Risk of Depression in People With Human Immunodeficiency Virus: A Nationwide Population-based Matched Cohort Study

Abstract

Background

Human immunodeficiency virus (HIV) infection is associated with depression. However, previous studies have not addressed familial factors.

Methods

Nationwide, population-based, matched cohort study of people with HIV (PWH) in Denmark between 1995 and 2021 who were matched on sex and date of birth with a comparison cohort randomly selected from the Danish population. Family-related factors were examined by inclusion of siblings of those in the cohorts. We calculated hazard ratios (HRs) for depression, receipt of antidepressants, electroconvulsive therapy (ECT), and suicide, as well as the yearly proportions of study cohorts with psychiatric hospital contact due to depression and receipt of antidepressants from 10 years before to 10 years after study inclusion.

Results

We included 5943 PWH and 59 430 comparison cohort members. Median age was 38 years, and 25% were women. We observed an increased risk of depression, receipt of antidepressants, ECT, and suicide among PWH in the 2 first years of observation (HR, 3.3; 95% confidence interval [CI]: 2.5–4.4), HR, 3.0 (95% CI: 2.7–3.4), HR, 2.8 (95% CI: .9–8.6), and HR, 10.7 (95% CI: 5.2–22.2), thereafter the risk subsided but remained increased. The proportions of PWH with psychiatric hospital contact due to depression and receipt of antidepressants were increased prior to and especially after HIV diagnosis. Risk of all outcomes was substantially lower among siblings of PWH than among PWH (HR for receipt of antidepressants, 1.1; 95% CI: 1.0–1.2).

Conclusions

PWH have an increased risk of depression. Family-related factors are unlikely to explain this risk.

Graphical Abstract

https://tidbitapp.io/tidbits/risk-of-depression-in-people-with-hiv-a-nationwide-population-based-matched-cohort-study

PWH, depression, antidepressants

The prognosis for people with human immunodeficiency virus (HIV; PWH) has improved substantially since the introduction of combination antiretroviral therapy (cART) [1]. HIV infection has been associated with depression [2–8]. However, previous studies lack well-matched comparison groups or have a small and selected sampling of study populations [2–4]. Further, no studies addressed whether familial factors might influence the association between HIV infection and depression [5–8]. Depression has a detrimental effect on quality of life and may lead to suicide [9]. Furthermore, depression among PWH has been associated with reduced adherence with cART, leading to reduced viral suppression, lower CD4 cell count, and increased mortality [10, 11]. Additional research is necessary to determine whether PWH have an increased risk of depression. We performed a nationwide, population-based, matched cohort study to analyze the risk of depression among PWH compared with a matched comparison cohort from the Danish population. We also examined family-related factors to see if they influence the risk of depression in PWH by analyzing the risk of depression among siblings of PWH compared with siblings of comparison cohort members.

METHODS

Setting

The study was conducted in Denmark during the study period 1 January 1995–1 September 2021, when the Danish population ranged from 5.2 million to 5.8 million, and the estimated prevalence of HIV infection in Danish adults ranged from 0.07% to 0.1% [1, 12]. Throughout the study, HIV treatment was provided at 10 specialized centers, where patients were followed with outpatient contacts in intervals of 12–24 weeks. In Denmark, all residents have access to tax-funded and cost-free healthcare services [13].

Data Sources

The data sources used were linked together by the unique 10-digit civil registration number assigned to each Danish resident at birth or immigration [13].

The Danish HIV Cohort Study (DHCS) was used to identify PWH [14]. The Danish Civil Registration System was used to form the comparison and sibling cohorts [15]. The Danish National Patient Registry was used to acquire data on affective mood disorder diagnoses, as well as comorbidity and treatment with electroconvulsive therapy (ECT) [16]. The Danish National Prescription Registry was used to acquire data on receipt of prescribed antidepressants [17] in accordance with the Anatomical Therapeutic Chemical (ATC) classification system [18]. The Danish Register of Causes of Death was used to capture data on suicide [19] (see Supplementary Table 1 for a detailed description of the different registries).

Study Population

The PWH cohort included all incident cases of PWH who were Danish residents and included in the DHCS between 1 January 1995 and 1 September 2021 (Supplementary Figure 1) [14]. The date of study inclusion was set to the last of the following: date of registration in the DHCS or date of first-time diagnosis of HIV infection in the Danish National Patient Registry.

The comparison cohort members were randomly extracted from the Danish population. We extracted 10 individuals who were alive and living in Denmark on the date of study inclusion without HIV diagnosis for every member of the PWH cohort. Comparison cohort members were matched on date of birth and sex of the PWH to whom they were matched and given the same date of study inclusion as the corresponding member of the PWH cohort.

To determine whether familial factors could affect the association between HIV infection and depression, we identified siblings of PWH and siblings of the comparison cohort members to form sibling cohorts that consisted of all full- or half-siblings with the same date of study inclusion as their relation in the PWH cohort and the comparison cohort.

Outcome Measures

Outcomes were time to depression (defined as a first-time depression diagnosis assigned in a psychiatric hospital: International Classification of Diseases, Eighth Revision, Clinical Modification [ICD-8], codes 296.09, 296.29, 298.09, 300.49 [20, 21] and ICD-10 codes F32.0-F33.9), frequency of psychiatric hospital contact due to depression diagnosis (inpatient as well as outpatient contacts), time to first-time redeemed prescription of antidepressants (ATC code N06A), frequency of redeemed prescriptions of antidepressants, time to a first-time treatment with ECT (codes BRXA1, BRXA10, and BRXA11), and time to death due to suicide (ICD-10 codes X60-X84 and Y87.0).

Other Information on Study Participants

The route of HIV infection for PWH was divided into 3 categories: men who have sex with men (MSM), intravenous (IV) substance use, and “other.” Other includes route of HIV infection due to heterosexual contact, hemophilia, blood transfusion, other, and unknown. Place of origin was divided into Danish origin and non-Danish origin, which was further divided into African, Asian, and other.

At study inclusion, the percentage of PWH and comparison cohort members with a Charlson comorbidity index score ≥2 was calculated. The Charlson comorbidity index score did not include HIV and AIDS [22].

Statistical Analyses

Time-to-Event Analyses

We performed time-to-event analyses of depression, receipt of antidepressant, treatment with ECT, and suicide. In these analyses, we calculated time from study inclusion until death, loss to follow-up, emigration, or outcome of interest, whichever came first. We excluded all members who already had the outcome of interest before study inclusion. For all PWH who were excluded, we also excluded their corresponding comparison cohort members.

As a measure of relative risk, we calculated cause-specific hazard ratios (HRs) and corresponding 95% confidence intervals (CIs) of the event of interest using Cox regression analyses, stratified according to matching sets in which we compared PWH with the comparison cohort members. Initial analyses suggested a violation of the proportional hazards assumption for the analyses of PWH compared with the comparison cohort members. Therefore, for PWH compared with the comparison cohort members, we performed a time-updated Cox regression in 2 time intervals (0–2 years after study inclusion and >2 years after study inclusion). Cox regression analyses on the sibling cohorts were adjusted for sex and age, as the sibling cohorts were not matched. The cumulative incidence function was used to calculate the absolute risk of the outcome of interest, considering death as a competitive risk.

Frequency of Outcomes

We determined the proportion of the PWH cohort and the comparison cohort members with psychiatric hospital contact due to depression and receipt of antidepressants each year from 10 years before study inclusion to 10 years after study inclusion. We then calculated absolute differences in proportions between the PWH cohort and comparison cohort members, with 95% (CIs). These analyses were chosen, as they might reflect disease burden [23] and preexisting psychiatric comorbidity [24]. These analyses were stratified according to calendar year of study inclusion (<2005 vs ≥2005). The analysis of psychiatric hospital contact due to depression was divided based on whether they were inpatient or outpatient contacts.

Sensitivity Analyses

We performed a range of sensitivity analyses stratified according to time of study inclusion, place of origin, route of HIV infection, sex, and time from initial HIV test to registration in the DHCS (Supplementary Table 2). We also performed a sensitivity analysis in which the outcome was changed to receipt of selective serotonin reuptake inhibitors (SSRIs) and according to different severities of depression (Supplementary Table 2).

Ethics Approval

The Danish Data Protection Agency approved the Danish HIV Cohort Study.

RESULTS

We included 5943 PWH and 59 430 comparison cohort members. A larger proportion of PWH had a Charlson comorbidity index score ≥2 and were born outside of Denmark compared with comparison cohort members (Table 1). We also included 5807 siblings of PWH and 82 411 siblings of the comparison cohort, who had approximately the same distribution of age, sex, and Charlson comorbidity index. More siblings of PWH than siblings of the comparison cohort had a non-Danish origin (Table 1).

Table 1.

Characteristics of People With Human Immunodeficiency Virus and Comparison Cohort Members at Study Inclusion

Characteristic	PWH (n = 5943)	Comparison Cohort (n = 59 430)	Siblings of PWH (n = 5807)	Siblings of the Comparison Cohort (n = 82 411)
Age at study inclusion,a y	38 (31–47)	38 (31–47)	36 (28–44)	36 (28–43)
Female sex, n (%)	1496 (25)	14 960 (25)	2774 (48)	39 765 (48)
Charlson comorbidity index score ≥2, n (%)	333 (6)	2423 (4)	202 (3)	2569 (3)
Origin, n (%)
Danish	3619 (61)	52 913 (89)	5229 (90)	78 477 (95)
Non-Danish	2324 (39)	6517 (11)	578 (10)	3934 (5)
Africa	984 (17)	…	…	…
Asia	503 (8)	…	…	…
Thailand	222 (44)	…	…	…
Other	837 (14)	…	…	…
Duration of follow-upa	9.8 (4.0–16.6)	12.5 (6.3–18.9)	…	…
Total years of follow-up	63 077	752 868	…	…
Route of human immunodeficiency virus infection, n (%)	…	…	…	…
Men who have sex with men	2508 (42)	…	…	…
Intravenous substance use	321 (5)	…	…	…
Otherb	3114 (52)	…	…	…

Values are numbers (percentages) unless stated otherwise.

Abbreviation: PWH, people with human immunodeficiency virus.

^aMedian, years (interquartile range).

^bOther includes route of human immunodeficiency virus infection due to heterosexual contact, hemophilia, blood transfusion, other, and unknown.

Depression

We observed an increased relative risk of depression among PWH compared with the comparison cohort members, especially in the first 2 years after study inclusion (Table 2, Figure 1). We observed an increased proportion of study participants with psychiatric hospital contact due to depression among PWH compared with the comparison cohort both prior to and after study inclusion, although the CIs for these estimates were wide (Supplementary Table 3). The most pronounced difference was observed the year of and the year after study inclusion (Figure 2, Supplementary Table 3).

Figure 1.

Cumulative incidence of depression (A), receipt of antidepressants (B), treatment with electroconvulsive therapy (C), and suicide (D) among PWH vs comparison cohort members. Abbreviation: PWH, people with human immunodeficiency virus.

Figure 2.

Proportion of study participants who had psychiatric hospital contacts due to depression (A), psychiatric outpatient hospital contacts due to depression (B), psychiatric inpatient hospital contacts (ie, admissions) due to depression (C), and receipt of antidepressants (D) each year from 10 years before study inclusion to 10 years after study inclusion.

Table 2.

Risk of Depression, Receipt of Antidepressants, Treatment With Electroconvulsive Therapy, and Suicide Among Study Participants, Stratified According to Characteristics of People With Human Immunodeficiency Virus

Outcome	2-Year Cumulative Incidence (95% CI)		HR0–2 Y (95% CI)	20-Year Cumulative Incidence (95% CI)		HR > 2 Y (95% CI)
	Cohort	Cohort		Cohort	Cohort
	PWH	Comparison		PWH	Comparison
	Total study population
Depression	1.2 (.9–1.5)	0.4 (.3–0.4)	3.3 (2.5–4.4)	5.9 (5.1–6.7)	3.4 (3.2–3.6)	1.9 (1.6–2.2)
Antidepressants	7.8 (7.1- 8.6)	2.8 (2.7–3.0)	3.0 (2.7–3.4)	30.6 (28.9–32.3)	22.5 (22.0–23.0)	1.5 (1.4–1.6)
Selective serotonin reuptake inhibitors	5.1 (4.5–5.7)	2.0 (1.8–2.1)	2.7 (2.4–3.1)	21.0 (19.6–22.5)	15.2 (14.7–15.6)	1.5 (1.3–1.6)
ECT	0.1 (.0–.2)	0.0 (.0–.0)	2.8 (.9–8.6)	0.6 (.4–1.0)	0.3 (.3–0.4)	2.0 (1.2–3.3)
Suicide	0.3 (.2–.4)	0.03 (.0–.0)	10.7 (5.2–22.2)	1.1 (0.8–1.5)	0.3 (.2–0.4)	3.6 (2.3–5.6)
	PWH from Denmark
Depression	1.4 (1.0–1.8)	0.3 (.3–.4)	4.1 (2.9–5.8)	5.5 (4.7–6.5)	3.1 (2.8–3.3)	2.0 (1.6–2.5)
Antidepressants	9.1 (8.0–10.2)	2.7 (2.5–2.9)	3.7 (3.2–4.3)	31.2 (29.1–33.3)	22.1 (21.5–22.7)	1.5 (1.4–1.7)
	PWH from Africa
Depression	0.8 (.3–1.5)	0.4 (.3–.6)	1.7 (.8–3.8)	7.1 (5.0–9.5)	4.2 (3.7–4.8)	1.8 (1.3–2.5)
Antidepressants	3.9 (2.8–5.3)	3.3 (2.9–3.7)	1.2 (.9–1.8)	28.5 (24.5–32.6)	24.6 (23.4–25.8)	1.2 (1.0–1.5)
	PWH from Asia
Depression	1.1 (.4–2.4)	0.3 (.2–.5)	3.5 (1.2–9.6)	5.5 (3.0–9.3)	4.4 (3.6–5.3)	1.3 (.7–2.4)
Antidepressants	6.7 (4.6–9.4)	2.8 (2.3–3.4)	2.4 (1.6–3.7)	25.2 (20.1–30.6)	23.3 (21.5–25.2)	1.1 (.8–1.4)
	PWH from other areas of the world
Depression	0.8 (.3–1.7)	0.4 (.3–.6)	2.1 (.9–5.2)	7.4 (4.3–11.7)	3.3 (2.7–3.9)	2.0 (1.2–3.3)
Antidepressants	8.5 (6.6–10.8)	2.5 (2.1–2.9)	3.7 (2.7–5.0)	34.4 (28.5–40.4)	20.0 (18.5–21.6)	2.0 (1.6–2.5)
	Men who have sex with men
Depression	1.3 (.9–1.8)	0.4 (.3–.4)	3.8 (2.5–5.8)	7.0 (5.6–8.5)	3.1 (2.8–3.4)	2.2 (1.8–2.8)
Antidepressants	8.1 (7.0–9.3)	2.5 (2.3–2.7)	3.5 (2.9–4.2)	32.6 (29.8–35.4)	20.5 (19.7–21.3)	1.7 (1.5–2.0)
	Intravenous substance use
Depression	1.3 (.4–3.1)	0.2 (.1–0.5)	6.5 (1.8–23.1)	4.3 (2.3–7.2)	3.4 (2.8–4.2)	2.0 (1.0–4.0)
Antidepressants	18.5 (13.8–23.7)	3.1 (2.4–3.9)	7.8 (5.2–11.7)	44.3 (37.4–51.0)	24.4 (22.5–26.5)	2.8 (2.1–3.9)
	Other
Depression	1.1 (.8–1.5)	0.4 (.3–.5)	2.8 (1.9–4.2)	5.3 (4.3–6.4)	3.6 (3.3–3.9)	1.6 (1.3–2.1)
Antidepressants	6.6 (5.7–7.6)	3.0 (2.8–3.3)	2.3 (2.0–2.8)	27.7 (25.5–29.9)	23.7 (23.1–24.4)	1.2 (1.1–1.3)
	Heterosexual PWH (women) from Denmark
Depression	1.8 (.7–3.6)	0.6 (.4–.9)	2.9 (1.1–7.1)	6.2 (3.9–9.3)	4.6 (3.8–5.4)	1.7 (.9–3.0)
Antidepressants	9.9 (6.8–13.8)	3.2 (2.5–3.9)	3.5 (2.3–5.5)	33.4 (27.2–39.7)	28.7 (26.6–30.1)	1.2 (.9–1.6)
	Heterosexual PWH (men) from Denmark
Depression	1.3 (.7–2.2)	0.3 (.2–0.5)	4.1 (2.0–8.6)	3.7 (2.5–5.4)	2.5 (2.1–3.0)	1.6 (1.0–2.7)
Antidepressants	7.4 (5.6–9.6)	2.6 (2.2–3.1)	3.1 (2.2–4.3)	24.6 (21.0–28.4)	21.2 (20.0–22.4)	1.3 (1.0–1.6)

Abbreviations: CI, confidence interval; HR, hazard ratio; PWH, people with human immunodeficiency virus.

There was no increased relative risk of depression among siblings of PWH compared with siblings of the comparison cohort members (Table 3, Supplementary Figure 2).

Table 3.

Risk of Depression, Receipt of Antidepressants, Treatment With Electroconvulsive Therapy, and Suicide Among Siblings of People With Human Immunodeficiency Virus (PWH) Compared With Siblings of the Comparison Cohort Members

Outcome	20-Year Cumulative Incidence, % (95% CI)		Hazard Ratio for the Total Study Period (95% CI)
	Siblings of PWH	Siblings of the Comparison Cohort
Depression	4.7 (4.0–5.5)	4.1 (4.0–4.3)	1.1 (.9–1.3)
Antidepressants	26.1 (24.5–27.7)	23.3 (22.6–23.7)	1.1 (1.0–1.2)
ECT	0.3 (.1–.6)	0.4 (.3–0.4)	0.8 (.4–1.5)
Suicide	0.3 (.2–.6)	0.2 (.2–.3)	1.4 (.8–2.7)
Siblings of PWH from Denmark and siblings of the comparison cohort from Denmark
	Siblings of PWH	Siblings of the Comparison cohort
Depression	4.5 (3.7–5.3)	4.0 (3.8–4.3)	1.1 (.9–1.3)
Antidepressants	25.9 (24.2–27.5)	23.3 (22.8–23.9)	1.1 (1.1–1.2)
Electroconvulsive therapy	0.3 (.2–.6)	0.4 (.3–.4)	0.9 (.4–1.6)
Suicide	0.3 (.2–.6)	0.2 (.2–.3)	1.3 (.7–2.6)

Abbreviations: CI, confidence interval; PWH, people with human immunodeficiency virus.

Receipt of Antidepressants

We observed an increased relative risk of receipt of antidepressants among PWH compared with the comparison cohort members, especially in the first 2 years after study inclusion (Table 2, Figure 1). An increased proportion of PWH had received antidepressants prior to and after study inclusion compared with the comparison cohort members; however, these differences increased substantially after study inclusion (Figure 2, Supplementary Table 3).

There was an increased relative risk of receipt of antidepressants among siblings of PWH compared with siblings of the comparison cohort members, but the difference was small (Table 3, Supplementary Figure 2).

Treatment With ECT

We observed an increased relative risk of treatment with ECT among PWH compared with the comparison cohort members (Table 2, Figure 1), although the estimates had wide CIs. We found no increased relative risk of treatment with ECT among siblings of PWH compared with siblings of the comparison cohort members (Table 3, Supplementary Figure 2).

Suicide

We observed an increased relative risk of death due to suicide among PWH compared with the comparison cohort members (Table 2, Figure 1). An increased relative risk of suicide was also observed among siblings of PWH compared with siblings of the comparison cohort members. However, the CIs for these estimates were wide (Table 3, Supplementary Figure 2).

Sensitivity Analyses

PWH from Asia seemed to have a lower relative risk of depression and receipt of antidepressants >2 years after study inclusion compared with PWH from other places, although the CIs for these estimates were wide (Table 2). PWH who were MSM and had IV substance use as the route of HIV infection seemed to have a higher relative risk of depression and receipt of antidepressants >2 years after study inclusion compared with PWH infected through heterosexual sex and other routes of infection (Table 2). Although the absolute risk of depression and receipt of antidepressants were higher among women compared with men, the relative risk of these outcomes was of the same order of magnitude across sex strata (Supplementary Table 5). The remaining sensitivity analyses did not lead to a substantial change in the Results section (Tables 2, Table 3, Supplementary Table 4, Supplementary Table 5).

DISCUSSION

In this nationwide, population-based, matched cohort study, we demonstrated an increased risk of depression, receipt of antidepressants, treatment with ECT, and suicide among PWH compared with the comparison cohort members from the background population. We also demonstrated an increased proportion of psychiatric hospital contacts due to depression and an increased proportion of receipt of antidepressants among PWH compared with the comparison cohort members, especially after, but also before, study inclusion. These findings were consistent across different strata of the study population. Finally, siblings of PWH had an increased risk of receipt of antidepressants and suicide compared with siblings of the comparison cohort members, but not to the same extent as for PWH.

The association between HIV infection and risk of depression might have several potential explanations. Reverse causality might contribute to our findings, as depression is associated with sexual risk behavior [25, 26]. This hypothesis is supported by the finding that more PWH than comparison cohort members had psychiatric hospital contacts due to depression and receipt of antidepressants before inclusion. Still, the proportion of both psychiatric hospital contacts due to depression and receipt of antidepressants increased substantially after study inclusion, which indicates that reverse causality is unlikely to fully explain the increased risk of depression among PWH.

HIV infection and some antiretroviral drugs are associated with the development of neuropathy [27]. Some antidepressants are used in the treatment of neuropathic pain, which could confound the association between HIV infection and receipt of antidepressants in our study. However, the fact that our sensitivity analysis on receipt of SSRIs resulted in essentially the same estimates as the main analysis indicates that treatment of HIV and cART-associated neuropathy with other antidepressants is unlikely to contribute substantially to our findings.

Depression is considered a multifactorial disorder [28], where both genetics and environmental factors are involved. The fact that siblings of PWH had an increased receipt of antidepressants compared with siblings of the comparison cohort members suggests that familial factors among PWH rather than the HIV diagnosis per se explain our findings. However, the receipt of antidepressants was much smaller for siblings of PWH than for PWH, which indicates that familial factors cannot explain all of our findings. Further, although speculative, receipt of antidepressants among siblings could be a result of chronic disease of the PWH and accompanying stigma rather than a marker of underlying depression risk in these families [29–31]. Taken together, we believe that HIV infection is associated with depression beyond what is explained by reverse causality and familial factors.

There might be several potential mechanisms behind an association between HIV and depression. The inflammatory response triggered by HIV infection might lead to depression, as chronic inflammation in general has been associated with mental disorders [32, 33]. Another possible mechanism could be a direct effect of viral replication in the brain due to an insufficient cART level [34]. Further, some antiviral drugs, such as efavirenz, are associated with neuropsychiatric side effects, which might explain the associations in our study. Finally, psychosocial factors such as stress, stigma, and discrimination, which are common among PWH, might lead to depression [5–7].

Gooden et al [7] demonstrated an approximately 2-fold risk of depression among PWH compared with a matched background population in the United Kingdom, whereas others have demonstrated an increased risk of suicide among PWH compared with the general population in China [35] and Switzerland [36]. Contrary to the study by Gooden et al, we demonstrated that the risk of depression is particularly high immediately following the time of HIV diagnosis when the clinician should be especially attentive to depressive symptoms.

Our study is the first to examine the massive burden of depression in the 10-year period leading up to HIV diagnosis among PWH. Wändell et al [8] studied visits to a psychiatry clinic and prescription of antidepressants 2 years before and after HIV diagnosis among PWH in Stockholm. However, the study suffered from potential selection bias and limited statistical power. In accordance with our study, Cholera et al [37] demonstrated an increased frequency of depressive symptoms in people before testing for HIV, but their study is limited by potential recall bias due to the use of a questionnaire to assess depression.

The reason that PWH from Asia were at lower relative risk of depression than PWH from other places in the world is not clear. One possible explanation is that immigrants from Asia, especially from Thailand, generally have a high socioeconomic status, which is associated with a lower risk of depression [38]. Another possible explanation is low awareness and use of the publicly funded Danish health care system among immigrants, which could result in underdiagnoses of depression. This theory is supported by recent data from the DHCS that demonstrate that HIV diagnosis is still substantially delayed in migrants to Denmark [39].

As in our study, the risk of depression is generally higher among women than men [5]. The association between HIV and depression in our study, however, was of the same order of magnitude across sex strata.

The large study size and long-term follow-up are important strengths of the study as they enabled us to study rare outcomes. Strengths also include the nationwide, population-based matched cohort design, which increases generalizability and reduces bias. It is an advantage that we used well-established national registries that are unbiased by our study hypothesis to capture study outcomes. Finally, by using the Danish National Patient Registry and the Danish National Prescription Registry, we were able to capture both more severe cases of depression (ie, those that led to referral to psychiatric hospitals) and less severe cases treated in the primary sector (ie, with the receipt of antidepressants).

It is a limitation that we have no information on alcohol use for the study cohorts because of the possible association between alcohol and depression [40]. Further, we have no information on drug use for the comparison cohort members. However, we were able to demonstrate an increased relative risk of depression, even among PWH who were not infected through IV substance use. We cannot exclude misclassification of the study outcomes, but we presume that the misclassification is nondifferential with minimal impact on relative risk estimates toward the null. When compared with the background population, PWH might be referred to other hospital-based specialties earlier, including psychiatric departments, as they are already enrolled in regular hospital-based visits. Accordingly, PWH might be diagnosed with conditions that would normally not have entailed a hospital referral and, thereby, a diagnosis in the Danish National Patient Registry [23]. The resulting ascertainment bias could affect the risk of a depression diagnosis. However, as antidepressants are most often prescribed by general practitioners, this outcome is less likely to be affected by ascertainment bias.

A final limitation is that siblings of immigrants might be less likely to be registered in the Danish Civil Registration System. Therefore, risk estimates for this subpopulation could be imprecise. However, in the sensitivity analyses that included only siblings of PWH and siblings of the comparison cohort members from Denmark, the results were essentially the same.

CONCLUSIONS

In this nationwide, population-based, matched cohort study, we demonstrated an increased risk of depression, receipt of antidepressants, treatment with ECT, and suicide among PWH compared with date of birth- and sex-matched comparison cohort members. Reverse causality and familial factors are unlikely to explain these risks. Our results demonstrate that clinicians should be attentive to symptoms of depression in people recently diagnosed with HIV. Additionally, more research and attention regarding mental health among PWH are required to understand the mechanisms behind our results and how to improve mental health among PWH.

Supplementary Data

Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Disclaimer. The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, and approval of the manuscript; and decision to submit the manuscript for publication.

Financial support. This work was supported by Gilead Sciences 2022 (Nordic Fellowship; unrestricted research grant to A. M. L.).

Author notes

Potential conflicts of interest. A. M. L. reports unrestricted research grants from Gilead and speaker honoraria/advisory board activity from Gilead, GSK, and Pfizer, with no relation to the current study; consulting fees from AbbVie and MSD (paid to institution); participation on a data and safety monitoring board for Novo Nordisk (paid to author); grants or contracts outside this work to institution from the Lundbeck Foundation and Novo Nordisk; payment or honoraria for lectures unrelated to this work from MSD, Pfizer, and GSK; and support for attending meetings from MSD, Chiesi, and Advanz. I. S. J. reports grants or contracts to institution from Odense University Hospital and consulting fees (to author) from Pfizer. M. M. T. reports a grant or contract from the Research Fund of Copenhagen University Hospital—Rigshospitalet (PhD scholarship) and travel grants for participation in a conference from GSK Pharma A/S. All remaining authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.