Abstract
Psychiatric illness and sleeping disorders are important co-morbidities of human immunodeficiency virus (HIV) infection, which impact both the individual and antiretroviral therapy (ART) selection. This systematic review aimed to assess the prevalence of psychiatric illness and sleep disturbance in people living with HIV (PLHIV) in the UK. Systematic searches for publications reporting epidemiological data for psychiatric co-morbidities and sleep disturbance with HIV were conducted in Embase, MEDLINE, Cochrane Library, eight key conferences (2013–2015), and by hand-searching references of included publications. Data were extracted from publications (2000 onwards) reporting the UK prevalence of depression, anxiety, suicide ideation, or sleep disturbance as a co-morbidity of HIV infection. Comparative UK general population data were obtained from the 2007 Adult Psychiatric Morbidity in England household survey, the 2012 Health Survey for England, and ‘PatientBase’ (epidemiological database). Sixteen publications met the inclusion criteria. Amongst PLHIV in the UK, the prevalence of depression varied from 17–47%, compared with a reported 2–5% prevalence for the UK general population. A similar disparity was observed in the prevalence of anxiety (22–49% PLHIV versus 4–5% general population), depression or anxiety (50–58% PLHIV versus 27% general population), difficulty sleeping (61% PLHIV versus 10% population), and suicide ideation (31% PLHIV versus 1% general population). This systematic review of UK data demonstrates that rates of psychiatric illness and sleep disturbance are substantially higher amongst PLHIV than in the general population. These data underline the importance of fully considering sleep and psychiatric issues prior to selection and prescription of antiretroviral drugs, as well as the need for ongoing psychiatric and psychological support for PLHIV on ART.
Introduction
Due to the introduction of combination antiretroviral therapy (ART) in the mid-1990s for the treatment of human immunodeficiency virus (HIV) infection, HIV infection is now a manageable chronic condition.1,2 However, treatment remains burdensome and can be associated with adverse events. Among people living with HIV (PLHIV), some data suggest that rates of general and psychiatric co-morbidities are high.3–5 Contributing factors to the high co-morbidity rate include a recent diagnosis of HIV, 6 burden of HIV-related stigma, 7 length of time with diagnosed HIV, 8 and HIV-mediated neural damage. 9 Furthermore, some treatments have been associated with adverse events like insomnia or depression. Depression is mentioned in the Summary of Product Characteristics for rilpivirine 10 and dolutegravir 11 (although they carry no warning for use in individuals with a prior history of psychiatric disorders). Several ART therapies, such as rilpivirine, 10 emtricitabine, 12 cobicistat, 12 dolutegravir, 11 lamivudine, 13 efavirenz (EFV), 14 and raltegravir, 15 may be associated with insomnia, which is described as a ‘common’ or ‘very common’ adverse event with these treatments. In fact, commonly-prescribed ARTs, including EFV and raltegravir, carry warnings regarding their use in individuals with a prior history of psychiatric disorders.14,15 In light of this warning, the British HIV Association (BHIVA) 16 and the US Department of Health and Human Services (DHHA) 17 HIV treatment guidelines recommend that individuals with a history of psychiatric illness or suicide ideation are treated with alternative regimens which do not contain EFV. So far, the prevalence of psychiatric illness suffered by PLHIV in the UK has not been systematically assessed, although it could be considerable. In 2015, there were an estimated 36.7 million people worldwide living with HIV. 18 As of June 2016, 18.2 million PLHIV were accessing ART. 18 In the UK, there are an estimated 101,200 PLHIV. 19 The impact of psychiatric illness, or even less serious issues like sleep disturbance, in PLHIV should not be underestimated. Such issues can lead to reduced adherence and thus development of drug resistance,20,21 complicating treatment and eventually impacting public health.
Our work aimed to assess the extent of the issue of psychiatric illness and sleep disturbance in PLHIV in the UK. This systematic review was therefore designed to identify recent UK data on the current prevalence of psychiatric and sleep disorders amongst PLHIV and to compare these with data from the UK general population.
Methods
Systematic review
Search methods
The following electronic databases were searched via OVID: Embase, MEDLINE In-process & Other Non-Indexed Citations, MEDLINE, and the Cochrane Library. Only English language publications published from 2000 onwards were searched on 2 April 2015; detailed search strings, which combine subject index headings with free text terms, are provided in the supporting information (please see online Supplementary Tables 1, 2 and 3). In addition, all abstracts published between January 2013 and April 2015 from BHIVA, the annual conference on Retroviruses and Opportunistic Infections, European AIDS Conference, International AIDS Society, Interscience Conference on Antimicrobial Agents and Chemotherapy, and International Society for Pharmacoeconomics and Outcomes Research conference proceedings were screened. Reference lists of all included studies were examined. Citations identified through the searches were assessed by a reviewer (NA) based on title and abstract using predefined eligibility criteria. Any uncertainties were resolved by discussion with a second reviewer (MK). Full publications of potentially relevant citations were obtained and examined by two reviewers (NA and MK). Disputes were resolved via discussion with a third party (AS). Relevant data from eligible publications were extracted into a data extraction table by a reviewer (NA) and verified by a second reviewer (MK).
Eligibility criteria
Key eligibility criteria were recorded a priori in a review protocol (not registered) and are provided in Table 1. Studies were included if participants were adults (≥18 years) with HIV, and they reported the incidence/prevalence of psychiatric co-morbidities and/or sleep disturbance.
Eligibility criteria.
HIV: human immunodeficiency virus.
Structured review
In addition to the systematic review conducted to identify the prevalence of psychiatric conditions in PLHIV, a structured review was conducted to determine the prevalence of psychiatric conditions and sleep disturbance in the general UK population. Comparative UK general population data were obtained from the 2007 Adult Psychiatric Morbidity in England household survey, the 2012 Health Survey for England, and ‘PatientBase’ epidemiological database.
For each outcome/co-morbidity data point identified in the systematic review (e.g. the prevalence of anxiety), comparable top-line data for the general population were sought and suitable data matched accordingly.
Results
Systematic review
A total of 16 publications met the inclusion criteria for the systematic review4,5,22–35 and reported data regarding the prevalence of psychiatric illness in PLHIV. Four publications reported results from the ASTRA study,27,29,33,34 while one publication reported data from the CIPHER study, 28 which was a sub-study of the ASTRA study. Two publications reported results from the PIVOT study.4,35 A further two publications reported data from the same study (no unique study name provided).5,32 The remaining seven publications reported on seven individual studies. Therefore, data were reported from a total of 11 unique studies. The majority of studies recruited PLHIV from HIV outpatient clinics.4,22–29,31,33–35 The recruitment setting was not reported for one study. 30 Two of these studies also included comparative general population data.28,33,34 These studies were largely similar in study design and patient demographics (Table 2). All data reported were extracted from the most recent publication associated with each study.
Study setting and patient characteristics of publications included from systematic review and structured review.
HIV: human immunodeficiency virus; MSM: men who have sex with men; PLHIV: people living with HIV.
aOf PLHIV population.
bPublication provided baseline patient data only, no prevalence data.
c3258 HIV+ pts screened for sleep disorder, depression, anxiety; 3151 HIV+ pts screened for anxiety or depression.
dAbstract superseded by full publication.
The study flow diagram is shown in Figure 1. A completed PRISMA checklist can be found in the supporting information (Supplementary Table 4).

Study flow diagram.BHIVA: British HIV association; HIV: human immunodeficiency virus; RCT: randomised clinical trial.
Structured review
The structured review identified four additional sources of psychiatric condition prevalence data for the UK general population: a 2007 Adult Psychiatric Morbidity in England household survey conducted for the NHS Information Centre for health and social care, 39 and three PatientBase epidemiological databases, which provided prevalence data for major depression, generalised anxiety disorder, and sleep disturbance.36–38
Outcomes
The main outcomes identified from the systematic review were the prevalence of depression,22–27 anxiety,23,24,27 anxiety or depression,4,32,34 suicide ideation, 5 and sleep disturbance 27 in PLHIV.
The cohort size in the studies presenting data for PLHIV ranged from 15 31 to 3258. 34 Studies reporting comparative data for the UK general population included between 45 28 and 7424 individuals.27,33,34 The study setting and patient characteristics of publications included from both the systematic review and structured review are presented in Table 2. The raw data and definitions of psychiatric co-morbidities from included studies are presented in Table 3.
Raw data and definition of psychiatric co-morbidities.
CI: confidence interval; CMD: common mental disorder; DRG: Decision Resources Group; EQ-5D: EuroQol five dimensions questionnaire; GAD-7: generalised anxiety disorder 7-item; HADS: Hospital Anxiety and Depression Scale; HIV: human immunodeficiency virus; MDD: major depressive disorder; MSAS: Memorial Symptom Assessment Schedule; MSM: men who have sex with men; PHQ-9: Patient Health Questionanaire-9; PLHIV: people living with HIV.
Depression
The prevalence of depression in PLHIV was reported in six studies22–27 and ranged from 27%26,27 to 47%. 22 In comparison, two studies reported lower rates of depression amongst the general UK population,39,38 ranging from 2% 39 to 5% 38 (Figure 2). Three studies reported the prevalence of depression in men who have sex with men (MSM),28–30 ranging from 16% 30 to 29%, 28 while a prevalence of 8% was reported for MSM in the general UK population 28 (Figure 2). In PLHIV who were aged ≥75 years, the prevalence of depression was 13% 31 compared with only 2% in the same subgroup in the general UK population 39 (Figure 2).

Summary of the reported prevalence of psychiatric illness and sleep disturbance in PLHIV compared with the general UK population. PLHIV: people living with HIV.
Anxiety
Three studies reported the prevalence of anxiety among PLHIV,23,24,27 which ranged from 22% 27 to 49%. 23 In the general population the reported prevalence was 4% 39 and 5% 37 (Figure 2). Three studies reported anxiety prevalence rates of 27% 30 and 21%28,29 in MSM, compared with 14% for MSM in the general UK population 28 (Figure 2).
Depression or anxiety
Three studies reported the prevalence of depression or anxiety in PLHIV,4,32,34 which ranged from 33% 4 to 58%. 32 This contrasts with a prevalence of 27% in the general UK population 34 (Figure 2). In the subgroup of white PLHIV, the reported depression or anxiety prevalence was 34%, 4 compared with 19% in white people in the general UK population 39 (Figure 2). The prevalence of depression or anxiety in black people in the general UK population was estimated to be 22%, 39 while a higher prevalence of 30% was reported in black PLHIV 4 (Figure 2).
Suicide ideation
Only one study reported the prevalence of suicide ideation in PLHIV (31%) 5 ; prevalence of the same psychiatric condition was estimated to be 1% in the general UK population 39 (Figure 2).
Sleep disturbance
The prevalence of sleep disturbance was 61% in PLHIV (based on a single study) 27 compared with 10% in the general UK population 36 (Figure 2).
Discussion
This systematic review identified 16 publications (from 11 unique studies) that reported the prevalence of psychiatric illness and sleep disturbance in PLHIV from the UK.4,5,22–35 Three studies27,33,34 and three epidemiological databases36–39 reported the prevalence of psychiatric illness and sleep disturbance in the general UK population. The results of our review show that the prevalence of depression, anxiety, depression or anxiety, suicide ideation, and sleep disturbance is substantially higher in PLHIV compared with the general UK population (Figure 2).
Our analysis is limited by the fact that the scope of the systematic review did not cover all psychiatric co-morbidities/adverse events associated with HIV infection or its treatments. The data obtained did not indicate any such selection bias; however, some trials were abstracts only and/or the study methods did not allow us to determine if there was any selection bias in the trials. Since our search was conducted in April 2015, we cannot exclude that the review may have missed more recent relevant publications. The prevalence data for the UK general population were obtained using a structured rather than a systematic review, and thus may not be comprehensive. Data were limited for certain psychiatric co-morbidities that were included in the scope, particularly for suicide ideation and sleep disturbance. Additionally, this review presents study-level data with no adjustment for individual confounding factors. This may impact the magnitude of difference observed for the prevalence of certain co-morbidities between PLHIV and the general UK population. The definitions of psychiatric conditions varied between studies. For example, different definitions were used in each of the five studies reporting the prevalence of depression in PLHIV (Table 3). Furthermore, the systematic review did not assess which treatment PLHIV were receiving. Therefore, the data do not allow us to determine the presence or extent of any association between the prevalence of psychiatric illness and any individual HIV ART.
Although ART has been an important and life-saving step change in the treatment of HIV, all current ARTs (like any pharmacological treatment) are associated with adverse events. In the case of ART, depression and insomnia, among other psychiatric illness, are of concern. EFV and raltegravir treatment regimens have been associated with a range of psychiatric side effects and they carry a special warning with regard to their use in individuals with a previous history of depression. Use of EFV has been associated with sleeping difficulty, 40 impaired cognitive function, 41 anxiety, 42 depression, 42 and suicide ideation. 43 Several other current ARTs list insomnia as a common side effect. Insomnia is known to adversely affect quality of life 44 and has been associated with non-adherence to ART, 45 which may be detrimental for viral suppression 21 and can promote drug resistance. 20 Other lifestyle factors, such as shift work, may compound these problems. PLHIV working shift hours may find it hard to adhere to a regular routine, which could affect the times that ART is taken. In the UK it is reported that 33% of men and 22% of women are engaged in shift work and given the prevalence of HIV in the UK (an estimated 101,200 PLHIV 19 ), it is likely that a proportion of PLHIV are engaged in shift work. Results from this study highlight that 61% of PLHIV in the UK have difficulty sleeping.
Besides ART, there are other factors which may contribute to the higher prevalence of psychiatric illness and sleep disturbance among PLHIV. In particular, people recently diagnosed with HIV may develop depression and anxiety as a result of their recent diagnosis. In a study by Weber et al., 6 increased anxiety and depression was observed in antiretroviral-naïve adults in the acute and early stages of HIV infection (median infection duration of 75 days) compared with HIV-negative adults. The length of time HIV infection has been diagnosed is also associated with depression and anxiety. A recent study by McGowan et al. 8 reported that a longer time since HIV diagnosis was associated with a trend for high prevalence of anxiety and depression (p < 0.001). Additionally, the strain of HIV-related stigma in PLHIV is associated with higher rates of psychological illness, particularly depression. 7 The impact of HIV-mediated neural damage can also elicit emotional and behavioural co-morbidities, such as depression, anxiety, and sleep disorders. 9
Possible solutions to reduce the prevalence of psychiatric illness among PLHIV have been suggested.16,46 Choosing the most appropriate ART treatment for individuals may help to reduce the prevalence of psychiatric illness. The identified data indicate that a substantial proportion of PLHIV in the UK (30–58%) requiring ART may benefit from an alternative to EFV. Substantiating this observation, recent guidelines by the DHHA 17 and BHIVA 16 for the treatment of HIV recommend that individuals with a history of psychiatric illness and suicide ideation are treated with alternative regimens which do not contain EFV. Additionally, the BHIVA guidelines recommend that EFV-containing regimens should be promptly switched to a viable alternative when PLHIV present with psychiatric symptoms. 16
Regular psychiatric screening utilising appropriate testing strategies and the provision of accessible psychiatric support may offer an opportunity to improve the clinical management of psychiatric illness in PLHIV. The 2012 standards for psychological support for adults living with HIV published by the British Psychological Society and BHIVA 46 highlight the importance of identifying psychological illness among PLHIV and providing interventions which promote mental health and well-being. In particular, the standard recommends that PLHIV should have access to regular annual screening to identify their psychological support needs and a range of psychological support services appropriate to their needs. 46 HIV treatment should improve the physical and psychological well-being of PLHIV. In light of the psychiatric adverse events possibly associated with a number of ART treatments, including key first-line treatments like EFV, it would clearly be beneficial to more widely implement psychiatric evaluations when assessing PLHIV for ART, as the presence of psychiatric illness may be exacerbated using these treatments. Better incorporation of psychiatric evaluations and counselling into annual patient reviews, better access to specialists, and training/education for non-specialists could help to address the issue of the high prevalence of psychiatric conditions (including depression), and issues like insomnia.
Conclusion
Our systematic review found that rates of depression, anxiety, suicide ideation, and sleep disturbance are substantially higher amongst PLHIV than in the general UK population. This should have implications for ART selection and psychiatric and psychological patient care.
Footnotes
Acknowledgements
Éanna Connaughton, an employee of DRG Abacus, contributed to the writing of this manuscript.
Authors’ contributions
NA and MK, employees of DRG Abacus, contributed to collection of data and data analysis. All authors contributed to the study design, interpretation of results and to the writing and final approval of this manuscript. NA provided writing assistance for the manuscript.
Declaration of conflicting interests
The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: LW, CR, and AS are employed by Gilead.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by Gilead.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
