The adjusted ORs represent the odds of each category with all else in the model being equal

These medical diseases are frequently considered chronic medical conditions, and are included in either the validated Katz chronic disease comorbidity questionnaire 15 or the Hierarchical Condition Category system and obstructive sleep apnea, which is recognized as one of the most prevalent chronic respiratory disorders and therefore included in our analysis. Medical multi-morbidity is defined in the literature as 2 or more chronic diseases . Using the above conditions to study medical multi-morbidity with the NSDUH has been performed in other studies . Drug use was assessed by NSDUH by self-report of cannabis , cocaine , heroin, inhalants, hallucinogens, and non-medical use of prescription medications . Non-medical use of prescription medications was defined as use of a drug that was not prescribed or used for the experience or feeling it caused. Nicotine dependence was defined based on dependence criteria of the Nicotine Dependence Syndrome Scale and alcohol dependence was defined by the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition . SUDs were determined by participant responses to a series of questions that determined if criteria would meet DSM-IV abuse or dependence categories for each drug. While NSDUH is not a diagnostic interview, this method provided a proxy diagnosis.Analyses include all respondents aged 18 and older. We aggregated data from the three separate cohort years into a single cross-section to be able to increase power for examining associations between variables with low prevalence . We first examined bivariable associations between demographic characteristics and number of self-reported conditions . Demographic measures included age , sex, race/ethnicity , education level ,roll bench total family income , marital status . NSDUH only provides pre-coded categorical responses for the age and income variables, and therefore could not be analyzed as a continuous variable.

Tobacco use , nicotine dependence, alcohol use , alcohol dependence, self-reported overall health , and self-reported mental health problems in the past-year were also compared among adults reporting 0, 1, and ≥2 chronic conditions. Bivariable analyses was also performed comparing past-year drug use and diagnosed substance use disorder for adults with 0, 1, and ≥2 chronic conditions. We evaluated correlates of past-year drug use among adults with medical multi-morbidity using binary logistic regression. We first estimated odds of each covariate separately, generating unadjusted odds ratios. We then fit covariates simultaneously using multiple logistic regression. All analyses were weighted as part of NSDUH to account for the complex survey design and to obtain accurate standard errors for estimates at the population level. More detailed information regarding the development of analysis weights are found elsewhere . Since our analyses utilized data from 3 cohorts, we divided the weights by 3 to obtain nationally representative estimates. Stata SE 13 was used for all analyses, and survey commands were utilized to provide accurate standard errors using Taylor series estimation methods . Secondary analysis of this publically available data was exempt for review by the New York University Langone Medical Center Institutional Review Board.The analytic sample included 115,335 respondents. Chronic medical conditions were common among respondents with weighted percentages of 29.5% of adults reporting 1 chronic condition and 17.4% reporting 2 or more conditions, and therefore 46.9% reported at least 1 chronic condition and 53.1% reported no chronic conditions . Most adults with medical multi-morbidity were middle-aged and older adults , while younger adults were more likely to report no chronic conditions . Compared with adults reporting no chronic conditions, adults with medical multi-morbidity were also more likely to be non-Hispanic white, less educated, married, have nicotine dependence, drink alcohol less frequently, report having worse health status, and more likely to have depression, anxiety, and need mental health treatment in the past year . Table 2 presents frequencies and weighted percentages of specific drugs used by 0, 1 or ≥2 chronic conditions. Overall, 16.1% of the full study sample reported past-year use of an illegal drug or cannabis and 2.6% reported meeting criteria for a SUD. The most common drug used in the past-year was cannabis, and cannabis use disorder was the most common SUD.

Non-medical use of opioid analgesics was the next-most common and opioid-related SUD was the second most common SUD. Among all adults with chronic conditions, past-year drug use was reported by 14.8% with 1 chronic condition and 11.6% with ≥2 conditions, and drug use was considerably higher among adults with no chronic conditions . Among adults reporting no chronic conditions, 5.5% used ≥2 drugs, while 4.3% of adults with 1 chronic disease and 3.2% of adults with ≥2 chronic diseases reported use of > 1 drug in the past year . Criteria for past-year SUD was met by 2.4% with 1 chronic condition and 2.1% with ≥2 conditions, and 2.9% of adults with no chronic conditions . In this large, nationally representative survey, we estimated the prevalence of past-year drug use among adults without chronic medical conditions and among those with one and two or more chronic medical conditions . While the prevalence of past year drug use was lower among adults with medical multi-morbidity, compared to adults with no chronic conditions, nearly 12% of adults with multi-morbidity reported engaging in past-year drug use. The lower prevalence of drug use among adults with multi-morbidity may be due to both the fact that younger adults were more likely to engage in drug use and less likely to have chronic medical conditions, and some of those with multi-morbidity may have stopped using drugs because of their morbidities . In alcohol studies for example, the “sick quitter” hypothesis proposes that adults may stop drinking due to medical illness, hospitalizations, or declining health, and therefore this group is not included as individuals with alcohol-related problems even though alcohol may have contributed to their illnesses . A similar phenomenon is likely occurring in observational studies with drug use including our study. However, we did find in multi-variable models that among adults with medical multi-morbidity, adults with alcohol dependence, current tobacco use, and adults with mental health problems were more likely to have engaged in past year drug use,drying rack cannabis and therefore be at particularly high risk for adverse effects of drug use. This emphasizes the importance of including SUDs and poly substance use disorders to the multi-morbidity framework as a distinct clinical profile that necessitates further research to better care for patients with complex multi-morbid disease . The intersection of substance use and medical multi-morbidity is complex. Drug use has a wide array of physiologic effects on the body, that may negatively impact existing chronic medical disease and complicate its management. For example, cocaine use can impact both the cardiovascular and cerebrovascular systems that can lead to stroke, disability or sudden death, particularly among adults with pre-existing disease . Second, substance use can also complicate the clinical management of existing chronic diseases.

Studies have shown decreased adherence to antiretroviral therapy among adults with HIV who engage in active substance use , and poor medication adherence for adults with psychosis who used cannabis . This is particularly relevant for adults with medical multi-morbidity, who often have complex medication schedules that demand careful monitoring and daily management. In addition, the complicated medical care that adults with medical multi-morbidity face places them at risk for medication management mistakes as well as adverse drug effects and interactions . This emphasizes the importance of substance use screening for adults with chronic conditions. In addition, since many adults in SUD treatment often have fragmented primary care , it is also important for SUD treatment providers to screen and ensure medical comorbidities are being managed. In our study sample, cannabis was the most commonly used drug among adults with chronic disease. This is not surprising since cannabis is the most common drug used in the NSDUH study sample and given more positive attitudes and policies related to cannabis use . Cannabis has been used and studied for medical treatment of chronic diseases including HIV, multiple sclerosis, chronic pain, seizure disorder, and other mental health disorders . Although this study cannot distinguish between medical use versus recreational use of cannabis, using cannabis for these types of conditions may partially explain its high prevalence use among adults with chronic disease . The risks of cannabis have not been well-documented, particularly among older adults with medical multi-morbidity who may be at higher risk for negative cardiovascular, pulmonary, and cognitive effects of cannabis use . Further study is needed to better assess the benefits and risks of cannabis use for specific chronic diseases and overall use by adults with medical multi-morbidity.

The statistically significant correlates of past-year drug use among adults with medical multi-morbidity identified in this study include younger age, male sex, low family income , current tobacco use, alcohol use and alcohol dependence, having had a major depressive episode, and having received mental health treatment. The demographic findings are similar to the overall results of NSDUH among the general adult population for illegal drug use including cannabis, such as adults with younger age and male sex having higher rates of use . However, our results are novel in that they identify a potentially very high-risk population of adults with the combination of medical multi-morbidity with poly substance use and/or mental health disorders. The strong association we found with alcohol dependence with concurrent drug use among adults with medical multi-morbidity is alarming. Unhealthy alcohol use itself can cause, exacerbate, and complicate the management of several chronic medical diseases , and therefore the combination of illegal drug use with alcohol use, especially if used concurrently, can be particularly dangerous for adults with existing chronic medical conditions. Further, the co-occurrence of substance use and SUD with mental health illnesses are well documented , and interventions for addressing patients with co-occurring conditions have been developed and studied . The added mental health comorbidity is particularly important as one recent study of 843,584 veterans in the VA system who had at least three visits to a mental health clinic found 30.6% had co-occurring psychiatric and SUDs . The added burden of poly substance use along with mental health problems highlights the need for a syndemic framework for caring for patients with compound multi-morbidity, where the focus is how multiple health conditions are adversely affected by behavioral, psychiatric, biological, and social conditions. The use of the syndemic framework can help identify intervention strategies to reduce harms and improve the management of chronic disease for adults with medical multi-morbidity, poly substance use, and mental health disorders. In practice, one approach using this framework could include the integration of primary care and chronic disease management into SUD and mental health treatment settings and vice-versa. There are important limitations to this study. First, the NSDUH responses are based on self-report and thus are subject to both recall and social-desirability bias; although the survey attempts to limit the latter via ACASI . Second, the NSDUH does not assess when a respondent fifirst experienced or was diagnosed with a chronic medical disease, but asks only for lifetime prevalence. Therefore, participants may have had a chronic medical condition years before being surveyed and that condition may have been resolved and not overlap with past year drug use. However, most of the conditions queried tend to be lifelong. In addition, while the specific chronic diseases chosen in the NSDUH study design were based on expert opinion of medical diseases that are often related to substance use , it does not include many chronic medical diseases that are often asked in clinical research and epidemiological studies to understand the burden of chronic disease in specific populations. It also does not specify “heart disease” further, thus this could include a wide range of cardiac diseases. Therefore, the generalizability of this study is limited to the specific chronic diseases asked of this study sample and how it was asked. Also, assessing DSM SUD criteria via surveys can also be limited as these are not full diagnostic interviews. Finally, our study classifies users of drugs if the individual used in the past-year and therefore does not distinguish between one-time use versus more frequent use, which may potentially have different risks and consequences.