Cholesterol & Lipids

How the LDL Reduction Calculator Works: Methods and Evidence

A readable explanation of the LDL calculator’s equations, statin dose estimates, combination therapy assumptions, study sources, and limitations.

By Mendel Jacobs, MD·Sep 16, 2026·11 min read
On this page

The LDL reduction calculator explores a straightforward question: if a treatment lowered LDL cholesterol by a given percentage, what would the resulting LDL level be? It combines a starting LDL-C value with estimates drawn from clinical trials and prescribing information.

The arithmetic is exact; the biological prediction is not. A displayed result such as 68.4 mg/dL describes a modeled scenario, not a promise about someone’s next blood test. This article explains the calculation, where the numbers come from, and how to interpret the result.

Educational use only. This tool is not intended for clinical care and does not replace professional medical advice, diagnosis, or treatment. Do not use its estimates alone to make medical decisions or change medication; consult a qualified healthcare professional.

Start with the right LDL value

The calculator has two starting points, and the distinction matters.

Before lipid-lowering medication: enter an untreated LDL value and select the treatments you want to model. Their estimated effects are applied to that starting value.

Already on treatment: enter the LDL measured while taking existing medicines, then select only new add-ons. The calculator assumes existing treatment continues unchanged. It does not reconstruct an untreated LDL, estimate a statin dose increase, or model switching or stopping a medicine.

For example, if someone already takes a statin and ezetimibe, their current LDL already reflects those treatments. Selecting ezetimibe again would count its effect twice. The existing-statin checkbox helps choose the appropriate model input for a new oral add-on; it does not apply a second statin reduction.

Why percentage reductions do not add up

Each additional percentage acts on the LDL remaining after the previous step.

Imagine a starting LDL of 160 mg/dL. A hypothetical treatment lowers it by 50%, leaving 80 mg/dL. A second treatment lowers that remaining 80 by 20%, leaving 64 mg/dL. The combined reduction is 60%, not 70%.

The equation for one step is:

LDL after treatment = LDL before treatment × (1 − estimated reduction / 100)

For several treatments, multiply the remaining fractions:

Final LDL = starting LDL × (1 − r₁) × (1 − r₂) × …

Here, each r is a reduction written as a decimal: 25% becomes 0.25. A medicine that is not selected contributes a multiplier of 1.

The calculator then reports:

  • Absolute reduction: starting LDL − final LDL.
  • Percentage reduction: 100 × (1 − final LDL / starting LDL).

The displayed steps run from statin to ezetimibe, bempedoic acid, and injectable therapy. That order explains the arithmetic; it is not a recommended prescribing sequence. Reordering the same multipliers does not change the answer.

A worked example

Consider an untreated LDL of 160 mg/dL, with atorvastatin 20 mg daily and ezetimibe 10 mg daily selected.

  1. The atorvastatin estimate is 43%, leaving 57%: 160 × 0.57 = 91.2 mg/dL.
  2. The ezetimibe estimate with a statin is 25%, leaving 75%: 91.2 × 0.75 = 68.4 mg/dL.
  3. The absolute reduction is 91.6 mg/dL. The combined percentage reduction is 57.25%, displayed as 57.3%.

This example demonstrates the model. It does not establish that this combination is appropriate for a particular person or that their measured LDL will reach 68.4 mg/dL.

Where the statin estimates come from

The calculator uses a fixed lookup value for each supported statin dose. It does not fit a new dose-response curve or interpolate between doses.

Six statins use the historical dose-response estimates in Law and colleagues’ 2003 meta-analysis, Table 2. Pitavastatin uses the adult dose-response study in the LIVALO prescribing information, Table 7. The studies differ in population and design, so the table should not be interpreted as a head-to-head comparison. Law et al.; LIVALO prescribing information.

Implemented statin inputs: daily dose → estimated LDL-C reduction
StatinDose and estimate
Atorvastatin10 mg → 37%; 20 mg → 43%; 40 mg → 49%; 80 mg → 55%
Rosuvastatin5 mg → 38%; 10 mg → 43%; 20 mg → 48%; 40 mg → 53%
Simvastatin10 mg → 27%; 20 mg → 32%; 40 mg → 37%
Pravastatin10 mg → 20%; 20 mg → 24%; 40 mg → 29%; 80 mg → 33%
Lovastatin20 mg → 29%; 40 mg → 37%; 80 mg → 45%
Fluvastatin20 mg → 21%; 40 mg → 27%; 80 mg → 33%
Pitavastatin1 mg → 31%; 2 mg → 39%; 4 mg → 44%

For any row, the multiplier is 1 − percentage / 100. For example, a 37% reduction uses 0.63.

Pitavastatin’s source study involved older adults. Fluvastatin 80 mg/day refers to XL 80 mg once daily or immediate-release 40 mg twice daily. Simvastatin 80 mg is excluded from the tool.

Intensity and the estimate are different pieces of information

“High intensity” or “moderate intensity” describes a guideline drug-and-dose category. The calculator’s percentage is a historical study estimate. One does not determine the other.

For example, atorvastatin 40 mg is classified as high intensity, while the historical lookup value used here is 49%. The tool preserves that distinction rather than changing the study estimate to match the category. The 2026 goal reference does not make these older dose-response estimates new 2026 data. 2026 dyslipidemia guideline.

Oral medicines beyond statins

The model changes the ezetimibe and bempedoic acid inputs according to whether a statin is present. A statin is considered present when one is selected with an untreated baseline, or when the existing-statin checkbox is selected in add-on mode.

Oral add-on inputs and the multiplier used at that step
MedicineWithout a statinWith a statin
Ezetimibe 10 mg/day18% reduction; LDL × 0.8225% reduction; LDL × 0.75
Bempedoic acid 180 mg/day21% reduction; LDL × 0.7918% reduction; LDL × 0.82

Ezetimibe: the 18% and 25% inputs correspond to treatment-group changes in the label’s monotherapy and statin add-on studies. The add-on study’s control group also had an approximately 4% reduction, so 25% is not the treatment-minus-placebo difference. IMPROVE-IT provides outcomes evidence for ezetimibe but is not the source of these two calculator inputs. Ezetimibe prescribing information, Tables 6–7.

Bempedoic acid: the statin-background input rounds CLEAR Harmony’s 18.1-percentage-point placebo-adjusted effect at 12 weeks to 18%. The other input rounds CLEAR Outcomes’ 21.1-percentage-point effect at six months to 21%. CLEAR Outcomes enrolled statin-intolerant patients; it was not a pure monotherapy trial, and some participants used low-dose statins or other lipid-lowering medicines. Using 21% for every no-statin scenario is a simplification. CLEAR Harmony; CLEAR Outcomes.

PCSK9 therapies and dose schedules

The calculator permits one PCSK9 therapy at a time. Selecting another replaces the current agent. The following inputs describe modeled LDL lowering, not comparative effectiveness or dosing advice.

Injectable inputs, schedules, and evidence
Agent and scheduleReduction / multiplierSource
Evolocumab 140 mg every 2 weeks or 420 mg monthly59% / 0.41FOURIER, pooled schedules, week 48
Alirocumab 75 mg every 2 weeks44% / 0.56ODYSSEY CHOICE I, statin-background cohort, week 12
Alirocumab 150 mg every 2 weeks62% / 0.38ODYSSEY LONG TERM, week 24
Alirocumab 300 mg every 4 weeks54% / 0.46ODYSSEY CHOICE I, statin-background cohort, week 12
Inclisiran 284 mg initially, at 3 months, then every 6 months50% / 0.50ORION-10/11, rounded day-510 estimate

Evolocumab: both schedules use FOURIER’s pooled 59% estimate in patients with established cardiovascular disease receiving statins. The calculator does not model a difference between schedules. FOURIER.

Alirocumab: the 75 mg and 300 mg inputs use CHOICE I results before protocol-driven dose escalation. The 150 mg input comes from a different trial and follow-up period. These numbers do not form a directly measured dose-response curve, and their differences cannot establish that one regimen is superior. PRALUENT prescribing information, ODYSSEY CHOICE I; ODYSSEY LONG TERM.

Inclisiran: the 50% input is rounded from placebo-adjusted reductions of 52.3% and 49.9% at day 510 in ORION-10 and ORION-11. The trial’s 300 mg inclisiran sodium corresponds to the labeled 284 mg inclisiran dose. The calculator includes one standard adult schedule. ORION-10/11; LEQVIO prescribing information.

What the model assumes—and cannot tell you

The main assumption is that each treatment produces an independent proportional reduction in the LDL remaining after the previous step. That is a useful way to illustrate combinations, but every combination shown has not been independently validated in a clinical trial.

The inputs also mix different populations, treatment backgrounds, follow-up times, and ways of reporting change. A within-group change compares treated participants with their own baseline. A placebo-adjusted effect also accounts for the change in the comparison group. Combining these measures does not make them equivalent. The result is an educational approximation, not a harmonized comparison of medicines.

The tool provides a single estimate without a prediction interval. It does not account for individual response, adherence, organ function, pregnancy, pediatric dosing, or rare lipid disorders. It does not calculate the chance of a heart attack or stroke. A percentage reduction in LDL is not the same thing as a percentage reduction in cardiovascular events.

There are limited interaction safeguards: the tool excludes bempedoic acid with selected simvastatin above 20 mg/day or pravastatin above 40 mg/day. Add-on mode warns about those combinations because it does not collect the existing statin dose. This is not a comprehensive interaction checker. NEXLETOL prescribing information.

Units, rounding, and goal comparisons

The calculator accepts values equivalent to 1–1,000 mg/dL. This checks the numerical input, not whether a scenario is clinically appropriate.

To convert mg/dL to mmol/L, it divides by 38.67; the reverse conversion multiplies by 38.67. Calculations retain unrounded values, with final LDL displayed to one decimal in mg/dL or two in mmol/L. Changing units rounds the converted starting input to four decimals. Decimal places make the arithmetic readable; they do not imply that an individual response can be predicted that precisely.

The optional goal comparison asks whether the unrounded estimate is strictly below the selected goal. A value equal to the goal is not below it. The 2026 goal guide is a separate reference: the calculator does not assign risk categories or check every guideline requirement, including percentage-reduction criteria. 2026 dyslipidemia guideline.

Using the result thoughtfully

Use the calculator to understand how a stated percentage translates into an LDL value, explore the arithmetic of a combination, or prepare questions about the evidence. Treatment decisions require a clinician’s assessment, and the actual response is established with follow-up measurements.

This methods description documents the calculator’s inputs and behavior as of September 16, 2026. It explains the model; it does not establish clinical validation.

References

Continue exploring

Mendel Jacobs, MD, MPH

Mendel Jacobs, MD, MPH

Menachem "Mendel" Jacobs, MD, MPH is an Internal Medicine Resident at Yale School of Medicine pursuing academic cardiology. He publishes under Menachem Jacobs.

Connect on LinkedIn