|
Abstract
|
Dear editors,
RE: Association of cannabis use with depression among cancer patients (Hu S et al., Prev Med Rep 2023; 35: 102305).
We read with strong interest Hu et al.’s (Hu et al., 2023), recent report examining the association of cannabis use with depression among individuals with a history of cancer in a secondary analysis of the US Centers for Disease Control and Prevention’s National Health and Nutrition Examination Survey (NHANES). In it, the authors report a moderately strong, positive correlation between marijuana use and “risk of depression”. Completely absent from the report are three critically important factors necessary for interpretation:
First, the authors in no uncertain terms, misuse the term “risk”, which means incidence, a measure that can only be obtained from a prospective analysis of the same individuals over time. Despite their utilization of multiple years of the NHANES survey, it’s important for readers to understand that the same individuals are not sampled year after year. That is, the study described is not a prospective design and incidence can neither be measured nor inferred (Lash et al., 2021). What results is an egregious overinterpretation of cross-sectional data.
Second, the authors fail to identify that the individuals who report a history of cancer within NHANES are not necessarily cancer patients (despite >80 uses of the term), and more likely a mixture of cancer patients (i.e., actively receiving treatment) and survivors (i.e., post-treatment) that is mathematically much more likely to be the latter. An examination of the authors’ materials and methods supplement does not indicate that NHANES participants actively undergoing treatment or who were recently diagnosed were separated from those with a history of cancer. This is an important distinction that has a significant impact on the interpretation of the authors’ findings.
Third, the authors utilized PHQ-9 scores >4 to indicate presence of depression. While it is widely understood that the PHQ-9 can act as a screener for depressive symptomology, a cutoff of 4 is inconsistent with depression; indeed, this is noted in the reference cited by the authors (Kroenke et al., 2001). A clear discussion of these limitations was lacking in this manuscript, which unfortunately appeared to lack any discussion of the study’s limitations whatsoever. Such introspection is necessary to allow for a nuanced interpretation of the study’s results in the context of its methods.
Sleep, pain, and emotional distress are all major reasons for cannabis-based product use among cancer patients, among who most report moderate to high perceived benefit (Brasky et al., 2023). The authors write, “Our study also draws conclusions consistent with previous studies that the use of cannabis in cancer patients may induce depression.” Unfortunately, the study’s underlying design simply has no capacity to test this hypothesis. We are left to wonder if the authors had considered interpreting their findings to imply that individuals with a history of cancer and who experienced depressive symptomology were at increased odds of using marijuana products, as others have done (Donovan et al., 2019), rather than using causal and politically charged language to suggest that use of marijuana products were “risk factors” for depression among such individuals?
Language: en |