Bad Science

Ben Goldacre, a British doctor and science journalist who writes the "Bad Science" column in The Guardian, presents a systematic critique of how scientific evidence is misunderstood, distorted, and exploited across a wide spectrum, from alternative therapists to the pharmaceutical industry to the mainstream media. The book follows what Goldacre calls a "natural crescendo" (xv), beginning with transparently absurd pseudoscience and escalating to cases in which the misuse of evidence has caused widespread suffering and death.

Goldacre opens by dismantling pseudoscientific products to teach the basics of experimental method. He demonstrates that the Aqua Detox footbath, which claims to extract toxins through the feet, simply produces brown water from electrolysis rusting its iron electrodes; when he has the water tested at a London hospital, no toxins are found. He debunks Hopi ear candles and detox foot patches using similarly straightforward tests and introduces a concept he calls the "hassle barrier" (11): Manufacturers produce impressive-looking documents full of graphs and scientific language but omit key details, creating superficially plausible barriers to discourage scrutiny. He then critiques Brain Gym, a proprietary exercise program used in hundreds of U.K. schools, whose materials contain transparently false claims such as that a special yawn increases "oxidation" (17), confusing it with oxygenation. Citing a 2008 study from the Journal of Cognitive Neuroscience, he shows that people find explanations more satisfying when they include logically irrelevant neuroscience jargon, a trick exploited throughout pseudoscience. He similarly critiques the cosmetics industry, arguing that all moisturizers work by hydrating the skin and that expensive ingredients are either diluted to ineffective concentrations or implausible given the skin's impermeability.

Goldacre uses homeopathy as a teaching device for evidence-based medicine. He explains the practice's origins with the 18th-century German doctor Samuel Hahnemann, who devised the principles of "like cures like" and serial dilution. A standard 30C homeopathic remedy is diluted 1 in 10 to the power of 60, far beyond the point where any molecules of the original substance remain. He explains blinding, in which neither patients nor researchers know who received the real treatment, and randomization, the random assignment of patients to groups, citing reviews showing that inadequate blinding exaggerates treatment benefits by 17 percent and poor randomization inflates results by up to 41 percent. He reviews systematic analyses showing an almost linear relationship between study quality and outcome: The worse the methodology, the more positive the result for homeopathy. He introduces meta-analysis, the pooling of data from many trials, and describes how the major meta-analyses of homeopathy conclude that it performs no better than placebo.

A chapter on the placebo effect argues that it represents the true story of mind-body healing. Goldacre surveys research showing that four sugar pills heal more ulcers than two, that saltwater injections outperform sugar pills for pain, that branded packaging enhances the effect of both aspirin and placebo, and that sham surgeries produce real improvements. He describes experiments showing that what doctors say and believe affects patient outcomes even when doctors are forbidden from communicating their beliefs, and raises ethical questions about the tension between informed consent and the therapeutic power of reassurance.

Goldacre identifies four errors that nutritionists and journalists repeatedly make: citing nonexistent studies, confusing correlation with causation, extrapolating from laboratory experiments on cells to human health, and cherry-picking supportive evidence while ignoring contradictory findings. He applies these errors to the antioxidant hypothesis, tracing how promising early data led to large-scale trials with devastating results. A Finnish trial of 30,000 participants found that beta-carotene supplements increased lung cancer and deaths, and the CARET trial of 18,000 was terminated early because subjects on antioxidants were 46 percent more likely to die of lung cancer. The most comprehensive review by the Cochrane Collaboration, an international organization that produces rigorous systematic reviews of medical evidence, found antioxidant supplements either ineffective or possibly harmful across 230,000 participants. Goldacre argues that the food supplement industry uses doubt-manufacturing tactics borrowed from the tobacco industry to discredit these findings.

As a case study in nutritionist pseudoscience, Goldacre profiles Gillian McKeith, a prominent U.K. television personality whose published work contains basic scientific errors and whose Ph.D. comes from a nonaccredited correspondence course. The book then takes its darkest turn with the story of Matthias Rath, a vitamin pill entrepreneur who promoted his products as a cure for AIDS in South Africa while denouncing antiretroviral drugs as poisonous. This chapter was absent from the original British edition because Rath sued Goldacre and The Guardian for libel. Rath's message found a receptive audience in the government of President Thabo Mbeki, who refused to roll out treatment programs, accept free drug donations, or apply for Global Fund grants. One study estimates that approximately 343,000 deaths could have been prevented between 1999 and 2007 had the government matched the Western Cape Province's antiretroviral rollout. Goldacre notes that not one alternative therapist or nutritionist anywhere in the world publicly criticized Rath's activities.

Goldacre then applies the same critical tools to the pharmaceutical industry. He details tricks companies use to bias trial results: studying drugs in younger, healthier patients; comparing against placebo rather than the best existing treatment; using surrogate outcomes instead of real-world endpoints like death; and manipulating statistical analyses. He explains publication bias, the tendency for positive trials to be published while negative results go unreported, and recounts the Vioxx scandal, in which the painkiller's manufacturer, Merck, understated cardiovascular risk. The FDA estimated the drug caused between 88,000 and 139,000 heart attacks during its five years on the market. Goldacre proposes a properly enforced public clinical trials register as a solution.

A chapter on cognitive psychology surveys how humans systematically err: perceiving patterns in randomness, mistaking regression to the mean for causation, seeking confirmatory evidence, and overweighting dramatic information. Goldacre describes Solomon Asch's classic conformity experiments, in which subjects went along with a group's clearly wrong answers, illustrating how social reinforcement compounds individual bias. He then examines how statistics are misused, showing how the same data can sound terrifying as a relative risk increase or unremarkable as a natural frequency. He explains the prosecutor's fallacy, the error of confusing the probability of the evidence given innocence with the probability of innocence given the evidence, through the case of Sally Clark, a British lawyer convicted of murdering her two babies based partly on a claim that double sudden infant death syndrome (SIDS, the unexplained death of an apparently healthy infant) was a "one in seventy-three million" event. This figure assumed the two deaths were independent, ignoring shared genetic and environmental factors; one paper estimates the actual odds at roughly two to one in favor of double SIDS over double murder.

The book's final major chapter traces the British media's nine-year scare over the MMR (measles, mumps, and rubella) vaccine. Andrew Wakefield's 1998 Lancet paper, a case series of just 12 children, could by design establish nothing about a causal link between the vaccine and autism. Journalist Brian Deer uncovered that Wakefield was paid legal aid money to investigate children whose parents were already suing vaccine manufacturers. The General Medical Council found him "misleading," "dishonest," and "irresponsible" (215), and The Lancet fully retracted the paper. Goldacre documents how the media systematically ignored reassuring epidemiological studies, including a Danish cohort of over 537,000 children that found no link between MMR and autism. Vaccination rates fell from 92 percent to 73 percent nationally, and England and Wales saw record measles cases and a mumps epidemic.

In his conclusion, Goldacre argues that the greatest harm from bad science is the cumulative "opportunity cost of bullshit": Obsession with dietary supplements distracts from genuine social determinants of health, political energy about pharmaceutical abuses is channeled into conspiracy theories rather than meaningful reform, and the public's ability to evaluate evidence is systematically undermined. He calls on scientists to engage directly with the public and contends that people are fully capable of understanding evidence when it is clearly explained.