Is Food Addiction Real? 19.9% by One Measure

A widely cited review found addictive-like eating symptoms in roughly one in five people screened with a validated tool. That is a real, replicated number. Whether it proves food is addictive the way a drug is addictive is a separate question the number does not settle on its own.

Here is what the evidence actually supports, and where it stops.

The short answer

A 2014 systematic review in Nutrients pooled 25 studies covering 196,211 people who completed the Yale Food Addiction Scale (YFAS), a 25 item questionnaire built by mapping the DSM criteria for substance dependence onto eating. The weighted mean prevalence of a food addiction diagnosis by that scale was 19.9 percent.

The number moved a lot by who was screened. People with overweight or obesity scored positive 24.9 percent of the time versus 11.1 percent among people at a healthy weight, and people in clinical eating disorder samples scored positive 57.6 percent of the time. Women screened positive more than twice as often as men, 12.2 percent versus 6.4 percent.

What the scale actually asks

The YFAS does not ask whether you like chips. It asks about loss of control over amount eaten, repeated unsuccessful attempts to cut down, continuing to eat a food despite knowing it causes a physical or emotional problem, and needing more of a food over time to get the same satisfaction, the same structure used to diagnose substance use disorders.

That borrowed structure is exactly what the debate below turns on. A questionnaire built from addiction criteria will, by construction, find some people who match those criteria. Whether matching the criteria means the same underlying process as a drug addiction is a different claim entirely.

Food addiction, or eating addiction

A 2014 paper in Neuroscience and Biobehavioral Reviews, co-authored by obesity researcher Johannes Hebebrand, argued the field has the label backwards. Their case: no single food or nutrient has been shown to be addictive on its own the way nicotine or alcohol is, so calling it food addiction implies a substance claim the evidence does not support.

Their proposed alternative is eating addiction, meaning the behaviour itself, the act of eating in a particular pattern, is what resembles an addictive process, not any specific food. It is a narrower and more defensible claim, and it changes what a high score should mean to you: not "this food is chemically addictive" but "this pattern of eating has some of the same behavioural features as one."

Who scores higher, and why that matters

The 2014 review's subgroup numbers are worth sitting with rather than skimming past. Prevalence more than doubled between people at a healthy weight (11.1 percent) and people with overweight or obesity (24.9 percent), and it more than doubled again between the general adult samples and clinical eating disorder samples (57.6 percent). Women screened positive at roughly twice the rate of men, 12.2 percent versus 6.4 percent.

That gradient tells you the scale is not picking up a random, evenly distributed trait. It is tracking something that clusters with weight status, clinical eating disorder history, and sex, which is exactly what you would expect if the questionnaire is measuring a real, if contested, pattern rather than noise.

What a high score does not mean

A positive YFAS score is not a clinical diagnosis. It has no entry in the DSM-5 or ICD-11, and researchers who study it disagree about whether it identifies a distinct condition or an overlapping description of binge eating, restrained eating, and ordinary cravings measured with different words.

It also is not a verdict on any single food. The foods most associated with high YFAS scores share an engineered combination, refined carbohydrate paired with fat in a ratio almost never found in whole foods, chips and candy rather than an apple or a chicken breast. That points at manufactured food environments, not at willpower, and it is a large part of why Hebebrand's group pushed back on the substance framing in the first place: no one is proposing that potatoes are addictive, only that a specific industrial combination of ingredients produces a specific behavioural pattern in some people.

Put together, the honest reading of a high score is closer to "this pattern of eating shares real features with an addictive process, in a population where it is more common than most people assume" than to "you are addicted to food" as a settled medical fact.

How OffRamp treats this without diagnosing anything

OffRamp is not a substitute for the YFAS or for a clinician, and it never tells anyone their eating is addictive. What it does track is far more modest: a daily check-in where you rate food noise, appetite, and energy on a plain 1 to 5 scale.

The insights engine behind that check-in has a deliberate floor. A week over week move in your average food noise rating has to reach at least half a point before the app will say anything about a trend at all. Smaller swings, the ordinary noise of a hard day or a good one, are treated as noise and stay quiet. That threshold exists because a tool that reacts to every daily fluctuation teaches you to distrust your own baseline, which is close to the opposite of what someone working through a difficult relationship with food needs.

If food noise or eating patterns feel like they cross into something you cannot manage on your own, that is worth raising with a therapist, registered dietitian, or your prescriber. OffRamp does not provide medical advice, diagnosis, or dosing guidance, and a trend line on a phone is not a substitute for that conversation.