In searching for the most impactful paper of 2015, PL went to the most highly cited general psychiatry journal, JAMA Psychiatry (formerly Archives of General Psychiatry). There one finds a list of the most widely read articles, among which the top article listed was this one, which had just been published online 6 days before PL found it. In about a week, this paper had already been read more frequently than other papers on in JAMA Psychiatry from earlier in the year.
This high level of attention led PL to choose it as the top paper of the year, not because we think it is the best or the most creative or impactful for the long run, but because it has had an impact. Thus, PL wants to bring the paper to the attention of our readers.
This study was headed by many of the same researchers (like Ronald Kessler PhD) who ran classic epidemiological studies in psychiatry, such as the National Comorbidity Survey (NCS). In this study, researchers in 17 countries conducted epidemiological surveys of the frequencies of psychiatric conditions, using DSM-IV diagnoses, in 47,609 people, and they did so repeatedly 18 times over a decade (2001-2011).
They then asked the persons who were interviewed about whether they had been diagnosed with a list of medical illnesses, and at what age they had been so diagnosed or had first experienced symptoms. This analysis describes the frequency of medical illnesses occurring AFTER onset of psychiatric diagnoses.
The key limitation to the study is that medical diagnoses are based completely on the self-report of patients, and were not independently verified or based on medical chart evidence. Another limitation, present with all such observational data, is that other confounding factors could exist and explain the association of mental and physical illnesses.
With those caveats in mind, using statistical regression modeling, the researchers found that mood conditions, anxiety conditions, and substance abuse each could explain about 3-13% of cases of a range of physical illnesss, such as cancer, diabetes, chronic lung disease, heart disease, and asthma.
When any psychiatric diagnosis was present, the risk of cancer was increased by an odds of 30% (odds ratio 1.3, 95% confidence intervals 1.1-1.5). The same was the case for diabetes. Heart disease risk was increased 70%, and chronic lung disease risk (presumably secondary to mediators like cigarette smoking) was doubled.
Many physicians focus on “physical” illnesses as if “mental” illnesses are unrelated. This report, within the constraints of its limitations as an observational study, reverses this discriminatory belief: If you want to control physical diseases better, diagnose and treat psychiatric conditions better.