Week 2: What Counts as a Recommendation

1 March 20277 min read

What a guideline is made of#

You look up the official advice on something, find a sentence telling you what to do, and act on it. But finding an official guideline does not mean every sentence in it is equally authoritative. Some sentences are the advice itself, the part a committee argued over with a body of evidence behind it. Others are examples, suggestions and explanations, written to help you picture the advice or to round out a booklet. The documents are often open about this, and Australia’s 24-hour movement guidelines print their recommendations and then a separate heading that says “Tips”. Once a sentence is quoted on its own, in a news story or by a friend or by you a fortnight later, the two kinds look identical. Telling them apart again is this week’s work, and it does not require reading any research. It requires asking what the document already says about itself, which is reliably the part nobody reads. Four questions do most of it, and they are ordered by how much each can save you. Is this still in force? Who is it written for? Is it a recommendation or a tip? How sure is it? The first can end the matter in seconds. The last may send you hunting for a separate evidence report. Asking them in that order means the expensive question only ever gets asked about a sentence that has already survived the cheap ones, and together they tell you how much weight that sentence will hold before you rebuild a morning around it.

Four questions, in order#

Is it still in force?#

Ask this first, because it is the only question here whose answer can end the matter outright, and it takes seconds. NHMRC states that its guidelines “have a limited life. They are regularly reviewed and will be updated or withdrawn in light of important new evidence that may emerge.” Withdrawn ones move to the Australian Government Web Archive, and the register of them carries entries like “Exercise and the older person (1994)”, which is to say advice about movement for older people that was once national guidance and is now history. The movement guidelines page does the opposite and prints the date it was last updated, currently 3 June 2026, so you can see at a glance that somebody has looked at it recently. Two limits on the check itself. NHMRC’s own register says “This list may not be fully comprehensive”, so absence from it proves less than you would like. And the Australian Clinical Practice Guidelines Portal, which existed to index guidelines, is no longer active.

Who is it written for?#

Next, because the answer decides which page you are reading at all. Usually it redirects rather than stops you: you find the document, discover the part in front of you is not addressed to you, and go to the part that is. Week 1 covered the three axes guidance narrows on before it ever reaches you. The check inside a single document is narrower still: read who the document says it is for, then read who each sentence inside it is for, because those are not always the same thing. The fourth Australian Dietary Guideline is “Encourage, support and promote breastfeeding”, addressed to a reader who may not be feeding anyone, sitting inside a document addressed to everyone. That is not an error. It is what a guideline is, a document written for a population, and a population contains people in circumstances you are not in. Australia’s movement guidelines make the same point structurally, splitting into separate pages for adults, older adults, pregnancy and people with disability, so that finding the page with your name on it is the first act of reading them at all. A sentence whose audience you have not checked is one you may be following on somebody else’s behalf.

Is it a recommendation, or a tip?#

Not one of the Australian Dietary Guidelines mentions breakfast. There are five of them, each described as “equally important in terms of public health outcomes”, and they cover energy balance and activity, variety across five food groups, limiting saturated fat and added salt and added sugars and alcohol, breastfeeding, and food safety. “Enjoy a healthy breakfast” is a practical tip in the summary booklet, a different document by the same authors, so citing it as a dietary guideline quietly promotes a tip into a recommendation. The movement guidelines keep the two kinds apart on the page, and a sentence lifted out of them still loses the difference. The adult recommendations are thirty minutes or more of moderate to vigorous activity on most days, muscle strengthening on two days or more, functional work for mobility and balance on three, several hours of light activity daily, limiting and breaking up sedentary time, and seven to nine hours of sleep with consistent bed and wake times. The seven thousand step figure is nowhere among them; it sits in a companion statement. Both numbers are worth knowing and neither is a recommendation, and the difference is not pedantry, because it tells you how much argument is standing behind the sentence.

How sure is it?#

A modern guideline rates itself twice, and the two ratings are independent. NHMRC uses GRADE, which it describes as an approach “to rate the certainty of evidence and the strength of recommendations”. Certainty comes in four levels: high, moderate, low and very low. Strength comes in two, strong and conditional, and those definitions repay attention. A strong recommendation means the panel expects almost all fully informed people to choose it. A conditional recommendation means the panel expects fully informed people to differ. Read that again, because it is this course’s premise written by the people who write the guidelines: a conditional recommendation is a formal statement, inside the guideline system itself, that this is a place where you are meant to decide for yourself. The two ratings also come apart more often than the words suggest. A review of eight Irish national clinical guidelines found that 151 of 202 recommendations were strong, and that only 55 of those 151 rested on high or moderate certainty evidence.

Two cautions before you carry that anywhere. Those are clinical guidelines, which are the most carefully graded kind of guidance there is, so the figure says nothing directly about what you eat or how far you walk; if anything it flatters the field, because these documents at least published both ratings. The authors also note that no guideline in the set said openly that it was making a strong recommendation on weak evidence. The Australian position for a general reader is different and in one respect worse. The adult movement guidelines page carries no certainty rating and no strength rating at all, not one mention of either, with the supporting evidence held in a separate final report. So for most public guidance the honest answer to how sure the document is, is that the page will not tell you and you have to go and find out.

All four questions are the same move: find out what a sentence actually claims before you act on it. That is this week’s lab as well, which asks you to operationalise tiredness, to say what you will count as tired before you try to measure it. A guideline that has not told you whether a sentence is a recommendation or a tip, and a measure that has not told you what counts as tired, fail in the same way. Both let you act with confidence on a sentence whose meaning was never fixed.