Meet gostrong: training and nutrition that cite their sources
gostrong plans training and nutrition together with a constraint solver, and ties every load-bearing constant to a published meta-analysis or position stand.
Most apps hand you a template or a calorie number and ask you to trust it. We built gostrong because we wanted to know why a number was what it was, and we wanted the answer to be a source, not a shrug.
Science over popular methodology
Every load-bearing constant in gostrong is anchored to a meta-analysis, a position stand or a systematic review. The rule we work by is simple: if a published meta-analysis disagrees with popular coaching material, the meta-analysis wins.
That is not a slogan. Our training templates were rewritten when the literature disagreed with them. Four examples:
| Choice | What the evidence says | What gostrong does |
|---|---|---|
| Proximity to failure | Hypertrophy drops meaningfully past about 4 reps in reserve (RIR), and gains are roughly equivalent across 0 to 2 RIR (Refalo 2023) | Flat RIR 2 for hypertrophy blocks |
| Scheduled deloads | A scheduled deload at the midpoint of a 9-week cycle worsened strength gains and did nothing for hypertrophy (Coleman 2024) | No calendar deloads; a deload fires on performance signals only |
| Weekly volume | About 0.24% hypertrophy per extra set at roughly 12 sets per week, with clear diminishing returns beyond that (Pelland 2025) | A volume objective that penalises overshooting more than undershooting |
| Volume across the cycle | No independent frequency effect when weekly volume is equated (Schoenfeld 2019) | The same weekly set prescription every week; load, not added sets, is the overload variable |
The nutrition side follows the same discipline. The protein floor is 1.93 g per kg of fat-free mass: the 1.62 g per kg bodyweight plateau found by Morton 2018, re-expressed for a lean trainee (about 84% fat-free mass). Target rates of weight change are 0.5 to 1.0% of bodyweight per week for fat loss (anchored to Garthe 2011) and 0.25 to 0.5% for muscle gain (Garthe 2013).
We also audit ourselves. A recent review of our constants against a large research corpus found zero uncited-and-wrong constants and zero hard conflicts. Where a topic is genuinely contested, such as the per-meal leucine dose, we say so and keep the hedge.
Training and diet, planned together
In gostrong the diet phase drives every calorie number, and the meal plan is solved against the training you actually have that day.
Here is what ships:
- Diet phases: fat loss, maintenance or muscle gain, with a rate and a duration. When a fat-loss phase ends, maintenance is seeded from your observed energy expenditure, and the app suggests a hold of about two thirds of the weeks you dieted. It is advice, not a lock.
- Meal plans: generated automatically, with a regenerate-day option.
- Shopping list: the week’s plan summed into one line per food, in grams.
- Routine generation: a weekly training template with exercises, sets, reps and RIR.
- Load progression and deloads: session targets that follow your data, not the calendar.
- Training journal: macrocycle down to per-set logging.
- Dashboard and statistics: today at a glance, and history across nutrition, body, sleep and training.
- Coach: a read-only page that tells you what to change and why.
A solver builds the plan, not a template
Meal plans and weekly training templates are combinatorial problems. Pick foods and grams so that protein, fat, carbohydrate, fish intake, mercury exposure and the weekly category limits all hold at once. Pick exercises and sets so that every muscle lands in a sensible volume range. A fixed template cannot do that for your body, equipment and schedule, so gostrong hands both problems to a constraint solver (OR-Tools CP-SAT, running as a separate service) and treats the science as the constraints and the objective.
A meal-plan solve takes around 30 seconds on a realistic food catalog, so it runs in the background and the page tells you when the result lands. The solver never decides your working weight: loads belong to the progression engine.
The plan adapts from what you log
A plan that never changes is just a different kind of template. gostrong closes the loop with two feedback mechanisms.
Training. After each week closes, the app looks at your session feedback. If performance drops at a matched load, it arms a deload. That deload cuts several variables at once (sets, reps, load and effort), following Bell 2024, and resumes from the last load you actually performed. Between blocks, each muscle’s volume landmarks are recalibrated from how you responded, never overridden mid-session.
Diet. The app compares your weekly weight trend with the rate your phase expects. The classic 3,500 kcal per pound rule is a static simplification that overestimates long-run change because energy expenditure adapts (Hall 2011), so we use it only as a starting seed and correct against your weigh-ins.
Load progression follows the same spirit. Reps progress before the load does: an easy session is banked as reps until the next plate is worth taking, and the load then moves by at most one step.
A week with gostrong
Setup, once. Onboarding is four steps: profile, training, diet, schedule. When you finish, your first meal plan and your first training mesocycle are already waiting on the dashboard.
Monday. Open the dashboard: today’s diet day shows what you have eaten against the day’s targets. The shopping list has the week’s groceries in grams.
Training days. Open the session. Each exercise shows a target such as 3 x 12 at 60 kg, 2 RIR, with the reason in words, and each set shows what that set did last time. The logger prefills from the prescription, and after your first set it prefills from the set you just did. Log weight, reps and RIR.
Every day. Tick off the meals you ate and swap a food if you need to. Log your weight and, optionally, body fat.
Midweek. Check the coach if a lift stalls. It reads training, diet, body and sleep together and tries to say which single thing to change first. Warnings come with a one-line note on the evidence behind them.
End of week. The weekly loop runs once the week is over. It may leave everything alone, adjust volume, or arm a deload. Next week’s meal plan is already there. Statistics show volume against each muscle’s landmarks and your weight trajectory against the phase target.
The details nobody reads
The unglamorous constraints matter. The weekly omega-3 floor is 250 mg of EPA plus DHA per day, or 1,750 mg per week (EFSA). The mercury ceiling is 0.7 micrograms per kg bodyweight per week, from the US EPA reference dose. The solver satisfies both while still giving you food you will eat.
What is coming on this blog
This post is the map. Over the next weeks we will go deeper on four topics:
- glycogen and the 28-day weight trend, and why your scale can mislead you for weeks;
- training volume via constraint solving, and how a solver picks sets;
- omega-3 against mercury in meal plans;
- velocity-based training, and what bar speed tells a plan.
Try it
If you have ever wondered why your app chose a number, gostrong is built for you.
References
- Pelland JC, et al. (2025). The Resistance Training Dose-Response: Meta-Regressions Exploring the Effects of Weekly Volume and Frequency on Muscle Hypertrophy and Strength Gain. Sports Medicine. https://pubmed.ncbi.nlm.nih.gov/41343037/
- Refalo MC, et al. (2023). Influence of Resistance Training Proximity-to-Failure on Skeletal Muscle Hypertrophy. Sports Medicine. https://pmc.ncbi.nlm.nih.gov/articles/PMC9935748/
- Coleman M, et al. (2024). Gaining more from doing less? The effects of a one-week deload period during supervised resistance training on muscular adaptations. PeerJ. https://peerj.com/articles/16777/
- Bell L, Pinto J, Helms E, et al. (2024). Deloading Practices in Strength and Physique Sports: A Cross-sectional Survey. Sports Medicine - Open. https://pmc.ncbi.nlm.nih.gov/articles/PMC9811819/
- Garthe I, et al. (2011). Effect of two different weight-loss rates on body composition and strength and power-related performance in elite athletes. Int J Sport Nutr Exerc Metab 21(2):97-104. https://pubmed.ncbi.nlm.nih.gov/21558571/
- Garthe I, et al. (2013). Effect of nutritional intervention on body composition and performance in elite athletes. Eur J Sport Sci 13(3):295-303. https://pubmed.ncbi.nlm.nih.gov/23679146/
- Morton RW, et al. (2018). A systematic review, meta-analysis and meta-regression of the effect of protein supplementation on resistance training-induced gains in muscle mass and strength. Br J Sports Med 52:376.
- Schoenfeld BJ, et al. (2019). How many times per week should a muscle be trained to maximize muscle hypertrophy? J Sports Sciences.
- Hall KD, et al. (2011). Quantification of the effect of energy imbalance on bodyweight. Lancet 378:826-837.
- EFSA (2010). Panel on dietetic products, adequate intake of EPA+DHA for adults.
- US EPA. Reference dose for methylmercury.