The Reece Critique

A critique of Reece & Hulse (2022) and the record of its lead author
Prepared for K I 49098/2020

I. The Study and Its Methodological Failures

The claim

Reece & Hulse published a three-part paper in Archives of Public Health (2022) titled "Geotemporospatial and causal inferential epidemiological overview and survey of USA cannabis, cannabidiol and cannabinoid genotoxicity expressed in cancer incidence 2003-2017." [6230] [6231] [6232] The paper claimed to show that cannabis use is causally associated with elevated rates of several cancers, including liver cancer (HCC).

The design: ecological correlation

The study correlated state-level estimates of cannabis use prevalence (drawn from the US National Survey on Drug Use and Health) with state-level age-standardised cancer rates, across 51,623,922 recorded cancer cases from 2003 to 2017. It did not examine individual cannabis users. It compared aggregate statistics at the state level and inferred from those correlations that individual cannabis exposure causes cancer.

Flaw 1: The ecological fallacy

This is the central and fatal error. Correlating group-level data to draw conclusions about individual causation is a textbook methodological error known as the ecological fallacy. A state with high cannabis use rates may differ from a low-use state in dozens of confounding ways -- hepatitis C prevalence, alcohol consumption, obesity rates, age structure, ethnic composition, healthcare access, smoking rates -- none of which are adequately controlled in this design. Chen, Barnes & Lewis (Arch Public Health, July 2022) [6215] published a formal letter to the editor making precisely this point: "The authors' conclusion of a causal association is limited due to biases inherent in ecological epidemiological studies. Though the researchers attempt to overcome these biases through validation and statistical manipulations, their approaches are insufficient to create conditions suitable for causal inferencing upon examination."

Flaw 2: Massive multiple comparisons

The study tested 19,877 age-standardised cancer rates across numerous cancer types, cannabinoid subtypes, and demographic sub-groups. With this volume of tests, statistically significant associations will arise by chance alone. No credible correction for multiple comparisons is applied.

Flaw 3: Exposure measurement is a proxy for a proxy

Cannabis "exposure" is measured via the National Survey on Drug Use and Health -- a self-report survey of cannabis use prevalence by state. This is not a measure of the cannabis consumption of the individuals who developed cancer. It is a population-level estimate used as a stand-in for individual exposure, compounding the ecological problem.

Flaw 4: Directly contradicted by individual-level evidence

The study's conclusion that cannabis is associated with elevated HCC is directly contradicted by the largest individual-level study available. ElTelbany A, Khoudari G, Al-Khadra Y, McCullough A & Alkhouri N [1735] (Cureus, 2022) analysed 101,231,036 patients in the US National Inpatient Sample and found cannabis users were 55% less likely to have HCC (adjusted OR 0.45, 95% CI 0.42-0.49, p<0.001), after adjusting for hepatitis B, hepatitis C, cirrhosis, smoking and obesity. The Reece ecological finding cannot survive comparison with this.

Flaw 5: Contradicted by established mechanism

CB1 and CB2 receptor overexpression in HCC correlates with improved prognosis (Perez-Gomez et al., 2007). [6224] THC and CB2-selective agonists reduce HCC cell viability via AMPK-dependent autophagy (Vara et al., Cell Death & Differentiation, 2011). [6216] A 2025 case series in the Journal of Cannabis Research reports durable complete response in two advanced HCC patients using cannabis oil. The mechanistic and clinical case literature runs entirely against Reece's ecological association.

Flaw 6: Undisclosed conflicts of interest

Reece declares no conflict of interest. However, he is a longstanding adviser to Drug Free Australia (DFA), a religious abstinence lobby that receives government funding and campaigns against harm reduction, medical cannabis, and drug law reform. His entire academic output on cannabis is aligned with DFA advocacy positions. The journal of choice for much of his earlier work is the Journal of Global Drug Policy and Practice, a non-peer-reviewed publication operated by DFA-aligned organisations.

II. The Record of Dr. Albert Stuart Reece

Religious advocacy as clinical practice

Reece is an evangelical Christian who wrote "Let My People Go: A Theology of Addiction" (2016), [6228] which offers biblical scripture as a treatment framework for drug addiction. The publisher's description states that Reece "placed into the hands of the church the wisdom of the scriptures which he has used with countless patients over years of clinical work." He describes himself as offering "biblically driven abstinence." This is not incidental to his scientific output. It is the operating framework. Every study he produces on cannabis, harm reduction, or drug policy arrives at a conclusion that supports prohibition and abstinence. The research is reverse-engineered from a predetermined theological position.

The Parliamentary testimony: Sodom and snakes

In April 2007, Reece appeared before the Australian Parliament's House Standing Committee on Family and Human Services, chaired by Bronwyn Bishop, to testify on illicit drug policy. He introduced himself by saying "I certainly know the science." He then displayed two images as evidence for his submission: a photograph described as "the archaeological site of Sodom" and a drawing of a tree with snakes in place of branches. His argument was that harm reduction policies -- including condoms, clean needles, methadone, non-punitive cannabis laws -- would lead Australia to the fate of Sodom. His stated conclusion: "There will be consequences." He also testified that condoms are "the real cause behind AIDS" and attacked internationally respected public health scientists Dr. Alex Wodak and Professor Wayne Hall, accusing Hall of "scientific fraud" under the protection of parliamentary privilege. Ray Moynihan reported on this testimony in Crikey under the headline: "Naltrexone II: no trials, just the power of prayer." [6227]

Twenty-five patient deaths in twenty months

Between 1999 and 2001, Reece operated a private clinic in Queensland prescribing rapid naltrexone detoxification for heroin dependence. Over twenty-five patients died in connection with his treatment in a twenty-month period. The Queensland Health Practitioners Tribunal opened proceedings. The case was adjourned indefinitely. Reece subsequently claimed, repeatedly and before Parliament, that he holds "the world safety record" in naltrexone prescription. He attributed the patient deaths to a government conspiracy. In September 1999, after being raided and threatened with closure, Reece declared that "the Queensland government will have blood on its hands" if he were shut down. Twenty months after getting exactly what he wanted, twenty-five people were dead.

Medical Board of Queensland: 2009 suspension

In 2009, the Medical Board of Queensland suspended Reece from practice. The grounds were supplying morphine to opiate-dependent patients and falsifying records to disguise this. The Medical Board found: "He has a somewhat evangelical approach to this area of medicine and because of that he does appear to lack a degree of insight and objectivity in relation to the treatment of his patients. Furthermore, he seems to feel that the ends justify the means in terms of treatment of patients."

Reducing medication if patients did not attend church

Patient testimony published online (and reported by Paul Gallagher in 2011 [6225] [6226]) includes the account of a woman in early pregnancy receiving Suboxone under Reece's supervision. According to this account, Reece reduced her medication dosage when she had not attended church. She reported being "scared that she could miscarry if this idiot messes around with her medication to suit his pathological world view."

Drug Free Australia and the funding pipeline

In October 2005, then Federal Health Minister Tony Abbott (Liberal Party) awarded Drug Free Australia $600,000 in government funding. The subsequent review found DFA had breached almost every condition of the grant, operating as a right-wing religious lobby rather than a treatment or research body. Reece is Drug Free Australia's scientific face. [6229] His papers appear in DFA parliamentary submissions and media replies. The organisation promotes abstinence, zero-tolerance enforcement, opposition to harm reduction, and the criminalisation of cannabis users.

Coined terminology

Reece coined the term "hedonistic delirium" to describe modern liberal democracies. He has argued that "sometimes the best way forward is the way back," meaning a return to pre-industrial social conditions. He described the advertising of women's nudity outside of a "strictly medical context" as "incredibly powerful pornography."

III. Summary Assessment

Reece's geotemporospatial study on cannabis and cancer is an ecological correlation study that cannot establish individual-level causation, suffers from massive multiple testing, uses a proxy-for-a-proxy exposure measure, and arrives at a conclusion directly contradicted by the largest individual-level study in the literature (ElTelbany et al. [1735], n=101 million) and by established receptor-level mechanism. It has been formally criticised in the same journal. Its lead author is a suspended-then-reinstated general practitioner with a documented evangelical approach to medicine that the Medical Board of Queensland found compromised his clinical objectivity, a record of 25+ patient deaths in twenty months, a parliamentary appearance featuring photographs of Sodom and snake-filled trees, a theological textbook prescribing scripture for addiction, and a career-long affiliation with a religious prohibition lobby funded by government money. His conclusion that cannabis elevates liver cancer risk should be weighed accordingly.

References

ElTelbany A, Khoudari G, Al-Khadra Y, McCullough A, Alkhouri N -- Lower Rates of Hepatocellular Carcinoma Observed Among Cannabis Users: A Population-Based Study. Cureus 2022;14(7):e26850. NIS 2002-2014, n=101,231,036
https://pmc.ncbi.nlm.nih.gov/articles/PMC9138632/ [1735]

Chen W, Barnes C, Lewis D -- Letter re: Reece & Hulse geotemporospatial study [methodological critique]. Arch Public Health 2022;80:160
https://pubmed.ncbi.nlm.nih.gov/35850921/ [6215]

Vara D et al. -- Anti-tumoral action of cannabinoids on hepatocellular carcinoma: role of AMPK-dependent activation of autophagy. Cell Death Differ 2011;18:1099-1111
https://www.nature.com/articles/cdd2011.31 [6216]

Perez-Gomez E et al. -- Overexpression of type 1 and type 2 cannabinoid receptors in human HCC: correlation with poor prognosis. Cancer Letters 2007;249(2):246-255
https://pubmed.ncbi.nlm.nih.gov/17074588/ [6224]

Gallagher P -- Drug Free Australia's Dr. Stuart Reece & Culpable Abuse of Patients. The Australian Heroin Diaries, 2011
http://theaustralianheroindiaries.blogspot.com/2011/05/drug-free-australias-dr-stuart-reece.html [6225]

Losing In The Lucky Country -- Dr. Stuart Reece: Drug Free Australia's shameful secret. 2011
https://luckylosing.com/2011/12/30/dr-stuart-reece-drug-free-australias-shameful-secret/ [6226]

Moynihan R -- Naltrexone II: no trials, just the power of prayer. Crikey, 15 May 2007
https://www.crikey.com.au/2007/05/15/naltrexone-ii-no-trials-just-the-power-of-prayer/ [6227]

Reece AS -- Let My People Go: A Theology of Addiction. WestBow Press, 2016
https://www.amazon.com/Let-My-People-Theology-Addiction/dp/1512744670 [6228]

MaterCare International -- MCI Welcomes Dr. Albert Stuart Reece
https://www.matercare.org/post/mci-welcomes-dr-albert-stuart-reece [6229]

Reece AS, Hulse GK -- Geotemporospatial and causal inferential epidemiological overview and survey of USA cannabis, cannabidiol and cannabinoid genotoxicity expressed in cancer incidence 2003-2017: part 1 overview. Arch Public Health 2022;80:100
https://link.springer.com/article/10.1186/s13690-022-00811-8 [6230]

Reece AS, Hulse GK -- Geotemporospatial and causal inferential epidemiological overview and survey of USA cannabis, cannabidiol and cannabinoid genotoxicity expressed in cancer incidence 2003-2017: part 2 categorical bivariate analysis and attributable fractions. Arch Public Health 2022;80:101
https://link.springer.com/article/10.1186/s13690-022-00812-7 [6231]

Reece AS, Hulse GK -- Geotemporospatial and causal inferential epidemiological overview and survey of USA cannabis, cannabidiol and cannabinoid genotoxicity expressed in cancer incidence 2003-2017: part 3 spatiotemporal, multivariable and causal inferential pathfinding. Arch Public Health 2022;80:102
https://pubmed.ncbi.nlm.nih.gov/35354499/ [6232]