Acne runs on hormones and heredity. Diet is a real signal, but a weak and messy one

Acne is produced by four interacting things happening inside a hair follicle: too much sebum, androgen signalling that drives it, abnormal keratinisation blocking the follicle, and bacterial overgrowth with inflammation behind it [s3]. Diet is not one of the four. It is a modifier, and the evidence that it modifies anything is weaker and more conflicted than either the "chocolate causes spots" folklore or the "diet is a myth" correction suggests.

The strongest risk factors are the ones you cannot change

A systematic review of acne epidemiology pooled the observational literature and found the largest and most consistent associations were with family history, age, body mass index and skin type [s1]. A history of acne in a parent carried a pooled odds ratio of 2.91 (95% confidence interval 2.58 to 3.28) against no parental history [s1]. Overweight or obese BMI carried 2.36 (1.97 to 2.83) against normal or underweight [s1]. Sex barely registered: males against females gave 1.07, with a confidence interval running from 0.42 to 2.71 — an interval so wide it conveys almost nothing [s1].

The same review's summary of diet is the sentence that should anchor any account of this question: the association between dietary factors and acne presentation or severity was less clear, with inconsistent results between studies [s1].

Acne itself is close to universal in adolescence. The Global Burden of Disease Study 2010 put global prevalence across all ages at 9.38%, while estimates for adolescents range from 35% to close to 100% depending on the population studied [s1].

What the diet literature actually contains

The largest single body of evidence concerns dairy. A meta-analysis pooled 14 observational studies covering 78,529 people aged 7 to 30 — 23,046 acne cases and 55,483 controls [s2]. Any dairy intake carried an odds ratio of 1.25 (1.15 to 1.36) against no intake; any milk 1.28 (1.13 to 1.44); yogurt 1.36 (1.05 to 1.77); cheese 1.22 (1.00 to 1.50), with the lower bound sitting exactly on no effect [s2].

The most counterintuitive result is the one that argues against the obvious mechanism. Low-fat or skim milk carried a higher odds ratio, 1.32 (1.16 to 1.52), than whole milk at 1.22 (1.06 to 1.41) [s2]. If dietary fat were the culprit, that ordering would be reversed. The authors' proposed mechanism is instead hormonal: milk-derived amino acids promote insulin secretion and induce hepatic production of insulin-like growth factor 1, and IGF-1 stimulates sebum production and follicular keratinisation, with plasma IGF-1 levels correlating with acne severity [s2].

There is a dose gradient of sorts. Compared with drinking milk less than weekly, two to six glasses a week gave 1.24 (0.95 to 1.62), one glass a day 1.41 (1.05 to 1.90), and two or more glasses a day 1.43 (1.09 to 1.88) [s2].

Two caveats sit inside the same paper and matter more than the point estimates. There was statistically significant publication bias across the included studies [s2]. And adjusted results were attenuated compared with unadjusted ones — meaning that when the studies controlled for confounders, the association shrank [s2]. The authors' own conclusion is that the results should be interpreted with caution because of heterogeneity and bias [s2].

Glycaemic load, and the shape of the trial evidence

A review of the diet–acne literature from 2009 to 2020 identified 53 eligible articles. Eleven were interventional — seven randomised controlled trials and four uncontrolled open-label studies — against 42 observational studies [s4]. That ratio is the field in one number: four out of five studies cannot establish causation by design.

The review's summary is that acne-promoting factors include high glycaemic index and glycaemic load foods, dairy products, fatty food and chocolate, while acne-protective factors include fatty acids, fruit and vegetable intake [s4]. It is candid that the role played by specific components within these food categories — whole against skim milk, cocoa against dark against milk chocolate — remains unsolved [s4].

A 2025 meta-analysis tested the Mediterranean diet, on the theory that if a Western high-glycaemic pattern promotes acne, a non-Western pattern should protect against it. Five case-control studies were found, together enrolling 765 participants — 340 cases and 425 controls [s3]. Higher adherence to the diet was not significantly associated with lower odds of developing acne: an odds ratio of 0.32 with a confidence interval of 0.08 to 1.28, and heterogeneity of 75% [s3]. A separate pooling of correlational studies did find a significant negative correlation between adherence and acne severity, a coefficient of −0.29 (−0.55 to −0.03), with heterogeneity of 79% [s3]. Higher vegetable consumption specifically was associated with lower odds of acne, 0.46 (0.29 to 0.74) [s3].

The authors graded the quality of evidence for both the development and severity findings as low [s3]. Five case-control studies with 765 people between them is not a foundation for a dietary claim, and the paper says so.

What the treatment guideline does with all this

The American Academy of Dermatology's guideline on acne management applied the GRADE framework and issued 18 evidence-based recommendations and five good practice statements [s5]. Strong recommendations went to benzoyl peroxide, topical retinoids, topical antibiotics and oral doxycycline, with oral isotretinoin strongly recommended for acne that is severe, scarring, causing psychosocial burden, or failing standard therapy [s5]. Conditional recommendations covered topical clascoterone, salicylic acid and azelaic acid, and oral minocycline, sarecycline, combined oral contraceptive pills and spironolactone [s5].

Diet does not appear among them. That absence is not the guideline calling the dietary literature wrong; it is the guideline declining to convert associations of this size and quality into clinical instructions.

The honest position

Heredity and hormonal biology explain most of who gets acne. Dairy shows a consistent but small association across a very large observational dataset that carries acknowledged publication bias and shrinks under adjustment. High-glycaemic eating shows a plausible mechanism through insulin and IGF-1, supported by a handful of small trials. Vegetables show a protective association in five case-control studies rated low certainty. Nobody has run the large randomised dietary trial that would settle it, and until someone does, the gap between what is known and what is asserted about acne and food will keep being filled by whoever is selling something.

This article is informational and is not medical advice.

Sources

• [s1] Systematic review of the epidemiology of acne vulgaris — Scientific Reports, 1 April 2020. https://doi.org/10.1038/s41598-020-62715-3
• [s2] Dairy Intake and Acne Vulgaris: A Systematic Review and Meta-Analysis of 78,529 Children, Adolescents, and Young Adults — Nutrients, 9 August 2018. https://doi.org/10.3390/nu10081049
• [s3] The mediterranean diet and acne vulgaris: a systematic review and meta-analysis — Nutrition & Metabolism, 5 November 2025. https://doi.org/10.1186/s12986-025-01033-9
• [s4] Diet and acne: review of the evidence from 2009 to 2020 — International Journal of Dermatology, 18 January 2021. https://doi.org/10.1111/ijd.15390
• [s5] Guidelines of care for the management of acne vulgaris — Journal of the American Academy of Dermatology, 30 January 2024. https://doi.org/10.1016/j.jaad.2023.12.017