality separately
Limitations
- Only 9 RCTs — still a small evidence base
- Blinding challenges inherent to psychedelic trials
- Cannot assess long-term durability from available data
- Heterogeneous dosing protocols and therapeutic support
Psilocybin shows real promise but is not yet the revolution early media coverage suggested. The effect attenuation in better-controlled trials is a methodological red flag that demands caution. Practitioners should follow the evidence but resist premature clinical adoption. The harm-reporting gap is particularly concerning — we don't yet have reliable safety data from the majority of trials.
The Anti-Inflammatory Threshold
Two studies identify specific thresholds for anti-inflammatory benefit. Weight loss reduces IL-6 — but only when sustained above 5%. And mind-body practices reduce CRP and TNF-α across 89 RCTs, confirming that yoga, meditation, and tai chi are legitimate immunomodulatory interventions, not just stress relief.
Weight loss is prescribed for inflammation reduction, but how much weight loss is enough? And which inflammatory markers respond? This meta-analysis focused on dietary weight-loss interventions with a minimum 12-month follow-up — a duration threshold that separates sustained from transient effects — and assessed IL-6 and TNF-α separately.
Study Design & Protocol
Design: Systematic review and meta-analysis of RCTs
Included trials: 12 RCTs (minimum 12-month follow-up)
TNF-α analysis: 3 studies, 629 participants
IL-6 analysis: Studies with >5% weight loss achieved
Key threshold: 5% body weight loss sustained
Only studies achieving and maintaining greater than 5% weight loss significantly reduced IL-6. TNF-α showed no response to dietary weight loss in any trial — changes ranged from −21% to +25% without between-group differences. This identifies a clear threshold: modest weight loss (2–4%) may improve metabolic markers but does not reduce systemic inflammation. The TNF-α null finding suggests this cytokine is regulated by different mechanisms than IL-6 in the context of obesity.
Strengths
- 12-month minimum follow-up — captures sustained effects only
- Separates IL-6 from TNF-α — different response profiles
- Identifies clinically actionable 5% threshold
- All dietary interventions — no confounding by medication
Limitations
- Only 12 eligible RCTs — small evidence pool
- TNF-α analysed in only 3 studies
- Cannot determine if exercise-based weight loss differs
- Threshold is correlational — not experimentally tested
The 5% threshold is clinically actionable: practitioners can set evidence-based weight-loss targets for anti-inflammatory benefit, rather than arbitrary goals. Clients who achieve 3–4% weight loss should be counselled that metabolic improvements may occur, but systemic inflammation (IL-6) requires crossing the 5% line. The TNF-α null finding also guides expectations — not all inflammatory markers respond to weight loss equally.
Mind-body practices — yoga, meditation, tai chi, qigong, mindfulness — are often recommended for stress, but their effects on objectively measured immune and neuroendocrine biomarkers have been less clear. This meta-analysis of 89 RCTs (screened from 1,697) provides the most comprehensive biomarker assessment to date, measuring CRP, TNF-α, IL-6, cortisol, BDNF, and sIgA.
Study Design & Protocol
Design: Systematic review and meta-analysis of RCTs
Included trials: 89 RCTs (from 1,697 screened)
Interventions: Yoga, meditation, tai chi, qigong, mindfulness
Biomarkers: CRP, TNF-α, IL-6, cortisol, BDNF, sIgA
Mind-body interventions significantly reduced CRP (SMD −0.13), TNF-α (SMD −0.37), and cortisol (SMD −0.33). IL-6 showed a reduction trend (SMD −0.24) but did not reach significance. BDNF and sIgA trended upward but also remained non-significant. The TNF-α and cortisol effects are noteworthy — particularly since the weight loss meta-analysis (Study 7) found TNF-α unresponsive to dietary intervention, suggesting mind-body practices access inflammatory pathways that diet alone cannot.
Strengths
- 89 RCTs — largest mind-body immunology meta-analysis
- Six distinct biomarkers assessed
- Multiple intervention types compared
- All included studies are RCTs
Limitations
- Small CRP effect size — clinical significance debatable
- Cannot determine which mind-body modality is most effective
- Blinding impossible for intervention type
- Heterogeneous intervention protocols and durations
Mind-body interventions are now supported by 89-RCT evidence as legitimate immunomodulatory tools — not just stress management strategies. The TNF-α reduction is particularly clinically interesting when juxtaposed with Study 7's finding that dietary weight loss does not affect TNF-α. This suggests complementary pathways: diet for IL-6, mind-body practice for TNF-α and cortisol. Practitioners can prescribe yoga, meditation, or tai chi as components of an anti-inflammatory protocol with confidence in the biomarker evidence.
Synthesis & Editorial Perspective
June’s evidence converges on a principle that should reshape how we design interventions: architecture matters more than category. Not "high protein," but which macronutrient fills the remaining plate. Not "low carb," but whether you cross the 50g/day ketogenic threshold. Not "eat nuts," but which nuts, and how much. The era of categorical dietary advice is giving way to architectural precision.
The gut-brain studies reveal a similar pattern of specificity. Probiotics work for anxiety, depression, and cognition — but only at sufficient dose (≥10⁹ CFU) and duration (≥6 weeks). The cognitive effect (SMD 0.48) exceeds the mood effect, which challenges the assumption that gut-brain interventions are primarily mood tools. Meanwhile, the natural extracts NMA shows that combination formulations (Cistanche + Ginkgo) outperform single compounds — synergy is a feature, not a marketing gimmick.
Perhaps the most integrative insight comes from juxtaposing the weight loss and mind-body studies. Dietary weight loss reduces IL-6 (above 5%) but not TNF-α. Mind-body practices reduce TNF-α and cortisol but not IL-6. These are complementary anti-inflammatory pathways that, prescribed together, would cover a broader inflammatory profile than either alone. The evidence is building for multimodal, systems-level intervention design — not single-target prescriptions.