How we cite sources

What counts as a source

In rough order of weight:

• Systematic reviews and meta-analysespublished in peer-reviewed journals. Highest weight when they pool well-conducted RCTs.
• Clinical practice guidelines from recognised bodies — American Academy of Sleep Medicine (AASM), NICE, European Sleep Research Society, the British Sleep Society. These represent expert consensus on the underlying evidence.
• Randomised controlled trials published in peer-reviewed journals. The methodological gold standard for causal claims.
• Large prospective cohort studieswhen RCTs aren't feasible (you can't randomise people to decades of short sleep).
• Mechanistic / laboratory studies where they support the chain of reasoning behind a clinical finding.

What we avoid

• Press releases and news articles standing in for the underlying research.
• Pre-prints (not yet peer-reviewed) as the sole basis for a strong claim — though we'll cite them when nothing better exists, with a note.
• Industry-funded research without independent corroboration.
• Anecdote, podcast transcripts, and supplement marketing copy.

How we cite

Each cited source includes:

• Author(s) and year
• Journal name and the study title
• Direct link (PubMed where available, journal otherwise)
• PubMed ID (PMID) when present, so the source is identifiable even if the URL changes

Sources appear at the bottom of each article in a numbered list. Inline claims that draw on a specific source link directly to the source page.

When studies disagree

Disagreement is common in sleep research — sample sizes are often small, populations vary, and effects can be modest. When the literature is mixed, we say so explicitly. Our default position:

• If a meta-analysis or systematic review exists, we weight it heavily and report its conclusion.
• If individual trials disagree, we identify the most methodologically sound and explain why we've weighted them as we have.
• Where genuine uncertainty exists, we label the evidence as Limited or Emerging — not Moderate or Strong.

Why we link to PubMed where possible

PubMed indexes the bulk of biomedical research and is freely accessible, with stable URLs and durable identifiers. Many journal links eventually rot or paywall; PubMed links don't. Journal pages are linked where they add value (e.g. for open- access full text).

How to flag a problem with a source

If you spot a citation that looks wrong, a study that's been retracted, or a source that doesn't support the claim it's attached to, please tell us. We'll investigate and correct.