PeakCalcs

Methodology & Editorial Policy

PeakCalcs is a health and fitness estimation platform. Every calculator on this site uses formulas drawn from peer-reviewed scientific publications. This page explains how those formulas are selected, implemented, verified, and maintained — and what this process cannot guarantee.

Formula Selection

Formulas are selected based on three criteria: peer-reviewed publication, widespread validation, and practical applicability to a general fitness audience.

PeakCalcs only implements formulas published in recognised scientific journals or official health organisation guidelines. Secondary sources such as fitness blogs, social media content, or commercially produced training manuals are never used as primary references. When a formula has been revised or superseded (as with Harris-Benedict, revised by Roza and Shizgal in 1984), PeakCalcs implements the most recent validated revision and cites both the original and the revision.

For calculators where multiple validated formulas exist (metabolic rate, one-rep max, body fat percentage), PeakCalcs implements all major formulas and presents results side by side. This multi-formula comparison is a deliberate design choice: it gives users a more honest picture of the inherent uncertainty in population-level estimates than any single number could.

Formula Sources

The following primary sources are used across PeakCalcs calculators:

Metabolic Rate (BMR / TDEE)

  • Mifflin MD, St Jeor ST, et al. "A new predictive equation for resting energy expenditure in healthy individuals." American Journal of Clinical Nutrition, 1990; 51(2):241-247.
  • Roza AM, Shizgal HM. "The Harris Benedict equation reevaluated: resting energy requirements and the body cell mass." American Journal of Clinical Nutrition, 1984; 40(1):168-182.
  • Katch F, McArdle WD. Introduction to Nutrition, Exercise, and Health. 4th ed. 1983.

Strength (One-Rep Max)

  • Epley B. Poundage Chart. Boyd Epley Workout. 1985.
  • Brzycki M. "Strength testing: predicting a one-rep max from repetitions to fatigue." JOPERD, 1993; 64(1):88-90.
  • Lander J. "Maximum based on reps." NSCA Journal, 1985; 6(6):60-61.

Body Composition

  • Hodgdon JA, Friedl KE. "Development of the DoD body composition estimation equations." Technical Document 99-2B. San Diego: Naval Health Research Center; 1999.
  • Jackson AS, Pollock ML. "Generalized equations for predicting body density." British Journal of Nutrition, 1978; 40(3):497-504.
  • Deurenberg P, Weststrate JA, Seidell JC. "Body mass index as a measure of body fatness." British Journal of Nutrition, 1991; 65(2):105-114.

Implementation Process

Each formula is implemented as a standalone TypeScript function with explicit parameter names matching the original publication's variable definitions. Internal calculations use metric units (kilograms, centimetres) as the primary standard, converting from imperial inputs where necessary. This matches the unit systems used in the original publications and avoids conversion errors in the calculation chain.

Formula implementations are isolated from the user interface. The calculation engine is a pure function: given the same inputs, it always produces the same outputs, with no side effects or state dependencies. This isolation makes testing straightforward and ensures UI changes cannot accidentally alter calculation results.

Verification and Testing

Every calculator undergoes a multi-stage verification process:

Stage 1: Implementation Review

The formula implementation is compared line by line against the original publication. Coefficients, operations, and variable definitions are checked. Common errors such as swapped gender coefficients, incorrect unit assumptions, and rounding at intermediate steps are specifically checked for.

Stage 2: Known-Value Testing

Each calculator has automated tests using reference inputs and expected outputs. These known-value tests verify that the implementation produces correct results for at least one primary scenario and one variant or edge case. Test values are derived from worked examples in the original publications where available, or from verified reference implementations.

Stage 3: Worked Example Verification

Every worked example on a calculator page uses real inputs that are run through the actual calculation engine. All intermediate values and final results in the worked example narrative are verified against the engine output. If a discrepancy is found, the worked example is corrected to match the calculation — never the other way around.

Stage 4: Source Link Verification

Every external citation and source link is periodically checked to confirm it still resolves to the correct publication. Broken links are replaced with working alternatives pointing to the same source material.

Health Content Guardrails

PeakCalcs enforces several non-negotiable rules for health content safety:

  • No clinical or diagnostic tools — PeakCalcs does not build GFR calculators, drug dosage tools, medical risk scores, or blood panel interpreters.
  • Estimation language only — results use "estimates", "predicts", "suggests", never "determines", "proves", or "confirms".
  • Calorie floors enforced in code — 1,200 kcal/day for women, 1,500 kcal/day for men. Results below these thresholds display a medical supervision warning.
  • No extreme deficit tools — no very-low-calorie diet planners or rapid weight loss calculators.
  • Body fat categories use neutral language — "Essential Fat", "Athletic", "Fitness", "Average", "Above Average".
  • Pregnancy calculators include enhanced disclaimers with IOM 2009 guideline citations and explicit direction to consult an obstetrician or midwife.

What This Process Cannot Guarantee

No verification process eliminates all sources of error. Population-level formulas are statistical models trained on specific study populations. An individual's actual metabolic rate, body composition, or strength may differ meaningfully from any formula's prediction.

PeakCalcs does not conduct original research, recruit study participants, or generate new empirical data. The site implements and presents existing published formulas. When the underlying research has limitations — and all studies do — those limitations carry through to the calculator's output.

The multi-formula comparison approach helps illustrate the range of estimates across validated methods, but it does not identify which estimate is correct for a given individual. That determination requires professional assessment using clinical methods beyond the scope of any online calculator.

About the Author

PeakCalcs is built and maintained by Dan Dadovic, a Commercial Director and PhD candidate in IT Sciences based in Northumberland, UK. Dan's expertise is in computational modelling, data analysis, and web application development. He is not a doctor, dietitian, or certified personal trainer. PeakCalcs provides estimation tools based on published research, not medical or nutritional advice.

For corrections, feedback, or questions about this methodology: contact@peakcalcs.com