Suffering from ≥2 chronic conditions is termed comorbidity when one of the conditions is regarded as primary or multimorbidity (without a primary condition) [1]. Such phenomena are associated with symptom severity, increased healthcare utilization, and a poorer prognosis [2, 3]. The vast majority of comorbidity/multimorbidity research understandably addresses physical illnesses in elderly populations [4, 5]. From a public health viewpoint and irrespective of age, mental health receives less attention than physical health, despite its substantial contribution to disease burden, further worsened by the COVID-19 pandemic [6] and the marked reduction in life expectancy among individuals with mental illness compared with the general population [7, 8]. Findings from the 27-country World Health Organization (WHO) Mental Health Surveys revealed that each lifetime mental disorder was associated with a substantially increased risk of subsequent mental comorbidity, with the risk persisting over >15 years [9].

Here we focus on three mental health conditions - anxiety, insomnia, and eating disorders (ED) - because they are relatively frequent in the general population, display a strong potential for comorbidity, and offer opportunities for prevention or treatment [10–12]. The most common among them is anxiety (encompassing generalized anxiety disorder, panic disorder, social anxiety, phobias, etc.) with prevalence estimates of 4%–25%, with higher rates among women, younger adults, and individuals with chronic illness [13]. During 1990–2019, disability-adjusted life years attributable to anxiety disorders increased worldwide by nearly 54% across age [14]. Next, a meta-analysis reported that not only anxiety but also many other mental disorders were concomitant with sleep disturbances, implying an imbalance in the arousal system [15]. Moreover, network analyses have provided evidence for connections among symptoms of anxiety, sleep disorders and ED [11].

The U.S. National Comorbidity Survey (baseline: 1990–1992; replication: 2001–2003) was among the first to address the prevalence and correlates of comorbid and multimorbid mental disorders [3, 16]. However, to date, few epidemiological studies have investigated the socio-demographic profiles of pure versus comorbid/multimorbid mental disorders in the general population [17–20] which is indispensable for identifying at-risk subgroups. To our knowledge, no prior epidemiological research has addressed the multimorbidity of anxiety, sleep disturbance/insomnia and ED in general-population adults. Only one small cross-sectional study with 130 Brazilian adults with overweight or obesity included measures of all three disorders and reported that adults aged <45 years with high trait anxiety also had high scores on binge eating and low scores on sleep quality, based on two-way correlations [21].

There is compelling need to advance mental multimorbidity research to inform targeted public health interventions. Therefore, this descriptive study investigated the prevalence and degree of mental multimorbidity, focusing on three mental health conditions and the socio-demographic characteristics of pure versus multimorbid cases. Given that women present higher rates of anxiety [13], insomnia [22] and ED [23], we explored sex-specific associations.

In total, N = 33,658 participants (25,269 women, 8,389 men) had complete and valid mental health and covariate data and were included in the analysis (Figure 1). The overall and sex-specific raw and weighted distribution of mental morbidity prevalence is summarized in Table 1. Overall, 40.6% (men 37.4%; women 43.6%) had ≥1 mental health condition; 2.3% had all three conditions (men 1.4%; women 3.1%). Across sex, the most prevalent comorbidity was anxiety and insomnia (men 6.1%; women 6.4%) whereas the least prevalent was insomnia and ED (men 0.9%; women 1.6%). The raw and weighted sociodemographic profiles associated with mental morbidity are presented in Table 2 (women) and Table 3 (men).

According to WHO, mental disorders are widespread, undertreated, and under-resourced [6]. To help address this issue, this large, descriptive study focused on anxiety, insomnia, and ED owing to their relatively high prevalence and comorbidity in the general adult population, and the potential for prevention or treatment [10–12]. Among 33,658 individuals, a weighted total of 40.6% presented ≥1 mental disorder and an absolute total of 821 participants presented anxiety-insomnia-ED multimorbidity. Women were twice as likely as were men to have the latter (3.1% vs 1.4%, respectively). For most pure and comorbid conditions, women were at higher risk than were men.

Some distinct sex-specific sociodemographic profiles of mental morbidity emerged. Among women, having any ≥2 mental health conditions was associated with a higher prevalence of smoking than having ≤1 condition. Such a trend was absent among men. The largest proportions of younger adults (aged 18–39 years) were found among women with comorbid anxiety and ED and among men with pure anxiety, whereas the largest proportions of adults with obesity were found among those with comorbid insomnia and ED, across sex. Multimorbidity was associated with more risk behaviors among men than among women. The anxiety-insomnia-ED multimorbidity group had the largest proportions of men who were overweight (52.1%) and current smokers (23.2%), while no such trends emerged in women. These findings are not fully consistent with prior research in college students, showing that those with anxiety and/or mood disorders were more likely to be daily tobacco users compared to their counterparts without such mental disorders [40].

The present findings are largely consistent with prior mental health research reporting a high level of psychiatric comorbidity in the general population, observing that risk factor profiles for comorbid disorders differed considerably from those for pure disorders; such observations have important implications for prevention and clinical practice [17]. In addition, prior research regarding the 12-month comorbidity of anxiety, mood, and substance use disorders reported that the odds ratios for parental psychiatric history and childhood trauma were higher for comorbid anxiety-mood disorder, suggesting increased vulnerability and a more severe condition, than for either disorder in its pure form [17]. Also congruent with the present findings are previous reports of a positive association of mental comorbidity with female sex, living alone, and being of low socio-economic status [17–20]. Low physical activity and obesity have likewise been individually and positively associated with anxiety, ED, and insomnia/sleep disorders [41–45]. Even though ED types were not assessed in this study, the results regarding the prevalence of obesity suggest that restrictive types were unlikely to have driven the associations.

The frequent comorbidity of mental disorders, the shared risk factors and genetic expressions, and the observed activation in the same brain regions, especially those involved in emotion, learning and memory [7, 9, 46] have led some researchers to challenge the traditional view of mental disorders as distinct entities. To explain the underlying structure of mental illness, researchers have even proposed a general psychopathology continuum termed the “p factor” [47]. Likewise, the WHO stresses that mental health exists on a complex continuum [6]. Many physiological, genetic, affective, and cognitive mechanisms have been implicated in mental multimorbidity. For example, the association between anxiety and insomnia has been postulated to be driven by common gene variant heritability, dysregulation in circadian clock gene expression, serotonin, dopamine, and inflammatory cytokine secretion, and by cognitive inflexibility, interpretational biases, rumination tendencies, and impaired social interactions [48–50]. In addition, the role of the microbiome and the gut-brain axis in mental health have been attracting attention due to their involvement in numerous neuroendocrine, immune, inflammatory, and neurotransmitter pathways [51]. The bidirectional gut-brain communication is mediated by neural and humoral mechanisms. Whereas much work remains to be done in this area, it is known that intestinal dysbiosis and behavioral impairment are bidirectionally linked [51].

This descriptive study provided weighted prevalence estimates which argues for the external validity of the findings and against any selection bias. Specifically, the weighting largely compensated for differences (e.g., age, sex, educational level) between the study participants and the general French population. However, the cross-sectional design precludes any inference of causality. Likewise, the chronology of the three mental health conditions was beyond the scope of the study. There is evidence of complex bidirectional and possibly mediated associations among anxiety, insomnia, and ED [52–55]. Future longitudinal research could shed light on all of these issues. Another limitation, common to epidemiological research, pertains to the use of self-reported data provided by volunteers. We relied on DSM-5 and ICSD-3 criteria for insomnia, and on validated tools for anxiety (STAI-T) and ED (SCOFF); however, data obtained with these tools cannot serve as evidence for clinical diagnoses. A limitation of SCOFF is the lack of distinction among the various ED types. It should also be acknowledged that racial/ethnic background is an important component of the sociodemographic profile, however, such information was not available in our database.

In conclusion, this weighted analysis provided information about the prevalence and degree of mental multimorbidity of anxiety, insomnia, and ED among adults in the general population. Some distinct sex-specific sociodemographic profiles of mental morbidity emerged, which could be taken into consideration by targeted prevention programs, future cross-sectional and longitudinal mental multimorbidity research, and could help generate novel moderation and mediation hypotheses. In addition, the prevalence of mental comorbidity and multimorbidity justifies future methodological work aimed at generating mental multimorbidity composite measures. The findings also support the adaptation of medical education, public health guidelines, and healthcare services for multimorbidity [56, 57]. The advent of artificial intelligence in mental health will further necessitate the elaboration of novel regulatory frameworks, guidelines, and policies [58].