The other day, my patient Nadia, 42, placed a stack of books on my desk as tall as a dictionary. "I've read everything, Diana. Keto, Atkins, Weight Watchers, paleo, intermittent fasting, cabbage soup — yes, even the cabbage soup. I've lost and regained the same stone four times in six years. Tell me the truth: which one actually works?" I took a deep breath. Because the honest answer is more nuanced than any diet book is willing to admit.
Here's what the science says — not Instagram, not sponsored testimonials, not influencers selling their "revolutionary method." Meta-analyses, meaning studies that compile and analyse dozens of clinical trials to extract reliable conclusions. This is the highest level of evidence in medicine — and it's the only thing that should guide your food choices.
In this article
- The truth nobody wants to hear
- The Mediterranean diet: the undisputed champion
- The DASH diet: the overlooked powerhouse
- The ketogenic diet: effective but at what cost?
- Weight Watchers (WW): the points system under the microscope
- The paleo diet: back to basics or marketing gimmick?
- Plant-based eating: ethical AND effective?
- The big comparison table
- Sustainability: the only criterion that truly matters
- Frequently asked questions
The truth nobody wants to hear
Let's start with the most robust finding from nutrition research over the past 20 years. In 2009, the POUNDS Lost study, published in the New England Journal of Medicine, compared four diets with vastly different macronutrient compositions (low-fat, high-protein, low-carb, etc.) in 811 participants followed for 2 years.
The result? All diets produced similar weight loss at 6 months (-6 kg on average). And all showed partial regain at 2 years. Macronutrient composition wasn't the determining factor — adherence was. The people who followed their diet most consistently, whichever it was, lost the most weight.
In other words: the best diet is the one you can maintain long-term. This sentence should be displayed in large print in every nutritionist's office. But it doesn't sell books, so nobody puts it on the cover.
The number that makes you think: According to a 2015 meta-analysis in the American Journal of Clinical Nutrition, approximately 80% of people who follow a diet regain the lost weight (or more) within 2 years of stopping. This phenomenon is called the yo-yo effect — and it's more harmful to health than the stable overweight it claims to treat.
The Mediterranean diet: the undisputed champion
If nutrition research had to award a gold medal, it would go to the Mediterranean diet — and it wouldn't even be close. No other dietary pattern has accumulated as much scientific evidence of health benefits.
What it is
Not a "diet" in the restrictive sense, but a dietary pattern inspired by the traditional eating habits of Mediterranean countries (Greece, southern Italy, Spain). Base: fruits, vegetables, legumes, whole grains, olive oil, fish, nuts. In moderation: dairy, poultry, eggs, red wine. Rarely: red meat, added sugars, ultra-processed foods.
What the science says
- Weight loss: Modest but sustainable. BMJ meta-analysis (2008): -1.75 kg over 12 months vs standard diet. Not spectacular, but maintained — which is remarkable.
- Cardiovascular health: The PREDIMED trial (7,447 participants, 5-year follow-up) showed a 30% reduction in major cardiovascular events. A colossal result.
- Type 2 diabetes: 23% risk reduction according to a 2014 meta-analysis.
- Cancer: 6% reduction in overall cancer risk and 7% for breast cancer (2017 meta-analysis).
- Longevity: 8-10% reduction in all-cause mortality in observational studies.
Who is it for?
Everyone, literally. It's the only diet recommended by the WHO, the AHA (American Heart Association), and most learned nutrition societies. It has no known contraindications (barring specific food allergies) and no documented negative side effects.
The limitation
It doesn't produce rapid weight loss. If your goal is to lose 2 stone in 3 months, the Mediterranean diet won't meet that expectation (no sustainable diet will, but that's another topic). It's designed for the long game — and that's precisely what makes it superior.
The DASH diet: the overlooked powerhouse
DASH stands for Dietary Approaches to Stop Hypertension. Developed by the NIH (National Institutes of Health), it was designed to treat high blood pressure — but its benefits go far beyond.
What it is
Fruits, vegetables, whole grains, lean proteins, low-fat dairy, nuts and seeds. Restricted: salt (< 2,300 mg/day, ideally < 1,500 mg), added sugars, saturated fats, red meat. Very close to the Mediterranean, with a particular emphasis on sodium reduction.
What the science says
- Hypertension: Blood pressure reduction of 6-11 mmHg systolic — comparable to a first-line antihypertensive medication.
- Weight loss: Modest (-2 to -3 kg over 6 months) but durable.
- Cardiovascular health: 20% reduction in heart failure risk.
- Diabetes: Improved insulin sensitivity.
Who is it for?
Particularly indicated for people with hypertension or cardiovascular risk. But like the Mediterranean, it benefits everyone. In the UK, average salt intake is 8g/day (the NHS recommends under 6g). Simply reducing salt has measurable effects.
Note: DASH can be challenging for people who eat lots of ready meals or takeaways — most dietary salt comes from processed foods (bread, cured meats, cheese, ready meals), not the salt shaker.
The ketogenic diet: effective but at what cost?
Keto is probably the most hyped diet of the past decade. It makes spectacular promises — and delivers on some. But the nuances are crucial.
What it is
Drastic carbohydrate reduction (< 50g/day, sometimes < 20g) in favour of fats (70-80% of calories) and moderate protein. The goal: force the body into ketosis, where it uses fat as its primary fuel instead of glucose.
What the science says
- Weight loss: Superior to low-fat diets short-term (6 months) — Cochrane 2022 meta-analysis: -1 to -2 kg more. But at 12 months, the difference almost entirely disappears.
- Blood sugar: Significant HbA1c reduction in type 2 diabetics — one of the most robust results.
- Epilepsy: The original medical indication. Proven effective since the 1920s for drug-resistant epilepsy.
- Blood lipids: HDL increase (good cholesterol), but also LDL increase in some — which concerns cardiologists.
The problems
- Catastrophic sustainability: The dropout rate is the highest of any studied diet. Very few people maintain strict keto beyond one year.
- Side effects: "Keto flu" (fatigue, headaches, irritability) at the start. Frequent constipation. Bad breath (acetone).
- Social impact: Extremely restrictive at restaurants, with family, in social settings. Bread, pasta, rice, fruit — almost everything is off limits.
- Potential risks: Fibre, micronutrient, and antioxidant deficiency. Increased renal load. Insufficient long-term studies.
My clinical view: Keto can be a legitimate therapeutic tool for epilepsy or type 2 diabetes under medical supervision. As a general weight loss strategy, I don't recommend it — the long-term failure rate is too high and the risks too poorly documented. If you want to reduce carbs, a moderately low-carb diet (100-150g/day) offers most benefits without the extreme constraints.
Weight Watchers (WW): the points system under the microscope
WW is the most scientifically studied commercial weight loss programme — and one of the few with results published in peer-reviewed journals.
What it is
A points system assigning a value to each food based on its nutritional profile (calories, saturated fat, sugar, protein). You have a daily points budget and manage it however you choose. The programme also includes community support (meetings, app, coaching).
What the science says
- Weight loss: -3 to -5 kg over 12 months in controlled trials. Superior to general dietary advice from GPs.
- Sustainability: Better than most strict diets — the points system offers flexibility, which supports adherence.
- Community: The group support effect is significant. Meeting attendees lose more than app-only users.
The limitations
- Cost: WW subscription is paid (£13-22/month depending on the plan). That's on top of food costs.
- System dependency: Some participants can no longer eat "normally" without counting points — which can perpetuate an anxious relationship with food.
- Weight focus: Despite evolving towards "holistic wellbeing," WW remains fundamentally a weight loss programme — not a global health improvement programme.
The paleo diet: back to basics or marketing gimmick?
The idea is appealing: eat like our hunter-gatherer ancestors, before agriculture and industrial food processing. But reality is more complex than the pitch.
What it is
Meat, fish, fruits, vegetables, nuts, seeds. Excluded: grains, legumes, dairy, refined sugar, industrial oils, processed foods. The central idea: our genome hasn't had time to adapt to foods that appeared with agriculture (~10,000 years ago).
What the science says
- Weight loss: Modest. 2019 meta-analysis: -3.5 kg over 6-12 months vs control diet. Not superior to Mediterranean.
- Blood sugar: Modest improvement in metabolic markers.
- Satiety: Participants report better fullness — probably due to high protein and fibre intake.
The problems
- Questionable scientific basis: The evolutionary hypothesis is simplistic. Humans continued evolving after the Neolithic — adult lactose tolerance in Europeans is one example. Moreover, the "Palaeolithic diet" varied enormously by region and era.
- Unjustified exclusions: Legumes and whole grains are among the most protective foods identified by nutrition research. Excluding them goes against the body of evidence.
- High cost: Lots of meat, lots of fish, no grains or legumes (the cheapest protein sources) = a hefty food bill.
My view: Paleo has a good basic instinct (avoid ultra-processed foods, eat more vegetables) but arrives at the wrong conclusion by excluding protective food groups. If you want to "eat natural," the Mediterranean diet does it better — with an incomparably stronger scientific dossier.
Plant-based eating: ethical AND effective?
Vegan eating (excluding all animal products) is often chosen for ethical or environmental reasons. But does it offer health and weight loss benefits?
What the science says
- Weight loss: Vegans have on average a BMI 1.5-2 points lower than omnivores (observational studies). Controlled trials show -2 to -4 kg over 6-12 months vs omnivore diet.
- Cardiovascular health: LDL cholesterol and blood pressure reduction.
- Diabetes: Improved insulin sensitivity.
- Cancer: Modest reduction in risk of certain cancers (though studies are observational, with significant confounders).
The risks if poorly planned
- Potential deficiencies: Vitamin B12 (supplementation MANDATORY), iron, zinc, calcium, omega-3 DHA/EPA, vitamin D. Without planning, a vegan diet can be nutritionally inadequate.
- Variable quality: Chips, white bread and Oreos are technically vegan. A vegan diet based on ultra-processed foods is probably worse than a balanced omnivore diet.
The consensus: A well-planned vegan diet (with B12 supplementation and monitoring of iron, zinc, calcium intake) is nutritionally adequate and health-beneficial. But the key phrase is "well-planned." If you're considering the transition, get guidance from a registered dietitian — at least for the first few months.
The big comparison table
Here's the comparison I wish I'd found when I first started exploring this subject — based exclusively on meta-analyses and randomised controlled trials.
Weight loss at 6 months:
- Keto: -5 to -7 kg (but significant water loss)
- WW: -4 to -5 kg
- Paleo: -3 to -4 kg
- Vegan: -3 to -4 kg
- Mediterranean: -2 to -3 kg
- DASH: -2 to -3 kg
Maintenance at 2 years:
- Mediterranean: ★★★★★ (best maintenance rate)
- DASH: ★★★★☆
- WW: ★★★☆☆
- Vegan: ★★★☆☆ (if ethically motivated)
- Paleo: ★★☆☆☆
- Keto: ★☆☆☆☆ (highest dropout rate)
Health benefits beyond weight:
- Mediterranean: ★★★★★ (heart, cancer, diabetes, longevity)
- DASH: ★★★★★ (hypertension, heart)
- Vegan: ★★★★☆ (heart, diabetes, if well-planned)
- Keto: ★★★☆☆ (blood sugar, epilepsy)
- Paleo: ★★☆☆☆ (limited data)
- WW: ★★☆☆☆ (weight-focused, not global health)
Sustainability: the only criterion that truly matters
Let's return to Nadia and her stone lost and regained four times. Her problem wasn't lack of willpower — that's a destructive myth we need to bury. Her problem was searching for a diet that "works" instead of searching for a way of eating she can maintain for life.
The 5 characteristics of a sustainable eating pattern
- It doesn't ban entire food groups — radical exclusions are the leading cause of dropout
- It's compatible with your social life — you can eat at restaurants, with family, with friends without stress
- It doesn't require constant calculations — if you need to weigh every food and count every calorie, it's not viable long-term
- It gives you energy, not fatigue — a good eating pattern improves your daily life, it doesn't degrade it
- It doesn't generate guilt — if a deviation makes you feel guilty, the relationship with food is dysfunctional
My clinical recommendation
For the majority of my patients, I recommend a flexible Mediterranean approach: lots of vegetables, good fats (olive oil, nuts, oily fish), varied proteins, whole grains, minimal added sugar and ultra-processed foods. No counting, no bans, no guilt. Results are slower but incomparably more durable — and the benefits for overall health are unmatched.
The grandmother test: If your grandmother wouldn't recognise what you're eating as "real food," it's probably ultra-processed. This simple filter is surprisingly effective for guiding your choices — and it requires no app, no subscription, and no diet book.
Frequently asked questions
Can you lose weight sustainably without a diet?
Yes — and it's actually the only way that works long-term for most people. Gradual changes to eating habits (more vegetables, fewer ultra-processed foods, mindful eating, cooking more) produce slower but infinitely more durable results than any restrictive diet. The goal isn't a temporary "diet" but a permanent lifestyle change.
Is calorie counting necessary?
No, for most people. Calorie counting can be useful short-term to build awareness of intake, but it's neither necessary nor desirable long-term. It can even become obsessive and anxiety-inducing. Approaches based on food quality (eating real foods, listening to hunger and satiety cues) are more effective and psychologically healthier.
Do "fat-burning" supplements work?
No. No supplement sold as a "fat burner" has demonstrated significant efficacy in controlled trials. Caffeine and green tea have a minuscule thermogenic effect (30-50 calories/day — the equivalent of eating a strawberry). Some "burners" contain potentially dangerous substances (ephedrine, sibutramine). It's a market estimated at £2 billion in Europe built on hot air.
Is intermittent fasting a diet?
Technically no — it's a time-based eating pattern (when to eat, not what to eat). But in practice, many people use it to create a caloric deficit by skipping a meal. If you eat the same number of calories in a shorter window, fasting alone won't make you lose weight. It's a tool, not a miracle solution.
Should I consult a professional before starting a diet?
Yes, ideally a registered dietitian (not an unqualified "nutritionist" or "nutrition coach"). A personalised nutritional assessment is worth infinitely more than any generic programme. The NHS provides access to dietetic services through GP referral for conditions like obesity and diabetes. Check with your GP.
Does gluten make you gain weight?
No. Gluten is a protein found in wheat, rye and barley. It doesn't cause weight gain or inflammation (except in people with coeliac disease or non-coeliac gluten sensitivity — roughly 1-6% of the population). Eliminating gluten without medical reason has no proven benefit and can lead to fibre and B vitamin deficiencies. It's one of the most persistent and profitable nutritional myths on the market.