# Keeping the Downcycle Going

Date checked: 2026-04-13
Scope: adult weight-loss adherence and "how to keep a calorie deficit going" methods, focusing on effect sizes, especially appetite-control and maintenance tools.

## Key takeaways

1. The biggest sustained effect sizes are from obesity medications, especially tirzepatide and semaglutide, not from supplements. They are in a different league from psyllium, TRE, caffeine, or most diet hacks.
2. Psyllium and other soluble fibers look real but modest. They are plausibly useful as support tools for satiety, not substitutes for GLP-1-class drugs.
3. Meal replacements are underrated because they reliably reduce decision fatigue and often improve 1-year weight loss by a few kilograms versus other diet programs.
4. Time-restricted eating has, at best, a small extra effect on body weight once total calories are controlled. It can still help if it makes adherence easier.
5. Sleep extension is not hype. In a randomized trial, increasing sleep by about 1.2 hours/night reduced energy intake by about 270 kcal/day, which is mechanistically relevant for keeping a deficit going.
6. Exercise is not a strong primary fat-loss lever by itself, but it matters for maintenance, appetite regulation, and preserving lean mass during a cut.
7. The base layer is still behavioral structure: self-monitoring, frequent contact, reduced-energy diet, high enough protein, and repeatable food routines. The flashy methods mostly work on top of that.

## Bottom line ranking

| Method | Typical effect size | Confidence | Best use | Main limitation |
|---|---:|---|---|---|
| Tirzepatide | about 15.0% to 20.9% mean weight loss at 72 weeks; placebo 3.1% | C1-C2 | Strongest appetite suppression and long-term downcycle support | Cost, access, GI adverse effects, likely regain after stopping |
| Semaglutide 2.4 mg | about 14.9% to 17.3% mean weight loss at 68 weeks; placebo about 2.0% to 2.4% | C1-C2 | Very strong appetite suppression | Cost, access, GI adverse effects, regain after stopping |
| Phentermine/topiramate | about 8.0% greater weight loss vs placebo; meta-analysis average 7.73 kg vs placebo | C2 | Strong non-GLP option | Side effects, teratogenicity concerns, prescription-only |
| Naltrexone/bupropion | about 4.1% greater weight loss vs placebo; mean about 4.5 to 6 kg at 28 to 56 weeks | C2 | Craving / reward-driven eating may fit some people | More discontinuation from adverse effects |
| Orlistat | about 3.1% greater weight loss vs placebo; older trial showed 10.2% vs 6.1% at 1 year | C2 | Cheap-ish adjunct, may help prevent regain while on-plan | GI side effects; weak compared with newer drugs |
| Meal replacements | about 1.4 to 6.1 kg extra at 1 year depending on support intensity | C2 | Great for reducing choice overload and making a deficit automatic | Monotony; adherence falls if the plan is too rigid |
| Psyllium / soluble fiber | about 1.25 to 2.5 kg lower body weight in meta-analyses; psyllium-specific MD about -2.1 kg over about 5 months | C2 | Satiety support, stool regularity, glycemic smoothing | Modest effect, needs water, can cause GI issues |
| Intensive behavioral treatment / self-monitoring | usually around 3% to 5% from lifestyle alone; stronger when visits are frequent | C2 | Foundation layer that makes everything else stick | Labor-intensive; effect often erodes without ongoing contact |
| Time-restricted eating | about -1.28 kg vs control in one meta-analysis | C2 | If simplifying eating windows reduces snacking | Small effect once calories are matched |
| Sleep extension | about -270 kcal/day intake in short RCT after +1.2 h/night sleep | C2 | High-value if sleep deprived and hunger-driven | Short-duration evidence; weight effect still indirect |
| Caffeine | signal for weight/fat loss exists, but evidence is noisy and heterogeneous | C3 | Possibly small appetite/energy bump | Tolerance, anxiety, sleep disruption, weak evidence |

## Claims

**Claim:** GLP-1/GIP-class medications have the largest effect sizes for keeping a downcycle going because they materially reduce appetite and raise the odds of sustained clinically meaningful weight loss.
- **Confidence:** C2
- **Sources:**
  - https://www.acc.org/Latest-in-Cardiology/Clinical-Trials/2022/08/04/15/32/SURMOUNT-1
  - https://jamanetwork.com/journals/jama/article-abstract/2821290
- **Date checked:** 2026-04-13
- **Cross-verified:** yes
- **Notes:** ACC summary of SURMOUNT-1 reports mean weight change at week 72 of -15.0%, -19.5%, and -20.9% for tirzepatide 5/10/15 mg versus -3.1% for placebo, and at least 20% weight loss in 30.0%, 50.1%, and 56.7% versus 3.1%. JAMA 2024 review reports placebo-subtracted weight loss of 12.4% for tirzepatide 15 mg and 11.4% for semaglutide across meta-analyses.

**Claim:** Semaglutide is highly effective while taken, but stopping it commonly leads to substantial regain, so it is better understood as chronic treatment than as a one-off cut accelerator.
- **Confidence:** C1
- **Sources:**
  - https://pmc.ncbi.nlm.nih.gov/articles/PMC9542252/
  - https://www.acc.org/Latest-in-Cardiology/Clinical-Trials/2021/02/18/19/23/STEP-1
- **Date checked:** 2026-04-13
- **Cross-verified:** yes
- **Notes:** The STEP 1 extension found mean weight loss of 17.3% at week 68 with semaglutide, followed by regain of 11.6 percentage points by week 120 after withdrawal, leaving a net loss of 5.6%; the paper states participants regained about two-thirds of prior weight loss. ACC notes 86.4% reached at least 5% weight loss on semaglutide versus 31.5% on placebo during STEP 1.

**Claim:** Psyllium has real but modest weight-loss effects and should be viewed as an adjunct for satiety and meal control, not an Ozempic substitute.
- **Confidence:** C2
- **Sources:**
  - https://pubmed.ncbi.nlm.nih.gov/37163454/
  - https://pubmed.ncbi.nlm.nih.gov/29092878/
- **Date checked:** 2026-04-13
- **Cross-verified:** yes
- **Notes:** Psyllium-specific meta-analysis reported MD -2.1 kg body weight, -0.8 kg/m2 BMI, and -2.2 cm waist circumference over mean dose 10.8 g/day for mean 4.8 months. Broader isolated soluble fiber meta-analysis reported -2.52 kg body weight versus placebo, but that includes more than psyllium and shorter interventions.

**Claim:** Soluble fiber is more credible than most supplements for helping a downcycle because it is low-risk, mechanistically coherent, and has repeatable though moderate effect sizes.
- **Confidence:** C2
- **Sources:**
  - https://pubmed.ncbi.nlm.nih.gov/29092878/
  - https://pmc.ncbi.nlm.nih.gov/articles/PMC9268533/
- **Date checked:** 2026-04-13
- **Cross-verified:** yes
- **Notes:** The 2017 meta-analysis found -2.52 kg body weight and -0.84 BMI. The 2022 prolonged soluble-fiber meta-analysis found -1.25 kg body weight versus control, showing a smaller but still significant pooled effect over longer interventions.

**Claim:** Meal replacements are one of the best non-drug adherence tools for keeping a deficit going because they reduce decision load and improve 1-year outcomes.
- **Confidence:** C2
- **Sources:**
  - https://pubmed.ncbi.nlm.nih.gov/30675990/
  - https://pubmed.ncbi.nlm.nih.gov/39433654
- **Date checked:** 2026-04-13
- **Cross-verified:** yes
- **Notes:** The 2019 meta-analysis found additional 1-year weight loss ranging from -1.44 kg to -6.13 kg depending on comparator and support intensity. The 2025 meta-analysis found pooled WMD -3.35 kg body weight, -1.12 BMI, and -2.82 cm waist circumference.

**Claim:** Time-restricted eating can work, but most of its benefit appears to come from making it easier to eat less, not from meal timing itself.
- **Confidence:** C2
- **Sources:**
  - https://pubmed.ncbi.nlm.nih.gov/36875837/
  - https://www.health.com/calorie-counting-vs-time-restricted-eating-8635835
- **Date checked:** 2026-04-13
- **Cross-verified:** partial
- **Notes:** A 2023 meta-analysis found TRE reduced body weight by -1.28 kg and fat mass by -0.72 kg versus controls. In a controlled-feeding study summarized by Health.com, equal-calorie groups lost similar weight, supporting the inference that TRE mostly works through lower intake rather than timing per se. The second source is secondary, so confidence stays at C2 rather than C1.

**Claim:** If you are chronically short on sleep, fixing sleep may be one of the highest-leverage non-drug methods for keeping a deficit going because it reduces spontaneous intake.
- **Confidence:** C1
- **Sources:**
  - https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2788694
  - https://pmc.ncbi.nlm.nih.gov/articles/PMC8822469/
- **Date checked:** 2026-04-13
- **Cross-verified:** yes
- **Notes:** In a randomized trial of 80 adults with overweight and habitual sleep under 6.5 h/night, sleep extension increased sleep by about 1.2 h/night and reduced energy intake by about 270 kcal/day versus control, with no significant change in energy expenditure.

**Claim:** Lifestyle treatment still matters, but lifestyle-only average effect sizes are usually modest unless contact is frequent and structured.
- **Confidence:** C2
- **Sources:**
  - https://pubmed.ncbi.nlm.nih.gov/25369490/
  - https://pmc.ncbi.nlm.nih.gov/articles/PMC12350384/
- **Date checked:** 2026-04-13
- **Cross-verified:** yes
- **Notes:** The 2014 JAMA review found 6-month mean losses from 0.3 kg to 6.6 kg in intervention groups, with larger effects when programs combined calorie restriction, physical activity, and behavioral therapy. The 2025 clinical practice guideline update states health behaviour changes alone generally achieve only modest 3% to 5% weight loss and are often not sustained long term.

**Claim:** Exercise is weak as a sole fat-loss driver but important for maintenance and lean-mass preservation, so it should be treated as a support layer rather than the main deficit engine.
- **Confidence:** C2
- **Sources:**
  - https://pubmed.ncbi.nlm.nih.gov/35156137/
  - https://pubmed.ncbi.nlm.nih.gov/25369490/
- **Date checked:** 2026-04-13
- **Cross-verified:** yes
- **Notes:** Reviews emphasize that physical activity alone is a poor strategy for major weight loss but is crucial for maintenance; primary-care behavioral interventions did better when they combined reduced energy intake with at least 150 min/week activity and behavioral support.

**Claim:** Among older non-GLP drugs, phentermine/topiramate is much stronger than orlistat or naltrexone/bupropion on average, but side-effect and safety tradeoffs are material.
- **Confidence:** C2
- **Sources:**
  - https://pubmed.ncbi.nlm.nih.gov/33864346/
  - https://jamanetwork.com/journals/jama/article-abstract/2821290
  - https://pmc.ncbi.nlm.nih.gov/articles/PMC10498912/
  - https://pubmed.ncbi.nlm.nih.gov/9683204/
- **Date checked:** 2026-04-13
- **Cross-verified:** yes
- **Notes:** JAMA review reports placebo-subtracted weight loss of 8.0% for phentermine-topiramate, 4.1% for naltrexone-bupropion, and 3.1% for orlistat. A phentermine/topiramate meta-analysis found average weight loss of 7.73 kg versus placebo. Older orlistat trial data show 10.2% vs 6.1% at 1 year and less regain while continuing treatment in year 2.

**Claim:** Caffeine is not a serious primary method for keeping a downcycle going compared with the options above.
- **Confidence:** C3
- **Sources:**
  - https://pubmed.ncbi.nlm.nih.gov/30335479/
- **Date checked:** 2026-04-13
- **Cross-verified:** no
- **Notes:** There is a positive signal in meta-analysis, but heterogeneity is very high and the evidence is much weaker and less clinically meaningful than for soluble fiber, meal replacements, or prescription obesity medications. I would treat it as optional and secondary because it can also worsen sleep, which may offset benefit.

## Practical synthesis

If the goal is specifically to keep the downcycle going, the evidence suggests thinking in layers rather than hunting for one perfect hack.

### Layer 1: make the deficit mechanically easier

- Use a repeatable food structure rather than improvising every meal.
- Meal replacements are high-evidence for this because they compress decisions and standardize calories.
- If hunger is the bottleneck, GLP-1 or GIP/GLP-1 drugs are the strongest lever by far, with phentermine/topiramate the strongest older non-GLP option.
- Psyllium or another soluble fiber is worth considering as a small helper, especially before meals, but expectations should stay modest.

### Layer 2: reduce biological pressure to overeat

- Fix short sleep first if relevant; the evidence here is stronger than most supplement talk.
- Keep protein reasonably high to preserve lean mass and satiety during a cut, although I did not find a single clean modern meta-analysis with a simple pooled effect size strong enough to summarize confidently here.
- Avoid relying on caffeine as a main strategy if it worsens sleep, anxiety, or rebound hunger.

### Layer 3: keep adherence from decaying over time

- Frequent self-monitoring and regular contact matter.
- Daily or near-daily weighing, logging, and a fixed meal template are boring but evidence-aligned.
- Exercise should be used mostly to preserve muscle, maintain function, and help maintenance rather than expecting huge scale changes from cardio alone.

## What seems most worth adding to the wiki

1. A ranked section on appetite-control tools, separating major-effect-size options from minor ones.
2. A dedicated comparison table for obesity medications with placebo-subtracted percentage weight loss and discontinuation caveats.
3. A "boring but works" section covering meal replacements, self-monitoring, sleep, and repeatable food routines.
4. A "small helpers" section for psyllium/soluble fiber, TRE, and caffeine, with honest language about modest effects.
5. A caution that stopping GLP-1s often leads to regain, so maintenance planning has to start before discontinuation.

## Sources used

- https://www.acc.org/Latest-in-Cardiology/Clinical-Trials/2022/08/04/15/32/SURMOUNT-1
- https://www.acc.org/Latest-in-Cardiology/Clinical-Trials/2021/02/18/19/23/STEP-1
- https://jamanetwork.com/journals/jama/article-abstract/2821290
- https://pmc.ncbi.nlm.nih.gov/articles/PMC9542252/
- https://pubmed.ncbi.nlm.nih.gov/37163454/
- https://pubmed.ncbi.nlm.nih.gov/29092878/
- https://pmc.ncbi.nlm.nih.gov/articles/PMC9268533/
- https://pubmed.ncbi.nlm.nih.gov/30675990/
- https://pubmed.ncbi.nlm.nih.gov/39433654
- https://pubmed.ncbi.nlm.nih.gov/36875837/
- https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2788694
- https://pmc.ncbi.nlm.nih.gov/articles/PMC8822469/
- https://pubmed.ncbi.nlm.nih.gov/25369490/
- https://pmc.ncbi.nlm.nih.gov/articles/PMC12350384/
- https://pubmed.ncbi.nlm.nih.gov/35156137/
- https://pubmed.ncbi.nlm.nih.gov/33864346/
- https://pmc.ncbi.nlm.nih.gov/articles/PMC10498912/
- https://pubmed.ncbi.nlm.nih.gov/9683204/
- https://pubmed.ncbi.nlm.nih.gov/30335479/