Advice about sleep changes direction regularly, and each new version is presented with confidence. The instability comes from how sleep is measured and how findings about groups are translated into instructions for individuals.
Measurement methods do not agree with each other
Laboratory measurement records brain activity, eye movement and muscle tone, and it defines sleep stages by specific electrical patterns.
Consumer wearables infer sleep from movement and heart rate, and self-reported diaries record what a person believes happened.
These three produce different numbers for the same night, so studies using different methods can genuinely disagree while all being accurate on their own terms.
Averages conceal wide individual variation
Recommended sleep durations are ranges drawn from population data, and the spread within any population is considerable.
A figure that describes the middle of a distribution says little about where any particular person sits, yet it is frequently reported as a target everyone should hit.
The result is guidance that feels wrong to a substantial share of the people receiving it, which erodes trust in the next piece of guidance.
Cause and effect are difficult to separate
Much sleep research is observational, recording that two things occur together rather than showing that one produces the other.
Poor sleep accompanies a great many conditions, and it can plausibly be a cause, a consequence, or a shared symptom of something else entirely.
Reporting frequently collapses that ambiguity into a simple causal claim, which is where a cautious finding becomes an overconfident headline.
Advice is filtered before it reaches the reader
A study reports a modest association with several caveats. A press release compresses it, an article compresses it further, and a summary reduces it to an instruction.
Each step removes qualification, and the qualification is usually the part that explained who the finding applied to.
Two studies with compatible findings can therefore arrive as contradictory advice, purely through the compression applied on the way.
Tracking can become part of the problem
Detailed nightly data invites comparison against a target, and anxiety about reaching that target is itself disruptive to sleep.
The devices producing the data are, as discussed above, estimating rather than measuring, so the number driving the anxiety carries real uncertainty.
Persistent difficulty sleeping is a matter for a clinician rather than for a different app, since several distinct conditions present as unrefreshing sleep and require different approaches.