Is Sugar More Addictive Than Cocaine? What the Science Actually Says

Sugar is not more addictive than cocaine in any sense that matters to a human being. The comparison comes from rodent studies using intermittent access to sugar solution, and it does not survive translation to people eating food. What is real — and more useful — is that ultra-processed food is genuinely hard to stop eating, for reasons that are about formulation rather than sugar alone.

What the claim actually says

The viral version is usually some form of “sugar is eight times more addictive than cocaine”. That specific figure traces back to animal work in which rats given intermittent access to sugar solution showed behaviours interpreted as bingeing, craving and withdrawal — including self-administration patterns compared with cocaine (Avena, 2008).

Two features of that research do most of the work, and both are worth understanding before you accept the headline. First, the animals were food restricted and given sugar intermittently — a schedule that produces bingeing with almost any palatable substance. Second, the comparison to cocaine was between different experiments with different designs, not a head-to-head equivalence.

Reviews of the field have been consistently more cautious than the headlines. A state-of-the-science review concluded that the evidence for sugar addiction in humans is much weaker than the animal literature suggests, and that the useful construct may be eating behaviour rather than a specific molecule (Westwater, 2016).

The evidence ladder

Claims about sugar addiction are often argued at the wrong level of evidence. This is where each strand of the argument actually sits, from weakest to strongest.

Notice what that ladder does to the debate. The cocaine claim lives at tier two. The defensible claim — that some foods are engineered to be hard to stop eating — lives at tier five, where the evidence is strongest and the policy implications are actually useful.

Where the animal evidence comes from

The intermittent-access model is genuinely interesting neuroscience, and it is worth knowing why it does not transfer.

• The schedule creates the bingeing. Rats with continuous access to sugar do not binge. Rats with intermittent access do. The behaviour comes from the restriction pattern, not the substance alone.
• Food restriction is part of the design. Caloric restriction itself increases reward sensitivity to palatable food. Human dieting does the same thing, which is why restrictive dieting predicts later overeating.
• The withdrawal data are contested. Signs interpreted as withdrawal in animals overlap with the stress of food restriction and the anticipation of access.
• Different studies, different designs. Comparing a sugar self-administration study with a cocaine study is not a controlled comparison, even when the numbers look comparable.

None of this means the neuroscience is wrong. It means it answers a narrower question than the headline implies.

Why the human picture differs

In people, the picture is messier and more interesting. Food addiction as measured by validated scales does identify a real subgroup — roughly one in ten to one in five depending on the population and the cut-off used (de Melo Barros, 2023).

But the substance implicated is not sugar. Reviews of the construct increasingly point at ultra-processed food as the more plausible target, because it combines several reward-relevant properties at once (LaFata, 2024). A clinical review of food addiction sets out how the diagnosis is used and where it remains contested (Krupa, 2024).

What actually drives overeating

Ultra-processed food affects appetite through several mechanisms at once, and that combination is the plausible explanation for why it is hard to moderate.

• Energy density without satiety. Fat and refined carbohydrate together are rarely found in nature and do not trigger the same fullness signals as fibre-rich food.
• Fast consumption. Soft textures mean you eat more before satiety signals arrive, which is the mechanism set out in how carbohydrates work.
• Reward prediction. Consistent flavour intensity keeps the reward system engaged in a way that whole foods do not (Hough, 2026).
• Disrupted homeostatic signalling. Reviews of ultra-processed food describe measurable effects on both hedonic and homeostatic appetite control (Ulug, 2025).
• Availability and cue density. The most boring and most powerful factor. Foods that are visible and cheap get eaten.

Blood glucose handling is a related but distinct issue — the mechanism of the crash-and-crave cycle is covered in blood sugar dysregulation and hidden blood sugar spikes.

What to do if you feel addicted to sugar

The practical advice follows from the reframe above: change the environment and the formulation, not your willpower.

• Stop restricting drastically. Restriction increases the reward value of the food you are avoiding, which is the same mechanism that produced the animal bingeing. See breaking a weight-loss plateau for the sustainable version.
• Eat protein and fibre at every meal. Both blunt the reward and the glucose response — see anti-inflammatory foods.
• Change what is in the house. This is more effective than any resolve formed in a supermarket.
• Do not buy the framing. “Sugar detox” products are a marketing category; the substance is not a toxin and does not require detoxification.
• Sleep and stress are part of this. Short sleep raises appetite and reward sensitivity — see sleep and weight and improving sleep.
• Ask for help if it is genuinely compulsive. Binge eating is a recognised condition with effective treatment, and it is not a willpower problem.

On the specific question of sugar versus other sugars, the distinction between added sugar and fruit sugar is real and covered in fruit sugar versus added sugar. What happens to glucose overnight is in the dawn phenomenon, and the medication side is in the GLP-1 review. Dietary patterns with outcome evidence are in the Mediterranean guide.

The dopamine argument, examined

The mechanism usually invoked is dopamine. Eating palatable food releases dopamine in the nucleus accumbens, drugs of abuse release more, therefore food is a drug. Each step in that chain needs examining, because the conclusion does not follow from the premises.

• Dopamine signals prediction, not pleasure. The modern understanding is that dopamine encodes reward prediction error — the gap between what you expected and what you got. Food produces a signal because it is a reward, which is true of every rewarding thing including water when you are thirsty.
• Magnitude is not the whole story. Drugs of abuse produce larger and more sustained dopamine responses than food, and they do so regardless of whether the person is hungry. Food responses depend heavily on metabolic state.
• Sensitisation is the addiction signature. Repeated drug exposure tends to sensitise the system, escalating the response. Repeated food exposure does not do this in the same way — if anything, repeated exposure to a specific food reduces its reward value, which is why variety matters.
• Human imaging shows overlap, not identity. Obese and non-obese people differ in reward responses to food cues, but the direction is inconsistent across studies and depends on whether you are looking at anticipation or consumption (Val-Laillet, 2015).
• The withdrawal story is weakest. Genuine substance withdrawal is physiologically dangerous for alcohol, benzodiazepines and opioids. Sugar “withdrawal” in humans is irritability and craving for a few days, which is the profile of a habit rather than a dependence.

The honest summary: food engages reward circuitry because it must, in order to keep you alive. Drugs of abuse hijack that circuitry. Both statements are true, and they are not the same statement.

What the sugar-detox industry gets wrong

Because the addiction framing is vivid, it has produced a commercial ecosystem of detoxes, resets and sugar-free challenges. Almost all of it rests on the framing rather than on the evidence.

• There is nothing to detoxify. Sugar is metabolised to glucose and used for energy. It does not accumulate as a toxin, and there is no withdrawal syndrome requiring medical clearance.
• Elimination tends to backfire. Restriction increases the reward value of the avoided food, which is exactly the mechanism the animal research demonstrates. The binge-restrict cycle is a well-recognised clinical pattern.
• Natural sugar is not different in kind. Honey, maple syrup and coconut sugar are sucrose and fructose with trace minerals. Agave is largely fructose. The physiology is the same.
• The real target is formulation. Reducing ultra-processed food reduces the reward-relevant combination without requiring you to police individual ingredients.
• Cutting fruit is not the point. Whole fruit is difficult to overeat and consistently associated with better outcomes — see fruit sugar versus added sugar.
• Shame is not a strategy. Framing overeating as an addiction can help some people and harm others by making lapses feel like relapse. If it is genuinely compulsive, treatment exists and it is not a detox.

The useful test for any sugar product: does it change what is in your kitchen, or only what is in your head? The first tends to work. The second tends to be sold twice.

Common questions about sugar and addiction