This is a reference class, not a person — and not everyone. It applies general evidence to one common profile; people who are leaner or further along than this need the conditionality flags below, not this exact list.
Who this is for
A Western adult in their late 40s sliding toward metabolic syndrome: central (visceral) adiposity, borderline hypertension, prediabetic-range glucose, raised triglycerides, low-grade inflammation, and early fatty liver (MASLD). Drifting, not diseased — which is precisely why the large, safe levers matter far more than fine-tuning.
Read this first — two framings that reorder everything
- Movement and adiposity are the headline; the carbs-vs-fat-vs-protein wars are the small print. The best-measured levers are activity and body fat; macronutrient composition is second-order — an isocaloric sugar swap is null on weight, low- vs balanced-carb differs by ~1 kg (not clinically meaningful), and cutting saturated fat moves cardiovascular events a little and mortality not at all.
- Sorted by robustness across the distribution, because advice for this drifting majority must not backfire on the leaner ~40%. Each lever is graded: (a) helps this profile, harms no one; (b) helps the metabolically impaired but is needless or mildly counterproductive for the healthy; (c) small or genuinely contested.
(a) The big rocks — pull these; they help this profile and harm no one
1. Aerobic activity — the largest well-measured lever.
- Total activity carries an all-cause mortality HR of 0.34 (95% CI 0.27-0.43) at high certainty — the most robust finding in this whole document (Physical Activity Dose and Mortality).
- Most of that benefit is reached by ~24 min/day of moderate-to-vigorous activity; the curve then flattens. You do not need to be an athlete.
- Sedentary time is a separate lever: risk climbs above roughly 6-8 h/day of total sitting.
2. Resistance training — for function, muscle retention, and a metabolic bonus.
- Its strongest evidence is proven but on surrogates: trials show it preserves lean mass during a weight-loss deficit (Protein and Resistance Training for Muscle and Strength) — you want to lose visceral fat, not muscle — and builds strength. And it harms no one.
- The metabolic bonus that lands here: skeletal muscle is the main site of insulin-stimulated glucose disposal, so building it improves glucose tolerance — directly relevant for this dysglycaemic profile, and why strength training tracks with lower diabetes risk (Muscle-Strengthening Activity and Mortality).
- Its own mortality signal (~RR 0.85 alone; ~0.60 combined with aerobic) is real but very-low certainty (observational, self-reported) — a weakly-warranted bonus, not the reason it earns a place. A couple of sessions a week suffices; no need to chase volume.
3. Lose the visceral fat (primarily via an energy deficit) — the dominant metabolic lever.
It acts on blood pressure, glycaemia, the fatty liver, and the overall drift simultaneously. The honest version:
- It did not reduce cardiovascular events in the definitive trial (Look AHEAD null; a 54-RCT meta-analysis agrees) — so don’t sell it as cardiac protection (Does Weight Loss Reduce Cardiovascular Events).
- It does lower all-cause mortality (RR 0.82, ~6 fewer deaths per 1,000) by a non-cardiovascular route — so pursue it for the liver, glycaemia, blood pressure, function, and that real survival benefit.
- Held targets: ≥5% loss clears liver fat; 7-10% reverses the steatohepatitis behind long-term liver damage (Fatty Liver MASLD and Weight Loss).
- Method matters less than a sustained deficit. The one diet with a clear positive recommendation is defined by energy, not macros — an 800-1,200 kcal total diet replacement for up to 12 weeks (Diets for Weight Loss - What NICE Recommends).
4. Adequate protein (~1.6 g/kg/day). Preserves lean mass during a deficit and supports the training. Benefit plateaus near 1.6 g/kg — more buys no additional muscle (Protein and Resistance Training for Muscle and Strength).
5. Sleep ~7-8 h. Avoid chronic short sleep. More is not a lever, and the elevated risk in long sleepers is a marker of underlying illness, not a target (Sleep Duration and Mortality).
(b) Metabolism-dependent — conditional, not universal
Carbohydrate restriction — for the dysglycaemic only.
- In the insulin-resistant or prediabetic, it lowers HbA1c ~0.47% and drives short-term type-2 diabetes remission (NNT ~3 at 6 months) (Carbohydrate Restriction and Type 2 Diabetes Remission).
- Two catches: the effect attenuates by 12 months, and it raises LDL (~+0.14 mmol/L). For the metabolically healthier member of this class it is needless, and that LDL rise is a net negative — the clearest “helps the impaired, could mildly harm the healthy” call.
- For weight alone, low- vs balanced-carb differs by ~1 kg — not worth the argument (Low-Carbohydrate vs Balanced-Carbohydrate Diets).
(c) Small, over-published, or contested — do not lead with these
- Saturated fat → unsaturated: a small effect on cardiovascular events (RR 0.83; ~15 fewer per 1,000
at high baseline risk, fewer here) and none on mortality (Saturated Fat Intake and Replacement).
- The mechanism is real: cutting saturated fat lowers apoB (the count of atherogenic, apoB-bearing particles), which causally drives atherosclerosis and accumulates over a lifetime (LDL ApoB and Cumulative Exposure). So a sustained small reduction is worth more than a short trial shows. For this metabolically-impaired profile, track non-HDL (a cheap apoB proxy), not LDL-C, which under-counts the particles here.
- Free sugars: the lever is energy, not sugar per se — an isocaloric swap is null on weight (Free Sugars Intake).
- Alcohol: moderate drinking is not protective — the apparent J-curve is an artefact of reverse causation and sick-quitter bias — so there’s nothing to chase upward, and for this fatty-liver-prone profile, less is better (Alcohol and Mortality and Vascular Disease).
- Sodium: a small blood-pressure lever, worth relatively more here (hypertensive-leaning: ~4-6 mmHg vs ~1.4 in normotensives) (Sodium Intake and Blood Pressure). Pharmacological BP-lowering has proven hard-outcome benefit even in primary prevention (Blood Pressure Lowering and Cardiovascular Events), but at borderline pressures the lifestyle route (weight, sodium, activity) is the move, not a drug.
- Red / processed meat: genuinely contested — one guideline family says reduce it, another says current intake is defensible on the same evidence; the clash is the standard of proof and whose values set the threshold, not whether meat causes cancer. Presented as the open question (Should Adults Reduce Red and Processed Meat).
Why diet composition is second-order for this person
This profile is higher-risk than a lean adult but still not high-risk, so the levers act mostly on metabolism, liver and function; the hard cardiovascular payoff of any single cardiometabolic intervention is modest at this risk level, with blood-pressure lowering the partial exception (Cardiometabolic Interventions and Hard CV Outcomes in Low-Risk People). Absolute benefit scales with baseline risk, so a large relative effect on a modest risk is a modest absolute one (Baseline Risk and the Relative-Absolute Split).
Gaps (stated honestly; routed for later)
- Pulses / plant-protein quality: the wiki holds almost nothing — the top acquirable gap for a nutrition guide.
- The “could harm the healthy 40%” sort is reasoned, not measured — it rests on mechanism and held metabolic-state differences (carb→LDL, remission only in the dysglycaemic), not quantified subgroup harm.
- No population statistics — the profile is a stated hypothetical, by design.
- No proven targets for waist or resting heart rate — treat them as progress markers, not cutoffs.
The honest limits
- This grades the evidence, not the outcome — never whether acting on it improves a real one.
- It appraises, it does not prescribe — carbohydrate restriction in particular needs medication management this cannot do.
- A reference class, not a person or everyone — outcome priorities and constraints are the individual’s; the tails need the conditionality flags, not this exact ordering.