Major Depressive Episodes and Psychiatric Symptoms Among HIV Controllers Compared With Antiretroviral-Treated People Living With HIV

Abstract

Background

Major depressive episodes are highly prevalent among people living with HIV (PWH), despite sustained virological suppression. However, psychiatric outcomes among HIV controllers (HIC), who maintain long-term spontaneous control of viral replication without antiretroviral therapy (ART), remain poorly characterized. We assessed major depressive episodes and psychiatric symptom profiles among HIC compared with ART-treated PWH and explored associated factors.

Methods

In this prospective observational study, HIC from the French multicenter ANRS CO21 CODEX cohort were compared with PWH receiving suppressive ART. Major depressive episodes were assessed using the Mini-International Neuropsychiatric Interview (M.I.N.I.), and psychiatric symptoms were evaluated using the Symptom Checklist-90–Revised. Between-group comparisons and univariable analyses were performed to identify factors associated with psychiatric outcomes.

Results

Ninety-eight participants were included (57 HIC and 41 ART-treated PWH). The prevalence of lifetime major depressive episodes was high in both groups and did not differ between HIC and ART-treated participants (35% vs 41%). Most psychiatric symptom dimensions were comparable between groups, although HIC showed higher interpersonal sensitivity and psychopathology composite scores on the SCL-90-R, reflecting higher psychological distress. In both groups, psychiatric outcomes were associated with preexisting psychiatric history, alcohol use, and sociodemographic factors, but not with immunovirological parameters.

Conclusions

Despite spontaneous viral control, HIC do not exhibit more favorable psychiatric outcomes than ART-treated PWH. The high prevalence of depressive disorders appears to be primarily driven by behavioral and sociodemographic factors and preexisting psychiatric conditions rather than immunovirological status, supporting routine psychiatric assessment in HIV care, including for HIC.

Major depressive episodes and psychiatric symptoms remain highly prevalent among people living with human immunodeficiency virus (HIV), even in the current era of sustained virological suppression. Recent global systematic reviews indicate that between one quarter and more than half of people living with HIV (PWH) experience symptoms of depression, highlighting depression as one of the most frequently reported psychiatric disorders in this population [1, 2]. Depression among PWH receiving antiretroviral therapy (ART) has been extensively described and is commonly associated with substance use and social vulnerability, including socioeconomic hardship and HIV-related stigma [3]. By contrast, associations between depressive symptoms and immunovirological parameters such as CD4 T-cell count or plasma HIV viral load appear inconsistent and generally weaker across studies [1, 4]. Emerging evidence suggests that inflammatory pathways may contribute to depression among PWH. Elevated levels of pro-inflammatory cytokines and markers of immune activation have been associated with depressive symptoms in several HIV populations [5, 6]. In addition, certain antiretroviral drugs, most notably efavirenz, have historically been associated with neuropsychiatric adverse effects, including depressive symptoms and suicidal ideation [7, 8]. Other mental health symptoms including anxiety and broader psychological distress have been reported in studies enrolling PWH worldwide, including Europe and sub-Saharan Africa [9–11].

However, psychiatric outcomes remain poorly characterized among HIV controllers (HIC), who represent approximately 0.5% of PWH and maintain long-term spontaneous control of viral replication in the absence of ART, with distinct virological and immunological profiles [12, 13]. Some HIC (∼20% in the French Agence Nationale de Recherche sur le SIDA et les Hépatites virales (ANRS) CODEX cohort) keep strictly undetectable HIV viral loads over time, without any history of viral blips (undetectable HIC, uHIC) [14], and exhibit very low levels of immune activation and inflammation. Conversely, other HIC might experience viral blips throughout their follow-up and display inflammatory marker levels comparable to those observed in ART-controlled PWH [14, 15]. The prevalence of non-AIDS-related events (nADEs) is broadly similar to that of ART-controlled PWH, including cardiovascular and oncologic disorders [16–20].

Data on psychiatric outcomes among HIC remain scarce, particularly regarding the potential contribution of viral blips to major depressive episodes. In a large retrospective cohort study conducted in the United States, HIC were reported to have a nearly fourfold higher risk of psychiatric hospitalization compared with PWH receiving ART, even after adjustment for demographic, behavioral, and clinical covariates [21]. Notably, psychiatric outcomes were identified using hospitalization codes, without standardized psychiatric assessment, and the study did not allow for a detailed characterization of depressive disorders or psychological symptom profiles. In France, a comparative analysis of 2 prospective ANRS cohorts enrolling HIC and ART-treated PWH followed since primary HIV infection revealed no increased risk of hospitalization for psychiatric causes or reported mental health diseases [18]. However, complementary psychosocial and quality-of-life studies conducted within ANRS cohorts suggested that spontaneous antiviral control does not necessarily translate into better mental health outcomes. Despite long-term viral control and favorable virological status, HIC were reported to experience psychological distress related to uncertainty about their health status and identity issues, as well as impaired mental quality of life compared with ART-treated individuals [22, 23]. Taken together, these findings indicate that, while spontaneous viral control might intuitively confer psychological benefits, mental health outcomes among HIC remain insufficiently characterized.

The PSYCHIC study was therefore designed to characterize psychiatric and psychological profiles among HIC using standardized diagnostic interviews and validated self-reported symptom questionnaires. Specifically, we aimed to assess major depressive episodes and psychiatric symptoms among HIC and to identify associated demographic, behavioral, and clinical determinants, as compared with PWH receiving ART.

METHODS

Study Design and Population

HIV controllers were enrolled from the French multicenter ANRS Cohorte des extrêmes (CODEX), as previously described [24]. This prospective cohort enrolls people living with HIV-1 for at least 5 years who have maintained spontaneous viral control—defined as 5 consecutive plasma HIV-1 RNA measurements below 400 copies per mL in the absence of ART—and who have never received ART, except for women treated during pregnancy, with treatment discontinued after childbirth. Among HIC, additional virological definitions were used for descriptive purposes. Individuals whose routine plasma HIV-1 RNA levels consistently remained below the detection threshold throughout their entire history of spontaneous viral control were referred to as uHIC [14]. Viral blips were defined as isolated plasma HIV-1 RNA measurements above the detection threshold that were not confirmed on subsequent testing. Individuals presenting with at least one blip in their history were referred to as blipper HIC (bHIC), in accordance with previous descriptions of the ANRS CODEX cohort [14]. For each participant, uHIC or bHIC status was defined using all available plasma HIV-1 RNA measurements up to the time of inclusion in the PSYCHIC study and was subsequently treated as a fixed binary variable in all analyses. The study was conducted in 30 centers of the CODEX cohort between 2022 and 2025 with consecutive recruitment of participants.

The control group consisted of PWH receiving ART with sustained virological suppression. Eligible participants were people living with HIV-1 who had maintained plasma HIV-1 RNA levels below the detection threshold (usually 20 copies/mL) for at least 12 months under stable ART at the time of inclusion in the PSYCHIC study; individuals with a history of virological failure within the preceding 12 months were excluded. Control participants were recruited contemporaneously with HIC, where feasible, among PWH attending routine HIV follow-up visits at each center, with the aim of achieving comparable distributions of sex assigned at birth, age (±5 years) and geographic origin.

All participants provided written informed consent, and the study protocol was approved by the Comité de Protection des Personnes Île-de-France VII (reference 11-033, ID-CRB 2011-A00734-37).

Data Collection

Clinical and demographic data were collected at study inclusion using standardized case report forms. Variables included age, sex assigned at birth, body mass index, geographic origin, educational level, duration of HIV infection, and virological status. Sexual behavior was self-reported. Participants reporting bisexual behavior were classified according to their predominant sexual behavior. Immunovirological and laboratory parameters, including CD4 and CD8 T-cell counts, CD4/CD8 ratio, ultrasensitive plasma HIV-1 RNA levels, total HIV DNA level, and co-infection serologies, were extracted from medical records. Ultrasensitive plasma HIV-1 RNA levels and total HIV DNA levels were available only for HIC and were not routinely measured in ART-treated participants. For participants receiving ART, treatment history was also collected. Data were obtained from medical records and participant interviews conducted during routine clinical visits.

Psychiatric Assessments

Information on substance use and addictive behaviors, as well as psychiatric history, including previous psychiatric care, diagnosed psychiatric disorders, and past or current psychotropic medication, was collected from all participants. Alcohol consumption was analyzed separately for frequency and quantity.

Psychiatric assessments were conducted at study inclusion. Major depressive episodes were assessed using the depressive disorder module of the Mini-International Neuropsychiatric Interview (M.I.N.I.), administered as a structured hetero-questionnaire using the validated French version [25, 26]. This module explores the presence of criteria for a major depressive episode over the preceding 15 days as well as a lifetime history of major depressive episodes, including age at first episode and the total number of depressive episodes.

Psychiatric symptoms at the time of assessment were further evaluated using the Symptom Checklist-90–Revised (SCL-90-R) [27, 28]. Ninety items were measured according to how much each symptom distressed the participant and scored on a 5-point Likert scale ranging from 0 = not at all to 4 = extremely. This self-administered questionnaire assesses 9 symptom dimensions: somatization, obsessive–compulsive symptoms, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, paranoid ideation, and psychoticism. In addition to these standard dimensions and indices, an additional composite dimension referred to as “psychopathology” was analyzed, as commonly used in French adaptations of the SCL-90-R [28]. This composite dimension includes items capturing depressive and neurovegetative features such as appetite and sleep disturbances, early morning awakening, feelings of guilt, and thoughts of death. Additionally, the SCL-90-R provide 3 global indices which reflect the degree of symptom severity: the Global Severity Index (GSI) summarizes the severity of all items as the average depth of impairment; the Positive Symptom Total (PST) index indicates the total number of symptoms experienced; and the Positive Symptom Distress Index (PSDI) reflects the level of distress corrected for the number of symptoms reported [28].

Statistical Analysis

Continuous variables are reported as medians with interquartile ranges (IQR) and categorical variables as numbers and percentages. Between-group comparisons were performed using the Wilcoxon–Mann–Whitney rank-sum test for quantitative variables and Fisher's exact test for categorical variables as appropriate.

Univariable analyses were performed to investigate associations between psychiatric outcomes and clinical, behavioral, and immunovirological parameters in the overall population of the PWH enrolled (n = 98) and within each group (HIC and ART), using nonparametric methods (Spearman's rank correlation and Wilcoxon–Mann–Whitney tests for continuous variables, or Fisher's exact tests for categorical variables).

No formal correction for multiple comparisons was applied in the primary analyses. To assess the robustness of the findings, P values for the SCL-90-R dimensions and global indices were additionally adjusted using the Bonferroni method. Statistical significance was defined as a 2-sided P < .05. All analyses were performed using R software (R Core Team, version 4.5.2). Figures were generated using GraphPad Prism (version 10.6.1; GraphPad Software, San Diego, CA, United States).

RESULTS

Study Population

A total of 98 participants were included in the analysis comprising 57 HIC from the ANRS CODEX cohort and 41 PWH receiving ART with sustained virological suppression. Baseline demographic, clinical, and behavioral characteristics of the study population are presented in Table 1.

Table 1.

Baseline Demographic, Clinical, and Behavioral Characteristics of the Study Population

	HIC (n = 57)	ART (n = 41)	P
Demographic characteristics
Sex assigned at birth (woman/man), n/n	34/23	23/18
Age, years [IQR]	58.0 [48.0–65.0]	59.1 [49.7–65.6]
Geographic origin, n (%)
Sub-Saharan African	17 (30)	14 (34)
Caribbean	3 (5)	1 (2)
European	33 (58)	23 (56)
Maghreb	4 (7)	3 (7)
Body mass index, kg/m2 [IQR]	25.6 [22.8–28.7]	25.6 [21.8–28.7]	.68
Sexual behavior category (predominant), n (%)			.63
Homosexual behavior	12 (21)	11 (27)
Heterosexual behavior	45 (79)	30 (73)
Mode of HIV acquisitiona, n (%)			.09
Sexual	44 (77)	40 (98)
Injection drug use	6 (11)	1 (2)
Blood exposure	3 (5)	1 (2)
Unknown	7 (12)	1 (2)
Education level, n (%)			.70
Primary education or less	3 (5)	4 (10)
Secondary education	30 (53)	21 (51)
Higher education	24 (42)	16 (39)
Substance use and addictive behaviors
Smoking, n (%)			.81
Never	34 (60)	23 (56)
Current	13 (23)	9 (22)
Former	9 (16)	9 (22)
Alcohol consumption, n (%)			.19
Never	15 (27)	17 (41)
Any drinking (past or not)	41 (73)	24 (59)
Frequency			.60
Once per month	14 (25)	9 (23)
2–4 times per month	13 (23)	5 (13)
2–3 times per week	9 (16)	3 (8)
≥4 times per week	5 (9)	7 (18)
Quantity			<.001
1–2 drinks per occasion	31 (54)	0 (0)
3–4 drinks per occasion	7 (12)	17 (41)
5–6 drinks per occasion	2 (4)	6 (15)
>7 drinks per occasion	1 (2)	1 (2)
Drug use, n (%)			.42
Never	46 (81)	37 (90)
Current	7 (13)b	3 (7)c
Past use	4 (7)b	1 (2)c
Non-substance-related addictive behaviors, n (%)	4 (7)d	0 (0)	.14
Medical history
Hypertension, n (%)	23 (41)	17 (41)	1.00
Diabetes mellitus, n (%)	6 (11)	5 (12)	1.00
Coronary artery disease, n (%)	3 (5)	2 (5)	1.00
Dyslipidemia, n (%)	8 (14)	15 (37)	.01
Chronic respiratory failure, n (%)	3 (5)	2 (5)	1.00
Cancer, n (%)	3 (5)e	1 (2)f	.64

Age, sex assigned at birth, and geographic origin were presented descriptively and were not compared statistically, as these variables were considered during recruitment to achieve comparable distributions between groups. Between-group comparisons were assessed using Fisher's exact test for categorical variables and the Wilcoxon–Mann–Whitney rank-sum test for continuous variables.

Abbreviations: ART, antiretroviral therapy; HIC, HIV controllers; HIV, human immunodeficiency virus.

^aSeveral PWH had dual exposure (sexual and injection drug use).

^bAmong CODEX participants, current drug use involved cannabis (n = 5), cocaine (n = 2), and synthetic cathinones (n = 1).

^cAmong ART-treated participants, current drug use included cannabis (n = 2) and cocaine (n = 1); information regarding the type of substances used in past drug use was unavailable.

^dOther addictions (n = 4): food, exercise, sex, and gambling.

^eCancer history includes breast cancer (treated with combined hormone and radiation therapy; remission > 20 y), squamous cell carcinoma of the larynx in remission, and thyroid cancer (status unknown).

^fLung cancer under immunotherapy.

Demographic and Clinical Characteristics

Consistent with previous descriptions of the CODEX cohort, the distribution of sex assigned at birth was 60% women and 40% men among HIC. Age and geographic origin were comparable between HIC and ART-treated participants (Table 1), consistent with the recruitment strategy aimed at achieving similar distributions between groups. Body mass index and educational level were also comparable between groups (Table 1).

Tobacco use and the frequency of alcohol consumption did not differ significantly between groups; however, the quantity of alcohol consumed per occasion was higher among ART-treated participants (P < .001). While self-reported current or past substance use was comparable between groups, injection drug use as the mode of HIV transmission was more frequent among HIC (Table 1).

The prevalence of most medical comorbidities was comparable between groups, except for dyslipidemia, which was more frequently reported among participants receiving ART (Table 1; P = .01). No participant in either group had a documented history or clinical suspicion of neurocognitive disorders at the time of inclusion in the PSYCHIC study, nor a known history of major central nervous system involvement, including progressive multifocal leukoencephalopathy, HIV-associated encephalitis, or cerebral toxoplasmosis.

Immunovirological Characteristics

Human immunodeficiency virus history and immunovirological characteristics of the study population are detailed in Table 2. The duration of known HIV infection was long and did not differ significantly between HIC and ART-treated participants (median 19.0 years [IQR 11.8–26.8] vs 23.4 years [16.2–31.4]). As expected, the nadir CD4 T-cell count was significantly higher among HIC than among ART-treated participants.

Table 2.

HIV History and Immunovirological Characteristics of the Study Population at the Time of PSYCHIC Enrollment

	HIC (n = 57)	ART (n = 41)
HIV history and baseline immunovirological status
Duration of known HIV infection (y)	19.0 [11.8–26.8]	23.4 [16.2–31.4]
Median year of HIV diagnosis [IQR]	2004 [1996–2011]	2000 [1992–2007]
CD4 nadir (cells/mm3) [IQR]b	551 [390–680]	249 [169–332]
CD4/CD8 ratio nadir [IQR]c	0.8 [0.6–1.0]	0.51 [0.3–0.9]
HCV co-infection, n (%)	2 (4)	4 (10)
CMV seropositivity, n/N (%)	32/41 (78)	19/25 (76)
Toxoplasmosis seropositivity, n/N (%)	24/40 (60)	12/28 (43)
Current immunovirological parameters at PSYCHIC enrollment
CD4 count (cells/mm3) [IQR]	879 [632–1073]	729 [634–924]
CD4/CD8 ratio [IQR]	1.2 [1.0–1.6]	1.1 [0.8–1.50]
HIV viral load (copies/mL) [IQR]	<20 [<20–30]	<20 [<20–30]
Blips during the two years prior to inclusion, n (%)	22 (39)	…
Ultrasensitive HIV RNA (copies/mL) [IQR]	1.5 [0–5.5]a	…
Total HIV DNA (copies/106 PBMCs)	20 [4–57.5]a	…
HIC group
Controller status, n (%)	11 (19)	…
Undetectable HIC (uHIC)	46 (81)	…
Blipper HIC (bHIC)	…	…
ART group
Duration of ART, years [IQR]	…	14.8 [10.1–18.2]
Number of ART regimens [IQR]	…	5 [4–6]
EFV exposure (past or current), n (%)	…	15 (37)
ABC exposure (past or current), n (%)	…	22 (54)
First-generation PI exposure (past), n (%)	…	14 (37)

Between-group comparisons were assessed using Fisher's exact test for categorical variables and the Wilcoxon–Mann–Whitney rank-sum test for continuous variables.

Abbreviations: ABC, abacavir; ART, antiretroviral therapy; EFV, efavirenz; HIC, HIV controllers; HIV, human immunodeficiency virus; PI, protease inhibitor.

^aTotal HIV DNA and ultrasensitive RNA correspond to the most recent available sample for each participant.

^b P < .001.

^c P = .001.

The prevalence of hepatitis C virus co-infection and seropositivity for cytomegalovirus and Toxoplasma gondii was similar in both groups (Table 2). At study inclusion, CD4 T-cell counts and the CD4/CD8 ratio were comparable between the 2 groups.

Among ART-treated participants, the median duration of ART was 14.8 years (IQR 10.1–18.2). Fifteen participants (37%) had current or past exposure to efavirenz. Twenty-two participants (54%) had current or past exposure to abacavir, and 14 (37%) had been exposed to first-generation protease inhibitors. The median number of antiretroviral regimens was 5 (IQR 4–6), reflecting long treatment histories.

Among HIC, 11 participants (19%) met the definition of uHIC while 46 (81%) were classified as bHIC. Viral blips within the 2 years preceding inclusion were observed in 22 HIC (39%). The median ultrasensitive plasma HIV-1 RNA level was 1.5 copies/mL (IQR 0–5.5), and the median total HIV DNA level was 20 copies per 10⁶ PBMCs (IQR 4–57.5).

Mental Health Assessments

History of psychiatric diseases, major depressive episodes, and SCL-90-R symptom scores are presented in Table 3. No significant differences were observed between groups regarding self-reported psychiatric history.

Table 3.

Mental Health Characteristics of the Study Population

	HIC (n = 57)	ART (n = 41)	P Value
Psychiatric history and treatments
Current psychiatric follow-up, n (%)	3 (5)	1 (2)	.64
Known psychiatric disorder, n (%)	8 (14)	4 (10)	.76
Anxiety disorders	1 (2)	0 (0)
Depressive disorders	7 (12)	4 (10)
Past psychotropic treatmenta, n (%)	12 (21)	8 (20)	1.00
SSRI	9 (16)	3 (8)
Tricyclic antidepressants	1 (2)	0 (0)
Benzodiazepines	2 (4)	5 (13)
Antipsychotics	2 (4)	1 (2)
Current psychotropic treatmenta, n (%)	7 (13)	2 (5)	.29
SSRI	6 (11)	1 (2)
Benzodiazepines	4 (7)	1 (2)
Antipsychotics	1 (2)	0 (0)
Psychiatric assessment
Major depressive disorder with M.I.N.I., n (%)
Any lifetime major depressive episode	20 (35)	17 (41)	.53
Current major depressive episode	2 (4)	0 (0)
History of major depressive episode	11 (19)	9 (22)
Recurrent major depressive disorder	7 (12)	8 (20)
Timing of major depressive episode relative to HIV diagnosisb, n (%)			.15
Before HIV diagnosis	9 (45)	5 (29)
After HIV diagnosis	10 (50)	7 (41)
Concomitant	1 (5)	5 (29)
Number of lifetime major depressive episodes [IQR]	1 [1–3]	1 [1–4]	.62
SCL-90-R symptom dimensions [IQR]
Somatization	3 [1–7]	2 [1–6]	.97
Obsessive–compulsive	2 [1–6]	2 [0–5]	.38
Interpersonal sensitivity	1 [0–4]	0 [0–2]	.04
Depression	3 [0–8]	2 [0–6]	.36
Anxiety	1 [0–4]	1 [0–3]	.43
Hostility	0 [0–1]	0 [0–2]	.72
Phobic anxiety	0 [0–2]	0 [0–1]	.16
Paranoid ideation	1 [0–3]	1 [0–3]	.79
Psychoticism	1 [0–2]	1 [0–2]	.51
Psychopathology (composite dimension)	4 [2–7]	2 [1–5]	.04
SCL-90-R global indices [IQR]
Global Severity Index	0.21 [0.11–0.49]	0.18 [0.10–0.40]	.23
Positive Symptom Total	15 [7–28]	11 [6–23]	.25
Positive Symptom Distress Index	1.31 [1.13–1.77]	1.30 [1.10–1.70]	.36

Between-group comparisons were assessed using Fisher's exact test for categorical variables and the Wilcoxon–Mann–Whitney rank-sum test for continuous variables.

Abbreviations: ART, antiretroviral therapy; HIC, HIV controllers; M.I.N.I., Mini-International Neuropsychiatric Interview; SCL-90-R, Symptom Checklist-90–Revised; SSRI, selective serotonin reuptake inhibitor.

^aSeveral treatments may apply to the same PWH.

^bPercentages calculated among participants with a lifetime history of major depressive episodes.

According to the M.I.N.I., the prevalence of lifetime major depressive episodes was 35% among HIC and 41% among ART-treated participants, and current major depressive episodes were infrequent in both groups (4% vs 0%). No significant differences were observed between HIC and ART-treated participants for depressive disorder outcomes. Among participants with a lifetime history of major depressive episodes, the temporal relationship between depression and HIV diagnosis was heterogeneous in both groups. In HIC, episodes occurred before HIV diagnosis (n = 9), after HIV diagnosis (n = 10), or concomitantly (n = 1). In ART-treated participants, episodes also occurred before HIV diagnosis (n = 5), after HIV diagnosis (n = 7), or concomitantly (n = 5). The median number of lifetime major depressive episodes did not differ between groups and was 1 in both groups (IQR 1–3 in HIC and 1–4 in ART-treated participants).

SCL-90-R symptom dimension scores were generally low in both groups (Figure 1). HIV controllers had higher scores for interpersonal sensitivity compared with ART-treated participants (median [IQR] 1 [0–4] vs 0 [0–2], P = .04). Consistent with M.I.N.I. findings, SCL-90-R depression scores did not differ between HIC and ART-treated participants (Table 3). No significant between-group differences were observed for somatization, obsessive–compulsive symptoms, anxiety, hostility, phobic anxiety, paranoid ideation, or psychoticism.

Figure 1.

SCL-90-R symptom dimension scores in HIV controllers (HIC; n = 57, gray) and people living with HIV receiving antiretroviral therapy (ART; n = 41, white). Data are presented as boxplots with overlaid individual data points. Boxes represent the interquartile range (25th–75th percentiles), center lines indicate the median, and whiskers show the distribution of values. Between-group comparisons were performed using the Wilcoxon–Mann–Whitney rank-sum test. Higher scores indicate greater symptom severity and poorer mental health. The y-axis is truncated at 30 for clarity.

HIV controllers also exhibited higher scores for the additional “psychopathology” composite dimension (median [IQR] 4 [2–7] vs 2 [1–5], P = .04). By contrast, no significant between-group differences were observed for the global SCL-90-R indices, including the GSI (0.21 vs 0.18), the PST (15 vs 11), and the PSDI (1.31 vs 1.30).

After Bonferroni correction for multiple comparisons across the SCL-90-R dimensions and global indices, the differences observed for interpersonal sensitivity and psychopathology scores were no longer statistically significant.

Factors Associated With Mental Health Outcomes

As expected, preexisting psychiatric history showed consistent associations with major depressive episodes in both HIC and ART-treated participants, as well as with higher symptom scores across multiple SCL-90-R dimensions and global indices (Supplementary Table 1A and B).

Among behavioral factors, higher alcohol consumption frequency was associated with major depressive episodes in the overall study population (P = .006; Supplementary Table 2).

Sex assigned at birth, age, and menopausal status—approximated by age ≥ 50 years as a surrogate marker—were explored in univariable analyses as correlates of psychiatric outcomes, with no significant associations observed within either group.

Geographic origin was associated with both substance use and psychiatric symptoms. Participants of European origin reported higher rates of tobacco use (31/56, 55% vs 10/42, 24% P = .002) and recurrent major depressive episodes (14/56, 25% vs 1/42, 1% P = .002) compared with non-European participants. In contrast, participants of sub-Saharan African origin reported lower levels of substance use but higher scores on the somatization dimension of the SCL-90-R (Supplementary Figure 1).

Finally, no associations were observed between immunovirological parameters (eg, DNA levels, presence or magnitude of blips, uHIC/bHIC status, CD4 T-cell count, or CD4/CD8 ratio) and psychiatric outcomes.

DISCUSSION

In this study, we analyzed for the first time the mental health outcomes of HIC by standardized questionnaires, compared with PWH receiving ART. We found a high prevalence of lifetime major depressive episodes, affecting approximately 40% of HIC, with no difference between groups. Importantly, psychiatric outcomes were not associated with immunovirological parameters. Despite long-term spontaneous viral control, very low cumulative viremia in some HIC, and the absence of ART exposure, HIC did not exhibit more favorable psychiatric outcomes than ART-treated participants. Regarding psychiatric symptoms, differences were limited to specific symptom dimensions, with HIC exhibiting slightly higher levels of interpersonal sensitivity and higher scores on the SCL-90-R psychopathology composite dimension, while most other symptom dimensions were comparable. Psychiatric vulnerability was primarily associated with preexisting psychiatric history, behavioral factors, particularly alcohol use, and sociocultural characteristics, rather than with HIC status or immunovirological parameters.

This high prevalence of major depressive episodes is consistent with previous reports among PWH [1]. Early meta-analyses demonstrated a markedly increased risk of depressive disorders in this population, while more recent reviews and large observational studies in the contemporary treatment era confirm that prevalence remains high, typically ranging from 20% to over 50% [2, 4]. This wide range of prevalence estimates is partly explained by substantial methodological heterogeneity across studies, including differences in study populations, assessment tools, and outcome definitions, as highlighted in systematic reviews focusing on psychological distress among PWH [1, 29]. By contrast, epidemiological studies in the general population estimate the lifetime prevalence of major depressive disorder at approximately 15%–20%, underscoring the significant psychiatric morbidity observed in our cohort [11, 30]. Notably, similar prevalence estimates were observed in HIC and ART-treated participants, supporting the notion that depressive disorders among PWH are driven by factors extending beyond virological control alone.

Data on psychiatric outcomes among HIC remain limited, with available evidence relying on heterogeneous study designs and outcome definitions. The HIC population included in the PSYCHIC study was consistent with the characteristics commonly reported in the French CODEX controller cohort [14, 24]. Differences in substance use and viral co-infections across HIC cohorts may influence reported psychiatric outcomes. In our study, hepatitis C virus (HCV) seroprevalence was low, in contrast to several other European HIC cohorts, particularly Spanish cohorts, where HCV co-infection is more frequent [20, 31]. Data on psychiatric outcomes among HIC derive from retrospective analyses. A large retrospective cohort study conducted in the United States reported a markedly increased risk of psychiatric hospitalizations among controllers compared with ART-controlled PWH. However, these outcomes were identified through diagnosis code data rather than through standardized psychiatric evaluations, limiting direct comparison with studies based on individual-level clinical evaluations [21]. Complementary insights from qualitative research suggest that HIC experience a distinctive psychosocial positioning, characterized by an ambiguous identity between health and illness and heightened sensitivity to social perceptions, which has been associated with poorer mental and social quality-of-life outcomes [22, 23]. In this context, the higher levels of interpersonal sensitivity observed among HIC in the present study may reflect this psychosocial positioning rather than an increased prevalence of overt psychiatric disorders. However, these differences did not remain statistically significant after correction for multiple comparisons and should therefore be interpreted with caution. By combining structured psychiatric interviews with validated self-reported symptom questionnaires and a contemporaneous comparison group of ART-controlled participants, the PSYCHIC study provides a quantitative and clinically grounded assessment of psychiatric outcomes among HIC.

Our analyses indicate that psychiatric outcomes among PWH are predominantly shaped by non-virological factors, since no association with HIV viral load or CD4 T cells were observed. As expected, reported preexisting psychiatric care was associated with major depressive episodes in both groups. Behavioral factors were also associated with major depressive episodes, notably alcohol use frequency. In addition, higher somatization scores observed among participants of sub-Saharan African origin may reflect culturally mediated patterns of symptom expression, characterized by a greater emphasis on somatic complaints rather than affective or emotional symptoms, as previously described among PWH in sub-Saharan Africa [32]. Although certain ART have historically been associated with neuropsychiatric adverse effects [7], ART exposure did not emerge as a determinant of psychiatric outcomes in the present study and may instead contribute to the higher prevalence of dyslipidemia observed among ART-treated participants [33]. These findings are consistent with previous studies showing that, among PWH, depressive disorders are more closely linked to the combined influence of preexisting mental health conditions, behavioral factors, and social adversity, rather than by HIV-related virological parameters [34–37].

HIV controllers may exhibit persistent immune activation and low-grade inflammation, sometimes at levels comparable to those observed in PWH receiving ART, including elevated levels of soluble CD14 or interferon-driven markers such as IP-10 [38]. These inflammatory pathways have been implicated in the pathophysiology of depression among PWH [39, 40]. Recent studies have also reported associations between depressive symptoms and inflammatory biomarkers, including pro-inflammatory cytokines such as IL-6 and TNF-α, as well as markers of monocyte/macrophage activation, among PWH [6, 41]. However, direct evidence specifically linking immune activation to mental health outcomes in HIC remains limited. Although depression among PWH is multifactorial, integrative approaches combining immunological and psychiatric data, such as those implemented within the PSYCHIC study, may help clarify the contribution of chronic immune activation to psychiatric outcomes among HIC.

These findings strongly support the need for systematic screening of depressive symptoms among PWH, including HIC. The use of brief, validated screening tools during routine care, followed by targeted clinical assessment when screening is positive, may facilitate early identification of depressive disorders, as recommended in current international clinical guidelines [42, 43]. A systematic consideration of behavioral factors and prior psychiatric history when evaluating mental health in PWH is mandatory to improve the timely recognition and management of depression, even among individuals with sustained virological control.

This study has several limitations. First, the sample size was modest, reflecting the rarity of HIC, which may have limited statistical power, particularly for association analyses. Second, psychiatric outcomes were assessed cross-sectionally and relied partly on self-reported measures, precluding causal inference. Third, immune activation markers were not directly explored in this analysis and warrant further investigation using integrated immunological and mental health assessments. Fourth, the ART-treated comparator group was heterogeneous with respect to treatment history, cumulative ART exposure, and duration of virological suppression. Therefore, similarities or differences between groups may reflect factors related to long-term treatment rather than HIV status alone. Finally, the absence of a general population control group limits direct comparison with individuals without HIV and precludes conclusions regarding the excess psychiatric burden attributable to HIV infection or HIC status.

CONCLUSIONS

Despite spontaneous control of HIV replication and the absence of ART, HIC do not exhibit more favorable psychiatric outcomes than PWH receiving ART. The high prevalence of depressive disorders in both groups appears primarily associated with behavioral factors, preexisting psychiatric conditions, and sociodemographic factors, including geographic origin. No associations were observed with the immunovirological parameters assessed in this study. These findings support routine screening for depressive disorders and the integration of mental health care into routine HIV care, including for HIC.

Supplementary Data

Supplementary materials are available at Open Forum 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.