Build a glutathione study card that preserves the actual question

Begin with a record you can identify

Use the paper’s title, authors, journal year and PubMed identifier. The source-card builder includes an example from Richie and colleagues, published in the European Journal of Nutrition in 2015. Its PubMed record is 24791752. Keeping that identifier makes it possible to distinguish the paper from a retailer’s retelling, a later review or a different formulation study with a similar title. A recognizable citation is the first part of a useful evidence trail.

Sources for this section: [1] Richie et al., 2015 — oral glutathione trial, indexed abstract

Fill in the study design and population

The example was a six-month randomized, double-blind, placebo-controlled study in 54 non-smoking adults. Its main question concerned oral glutathione and body stores of glutathione. Record those features together. Do not quietly replace the participants with people who have a disease the trial did not study, or replace the oral preparation with a different route. This brief guide uses the indexed abstract and citation rather than claiming an independent review of the complete trial dataset.

Sources for this section: [1] Richie et al., 2015 — oral glutathione trial, indexed abstract

Name the measured outcome precisely

The abstract reports increases in glutathione stores in measured body compartments. That is a biochemical finding. It does not by itself demonstrate that the intervention prevents a disease, improves a symptom or produces a longer-term clinical outcome. In the card, put the measured outcome in one field and the unanswered clinical question in another. This preserves the finding while keeping its boundary visible.

Sources for this section: [1] Richie et al., 2015 — oral glutathione trial, indexed abstract

Use the form as a reading aid

Choose the study-design category, preparation or route and outcome category. Then enter the population, measured outcome and unresolved question in your own words. Build the card and read it back. If your summary promises more than the endpoint supports, revise it before downloading. The tool does not validate your interpretation against a paper or fill missing details with an AI-generated conclusion. It organizes the text you supply.

Write an unresolved question that can be investigated

“Does it work?” is too broad for the final field. A more useful question names the gap: “Does this biochemical change translate into a meaningful outcome in the population I am reading about?” Another asks whether the marketed preparation matches the studied one. These questions are not proof of benefit or harm. They define what further evidence would need to answer. Leave a field marked unknown when the available source cannot settle it.

Save a version you can compare later

Download the source card with its citation. If you change an input after building, rebuild the card so the output matches the current fields. For a later paper, create a separate card and compare population, formulation, comparator, endpoint and follow-up. Avoid combining the most favorable result from several different studies into one claim that no individual study tested. The resulting collection should make disagreement and missing information easier to see.