We defined having a need for mental health treatment by having a positive screen for depressive symptoms or post traumatic stress disorder symptoms or reporting symptoms of other mental health problems, including anxiety, hallucinations, thoughts of suicide, or attempted suicide in the past 6 months. To assess current depressive symptoms, we used the Center for Epidemiologic Studies Depression Scale , considering a score of ≥22 to be evidence of depressive symptoms. We evaluated current PTSD symptoms using the Primary Care PTSD Screen , which asks participants to report whether they experienced any of four symptoms in the previous month due to a past experience: nightmares, avoidance of situations that reminded them of it, hypervigilance, or emotional numbing to their surroundings. We considered a score of four to be consistent with PTSD symptoms. To assess additional mental health problems , we used questions from the National Survey of Homeless Assistance Providers and Clients , as adapted from the Addiction Severity Index  and considered a report of any of those symptoms to be evidence of other mental health problems. We considered anyone who met criteria for depressive symptoms, PTSD symptoms or other mental health problems to have a mental health need.Drawing on Gelberg and Anderson’s model, we examined factors associated with not having received mental health treatment among those with a mental health need. We included the factors listed above,vertical grow system which we identified a priori. In the model with unmet need for mental health services, we examined whether having an alcohol or drug use problem was associated with unmet need, considering them to be need factors.

We conducted a separate analysis to examine factors associated with not having received substance use treatment amongst those with an identified need; we again used the Gelberg and Anderson model and used factors listed above, which we identified factors a priori. In the substance use model, we tested whether having depressive symptoms, PTSD symptoms, or additional mental health problems, conceptualized as need factors, were associated. We used logistic regression in these analyses. To construct our models, we included only hypothesized variables with a bivariate p value of <0.20 in the full multivariate model. To define our reduced model, we conducted backward elimination, retaining independent variables with p ≤.05. Due to a skip pattern error, we incorrectly assessed 33 individuals using the AUDIT. To correct for this, we used multiple imputation to estimate the relationship between the treatment variables and the total AUDIT scores. We conducted multiple imputation analysis in STATA 14.2. We used SAS 9.4 to conduct our descriptive and logistic regression analyses.In a population‐based sample of older adults experiencing homelessness, we found a high prevalence of unmet need for mental health and substance use treatment. While the majority of participants had mental health and substance use problems, few received treatment. One‐third of those with mental health need received mental health care.We identified predisposing and enabling factors associated with unmet treatment need. Adults aged 65 and over had a higher odds of unmet need for mental health treatment. Older adults are more likely to have competing demands, including higher physical health needs, which can interfere with receiving behavioral healthcare. Due to a shortage of geriatric psychiatrists and geriatric mental health care services, older adults may not have access to treatment when they seek care. 

The homeless population age 65 and older is expected to triple by the year 2020. Thus, there is a need to design care that meets the needs of this growing, but underserved, population. We found that having a regular healthcare provider was associated with less unmet need. Having a regular provider can increase engagement because primary care providers may help identify needs and refer to care. In safety‐net systems, such as the ones in which our participants receive care, primary care providers may be the primary source of mental health treatment, by prescribing psychotropic medication. Primary care providers are responsible for an increasing proportion of prescriptions for psychotropic medication. In addition to prescribing medication for mental health conditions, primary care providers can refer patients to outpatient mental health counseling and treatment with specialist staff or providers. In some safety‐net settings, mental health services may be colocated with physical health services via collaborative care models.Collaborative care models can enhance information sharing and treatment plan collaboration and reduce barriers to care. CCMs are effective at reducing depressive symptoms and suicidal ideation among older adults. CCMs are cost‐efficient and can increase the capacity of resource‐constrained settings to provide care for patients with complex needs. Federally Qualified Health Centers can bill for both a medical and mental health visit on the same day , and recent changes to FQHC payment codes allow billing for behavioral health care management services in addition to the FQHC billable visit. Pay‐for‐performance programs link public hospitals’ payments to care coordination and mental health treatment metrics. It is possible that participants in our study were obtaining care in safety‐net primary care settings with CCMs.

Alternatively, the reduced odds of unmet need amongst those who had regular care providers could reflect other factors that we did not measure. For example, having a regular care provider may be a marker for increased system engagement and reduced barriers to any type of care. Those who seek primary care may be more organized, knowledgeable about safety‐net service availability, and have more access to transportation and other enabling resources.. Having a case manager was associated with less mental health and substance use treatment need. In the case management brokerage model, case managers help people navigate care systems and provide a linkage to services. In the clinical case management model, case managers serve as care providers and may provide both mental health and substance use services directly. In some models,mobile grow systems such as intensive case management, case managers provide both brokerage and direct services. It is possible that the association between having a case manager and decreased odds of unmet need for both mental health and substance use services is a result of reverse causality; treatment programs may assign a case manager. We found that participants who first became homeless at age 50 or older had a higher odds of unmet substance use treatment need. Those with late onset homelessness had led more “typical” lives, with a higher likelihood of having been continuously employed and having been married or partnered. They were less likely to have had early onset of substance use problems, thus, they may have developed substance use problems more recently. These individuals may have been less aware of safety‐ net resources in general or resources for substance use treatment in particular. Spending time in jail/prison in the past 6 months was associated with reduced unmet substance use treatment need. It is possible that participants initiated substance use treatment while incarcerated. However, most incarceration settings do not provide adequate treatment services. Alternatively, as a condition of release, participants may have been required to engage in substance use treatment. Our findings indicate there is a lack of community‐based pathways into substance use care. By giving medication‐assisted treatments, such as buprenorphine for opioid use disorder and naltrexone for alcohol use disorder in primary care settings, primary care providers can begin to address this unmet need. However, there is a need for greatly expanded substance use services. Our study has several limitations. We did not use a full psychiatric diagnostic interview. However, screening measures are important empirical tools for the referral of individuals to mental health treatment, especially when integrated care is available. We did not ask participants where they received mental health services, thus we cannot determine whether they received care colocated with primary care, or treatment in mental health specific settings.Due to the success of antiretroviral therapy and an increase in the incidence of HIV infection among older adults, the proportion of older persons living with HIV in the United States is rapidly growing. Therefore, it is important to evaluate physical and emotional health among the changing demographics of PLWH. One of the most prevalent psychiatric conditions among PLWH is major depressive disorder , with PLWH at a two- to seven-fold greater risk for depressive disorders compared to the general population. PLWH have a higher prevalence of both MDD and subsyndromal depression symptomatology than HIV- individuals of the same age or the general population. A multi-site cohort study of over 1500 PLWH found lifetime depressive symptom rates of 63% and across multiple studies diagnosis of lifetime MDD ranges from 22–54% in PLWH, compared to 4.9–17.1% lifetime MDD diagnosis in the general U.S. population. 

These high rates of depression among PLWH represent a major public health concern, as depression has been linked to worse psycho logical and medical outcomes in PLWH, including lower reported quality of life, increased viral load, and a higher likelihood of mortality. Untreated depression in PLWH has also been related to increased cognitive com plaints and worse reported daily functioning compared to PLWH without depression. These medical and psychological factors may be exacerbated in older PLWH who are often burdened to a higher degree with HIV-related medical and psychological factors, in conjunction with aging related problems. Despite the high prevalence rates of depressive disorders among PLWH, depression is often under diagnosed and in adequately treated within this population, though. Given the prevalence of depression among PLWH, it is vital to evaluate other co-occurring factors that may be associated with elevated depressive symptoms. Multiple studies have found an association between higher depressive symptoms and worse quality of life , even after controlling for demographic factors. PLWH with elevated depressive symptoms report lower mental and physical health-related quality of life , supporting the idea that depression affects multiple aspects of quality of life. However, there is a dearth of research regarding the association between depression and positive psychological factors, e.g. resilience, grit, and self-rated successful aging among PLWH. Two studies have found an association between higher resilience and lower depressive symptoms among PLWH. Similarly, in PLWH greater grit has been negatively associated with major depression. In older adult persons without HIV, lower levels of depressive symptoms have been associated with increased self-rated successful aging ; however, few studies have been conducted to evaluate positive psychological factors and quality of life in relation to depressive symptomatology in PLWH compared to control participants. Given there is an increase in the population of older PLWH and that depression is a highly comorbid condition among PLWH, assessing the relationship between depressive symptoms and other psychological factors across different age decades may provide insights for clinical interventions. Therefore, we hypothesized that: 1) PLWH aged 56–65 would have the highest proportion of elevated depressive symptoms compared to HIV- participants; and 2) elevated depressive symptoms would be associated with lower ratings of HRQoL and positive psychological factors across groups, with strongest associations in the oldest PLWH.One hundred twenty-two PLWH and 94 HIV- individuals from the Multi-Dimensional Successful Aging Among HIV-Infected Adults study conducted at the University of California, San Diego HIV Neurobehavioral Research Program and the UCSD Stein Institute for Research on Aging participated in this study. The study was approved by the UCSD Institutional Review Board, and all participants provided written informed consent after the study was explained to them by a trained staff member. In order to enroll a representative cohort of participants, minimal exclusion cri teria were applied and included: 1) neurologic condition other than HIV known to impact cognitive functioning , 2) psychotic disorders , and 3) positive urine tox icology on the day of testing for illicit substances other than cannabis. Inclusion criteria were: 1) aged 36–65 years, 2) fluent in English, and 3) ability to provide informed consent.The present study provides unique findings on the interplay of de pression, HRQoL, and positive psychological factors among middle aged and older PLWH and HIV− individuals in a multi-cohort design structure. In our sample, PLWH were significantly more likely to report elevated depressive scores compared to HIV− individuals. This finding supports prior studies that have found PLWH endorse more depressive symptoms than HIV− individuals. Contrary to our hypothesis, the youngest cohort seemed to drive this finding, with a significantly larger proportion of PLWH reporting elevated depressive symptoms compared to HIV-individuals within this age group. That is, the proportion of elevated depressive symptoms did not differ by HIV status among the middle aged and older age cohorts.