Introduction
Nonalcoholic fatty liver disease (NAFLD) is a chronic liver disease with a spectrum of clinicopathological presentations. The clinical course and progression of NAFLD often begins with the accumulation of triglycerides in hepatocytes which predisposes these cells to inflammation and necrosis resulting in nonalcoholic steatohepatitis (NASH).Reference Brunt, Wong and Nobili 1 , Reference Machado and Diehl 2 The progression of the disease can lead to eventual cirrhosis and possibly hepatocellular carcinoma.Reference Brunt, Wong and Nobili 1 Although several histopathological features of NAFLD and NASH are similar to what are observed in alcoholic fatty liver disease (eg. steatosis, inflammation, hepatocyte ballooning, Mallory–Denk bodies, and fibrosis within the lobules), NAFLD/NASH occurs in the absence of excess alcohol use.Reference Rostama, Beauchemin and Bouchard 3
The pathology of NAFLD is conceptualized as a complex multifactorial disease state involving metabolic-inflammatory effectors often observed across disparate disorders (eg. diabetes, mood disorders, obesity, and cardiovascular disease).Reference Rostama, Beauchemin and Bouchard 3 , Reference Lonardo 4 It is currently estimated that approximately 25% of the world’s population meet or exhibit histopathological evidence of NAFLD.Reference Younossi, Koenig and Abdelatif 5 The rising prevalence of obesity and associated metabolic syndrome in both middle- and high-income countries is predicted to result in a higher number of NAFLD cases globally over the coming years.Reference Lazarus, Mark and Anstee 6 Overarchingly, the highly comorbid, insidious, and prevalent nature of NAFLD makes it a significant healthcare concern.Reference Brunt, Wong and Nobili 1
It is posited that the pathogenesis of mood disorders includes, but is not limited to, disturbances across metabolic and inflammatory systems.Reference Mansur, Brietzke and McIntyre 7 Although the association between the metabolic-inflammatory pathways and mood disorders is robust, the exact causal direction and longitudinal presentations remain incompletely understood. Therefore, it is postulated that a bidirectional link exists between metabolic-inflammatory syndromes and mood disorders.Reference Mansur, Brietzke and McIntyre 7 – Reference Mansur, Brietzke and McIntyre 9 A separate line of evidence unequivocally demonstrates that persons with mood disorders are differentially affected by obesity, diabetes, and metabolic syndrome.Reference Otte, Gold and Penninx 10 , Reference Soczynska, Kennedy and Woldeyohannes 11 A derivative of the foregoing observation is that persons with mood disorders would be predicted to be at greater risk for NAFLD and NASH.Reference Rostama, Beauchemin and Bouchard 3 , Reference Soczynska, Kennedy and Woldeyohannes 11
This narrative review aims to summarize the extant literature pertaining to the association of NAFLD and NASH with psychiatric disorders, the shared pathophysiological nexus, and potential therapeutic and research vistas.
Association of NAFLD with psychiatric disorders
To better understand the association of NAFLD with psychiatric disorders, we subdivided the available literature across the following major psychiatric disorders: major depressive disorder (MDD), bipolar disorder (BD), schizophrenia, and cognitive dysfunction.
Major depressive disorder
MDD is a chronic illness associated with affective, somatic, and cognitive alterations.Reference Otte, Gold and Penninx 10 The global disease burden of MDD has increased, with an annual >50 million years lived with disability. 12 Many recent cross-sectional studies have established a significant association of MDD with co-occurrence of NAFLD (Table 1). Furthermore, a recent meta-analysis synthesizing cross-sectional studies has shown that depression increases the odds of developing NAFLD by up to 42% (pooled OR = 1.46, 95% CI: 1.15, 1.85, p = 0.002).Reference Gu, Zhang and Hu 13 The foregoing meta-analysis also reported that patients with NAFLD were more likely to report depressive symptoms when compared with their non-NAFLD counterparts in general (pooled OR = 1.13, 95% CI: 1.03, 1.24, p = 0.007).Reference Gu, Zhang and Hu 13
Characteristics of Key Studies Ascertaining NAFLD Across Patients with Depression
Abbreviations: ALD, Adrenoleukodystrophy; BDI, Beck Depression Inventory; BMI, Body mass index; CH-B, Chronic hepatitis B; CH-C, Chronic hepatitis C; CI, Confidence interval; CLD, Chronic liver disease; CT, Computerized tomography; DM type 2, Type 2 diabetes mellitus; DSM-5, Diagnostic and Statistical Manual of Mental Disorders, fifth edition; DSM-IV, The Diagnostic and Statistical Manual of Mental Disorders, fourth edition; DSM-IV-TR, Diagnostic and Statistical Manual of Mental Disorders, fourth edition, text revision; FIB-4, Fibrosis-4; FLI, Fatty liver index; GAD, Generalized anxiety disorder; GGT, Gamma-glutamyl transferase; HADS, Hospital Anxiety and Depression Scale; HCV, Hepatitis C virus; HS, Health Subjects; hs-CRP, High-sensitivity C-reactive protein; ICD-109, International Classification of Diseases; MDD, Major Depressive disorder; MRI, Magnetic Resonance Imaging; NAFLD, Non-alcoholic fatty liver disease; NASH, Non-alcoholic Steatohepatitis; NHANES, The National Health and Nutrition Examination Survey; OR, Odd ratio; PHQ-9, Patient Health Questionnaire; RNA, Ribonucleic acid; WPAI, Work Productivity and Activity Impairment.
To our knowledge, two recently published cohort studies have established the risk of developing NAFLD in individuals diagnosed with depression, and vice versa. Cho et al. followed Korean adults (n = 142 005) with no NAFLD and no excessive alcohol consumption at baseline for a median of 4 years. Overall, depression modestly increased the risk for developing NAFLD (aHR = 1.07, 95% CI: 1.03–1.11, p < 0.001). However, when stratified by BMI category, the risk increased significantly for overweight (BMI ⩾ 25 kg/m2) individuals (aHR = 1.25, 95% CI: 1.16–1.35, p < 0.001) in comparison to normal weight individuals (BMI ≤ 25 kg/m2; aHR = 1.02, 95% CI: 0.98–1.07, p = 0.072).Reference Cho, Chang and Sung 14
Another longitudinal study that followed adolescents (n = 160) with NAFLD for a mean duration of 3.8 years reported a steady increase in the incidence of depression among participants over time from baseline (i.e., 8.1% had depression initially).Reference Noon, D’Annibale and Schwimmer 15 During the follow-up period, an additional 9.5% (95% CI = 4.7%–14.3%) of the adolescents with NAFLD developed depression. However, the study by Noon et al. was limited due to the lack of control group (i.e., adolescents without NAFLD) and relatively small number of participants.Reference Noon, D’Annibale and Schwimmer 15 Further information regarding important diagnostic definitions, patient characteristics, and relevant findings of these two cohorts is delineated in Table 1.
Multiple cross-sectional and cohort studies have also described a positive bidirectional association between the severity of depression and the degree of NAFLD fibrosis. In the cohort by Cho et al., obese people with greater severity of depression exhibited an increased risk of having NASH with greater degrees of fibrosis when compared to obese people having lesser to no depression (aHR = 3.55, 95% CI: 1.38–9.13, p < 0.005).Reference Cho, Chang and Sung 14 Similarly, in the second longitudinal study, adolescents who developed depression reported worse outcomes based on alanine transferase liver enzyme serology (indicator of liver injury) when compared with those who did not develop depression (i.e., improved 21%, stable 36%, worse 43% versus improved 42%, stable 35%, worse 23%; p < 0.01, respectively).Reference Noon, D’Annibale and Schwimmer 15
Moreover, in the majority of cross-sectional studies, a dose-dependent correlation was found between severity of depression and the histological stage of NAFLD even after adjustment for confounding factors (ie. diabetes, gender, age, obesity, hypertension, etc.).Reference Choi, Chung and Kang 16 – Reference Youssef, Abdelmalek and Binks 23 Further relevant details are described in Table 1.
Bipolar disorder
BD is a serious and debilitating illness formally differentiated from MDD by the presence of hypomania or mania.Reference McIntyre, Berk and Brietzke 9 Accumulating research has shown that patients with BD are more likely to develop insulin resistance, obesity, and other metabolic disruptions when compared with patients with MDD.Reference McIntyre, Berk and Brietzke 9 Notwithstanding, studies investigating the association of BD with NAFLD are limited. A retrospective cross-sectional electronic chart review of veterans (n = 10638) reported that patients with BD (both I and II) had significantly increased odds of developing NAFLD when compared with healthy controls (aOR = 2.57, SE = 0.316, p = 0.03).Reference Fuller, Rodriguez and Linke 24
In another recently published study, Godin et al. analyzed the data of 1969 BD (both I and II) patients to ascertain the association with NAFLD. The prevalence of NAFLD was found to be 28.4%, with a higher rate in males compared to females (OR = 3.5; 95% CI = 2.54–4.72; p = N/A). When compared with the prevalence of NAFLD in the general population, which is estimated to be at 17%, the likelihood of presenting with NAFLD was significantly greater in patients with BD.Reference Godin, Leboyer and Belzeaux 25 , Reference Nabi, Lacombe and Boursier 26 The prevalence increases further when comorbid diabetes, metabolic dysfunction, and obesity are present, along with BD.Reference Godin, Leboyer and Belzeaux 25 An important consideration is that a significant proportion of patients with BD are prescribed antipsychotic agents for maintenance treatment, and many of these agents (eg. risperidone, olanzapine, or quetiapine) are associated with significant metabolic alterations and could play a part in overall pathogenesis of NAFLD.Reference McIntyre, Berk and Brietzke 9 ,27–29 Parenthetically, the relationship between valproate use and risk of NAFLD is not established, although several studies report such an association.Reference Bayat, Jalali and Poursadeghfard 30 , Reference Luef, Waldmann and Sturm 31 Moreover, the included studies remained limited in discerning the difference between the association of BD-I and BD-II with NAFLD, respectively.
Schizophrenia
Schizophrenia is a chronic psychotic illness that often requires life-long management with antipsychotic agents.Reference Jawad, Alnefeesi and Lui 32 Long-term exposure to antipsychotic agents is associated with progressive disruption of glucose-insulin homeostasis, dyslipidemia, and weight gain.Reference Casey, Haupt and Newcomer 27 , Reference Cha and McIntyre 28 As such, NAFLD is more prevalent among patients with schizophrenia as compared to general population.Reference Rostama, Beauchemin and Bouchard 3 Like MDD and BD, multiple cross-sectional studies have shown a strong association between schizophrenia and NAFLD. For pertinent details of key cross-sectional studies, refer to Table 2.Reference Fuller, Rodriguez and Linke 24 , Reference Hsu, Chien and Lin 33 – Reference Yang, Cheng and Zhang 37
Characteristics of Key Studies Ascertaining NAFLD Across Patients with Cognitive Deficits, Bipolar Disorder and Schizophrenia
Abbreviations: ALD, Adrenoleukodystrophy; BDI, Beck Depression Inventory; BMI, Body mass index; CH-B, Chronic hepatitis B; CH-C, Chronic hepatitis C; CI, Confidence interval; CLD, Chronic liver disease; CT, Computerized tomography; DM type 2, Type 2 diabetes mellitus; DSM-5, Diagnostic and Statistical Manual of Mental Disorders, fifth edition; DSM-IV, The Diagnostic and Statistical Manual of Mental Disorders, fourth edition; DSM-IV-TR, Diagnostic and Statistical Manual of Mental Disorders, fourth edition, text revision; FIB-4, Fibrosis-4; FLI, Fatty liver index; GAD, Generalized anxiety disorder; GGT, Gamma-glutamyl transferase; HADS, Hospital Anxiety and Depression Scale; HCV, Hepatitis C virus; HS, Health Subjects; hs-CRP, High-sensitivity C-reactive protein; ICD-109, International Classification of Diseases; MDD, Major Depressive disorder; MRI, Magnetic Resonance Imaging; NAFLD, Non-alcoholic fatty liver disease; NASH, Non-alcoholic Steatohepatitis; NHANES, The National Health and Nutrition Examination Survey; OR, Odd ratio; PHQ-9, Patient Health Questionnaire; RNA, Ribonucleic acid; WPAI, Work Productivity and Activity Impairment.
In a prospective interventional study, patients with schizophrenia (N = 191, with 180 being antipsychotic-naive) were maintained on antipsychotic agents (i.e., aripiprazole, risperidone, quetiapine, or ziprasidone), and followed for three years.Reference Morlán-Coarasa, Arias-Loste and Ortiz-García de la Foz 38 At the end of the follow-up period, 25.1% (n = 48/191) developed liver steatosis, and 19.4% (n = 37/191) were determined to have indeterminate steatosis (i.e., at high future risk). No statistical difference was found between the incidence of NAFLD and the initial choice of antipsychotic agent.Reference Morlán-Coarasa, Arias-Loste and Ortiz-García de la Foz 38 Results from this study emphasize the hidden burden of NAFLD among patients with schizophrenia.
Cognitive dysfunction
Extensive research in the last decade has ascertained that cognition is paramount when patient-reported functionality and well-being are considered.Reference Pan, Park and Brietzke 39 The residual cognitive deficits following remission of an MDD or BD episode hamper the recovery of an individual to premorbid functionality.Reference McIntyre, Cha and Soczynska 40 Even in the absence of a mood episode, patients suffering from metabolic syndrome (i.e., obesity, insulin resistance, diabetes mellitus, or hypertriglyceridemia) perform worse on standardized cognitive tests compared with healthy controls.Reference Huang, Fan and Shi 41 , Reference Tahmi, Palta and Luchsinger 42
A relatively small number of studies have studied the association between cognitive deficits and NAFLD. A cross-sectional study by Celikbilek et al.Reference Celikbilek, Celikbilek and Bozkurt 43 reported lower Montreal Cognitive Assessment (MoCA) scores in NAFLD patients when compared with healthy controls (18.17 ± 5.20 versus 21.08 ± 3.93; p < 0.001, respectively). On further sub-domain analysis, patients with NAFLD did significantly worse on visuospatial and executive items in comparison to healthy controls. Moreover, a negative correlation between liver fibrosis and cognitive measures was observed (i.e., r = − 0.359; P < 0.05).Reference Celikbilek, Celikbilek and Bozkurt 43
Similarly, another study reported that patients with NAFLD were at a significantly higher risk for cognitive impairment than healthy individuals (RR = 3.9, 95%, CI = 1.815–8.381, and p = 0.0005) when assessed through MoCA.Reference Filipović, Marković and Đurić 44 Furthermore, patients with NAFLD performed poorly in the work environment relative to healthy individuals when measured through Work Productivity and Activity Impairment Questionnaire-General Health (WPAI-GH). These patients reported more absenteeism (28.5% versus 12.4%, p = 0.003), presenteeism (33.7% versus 23.0%, p = 0.006), overall work impairment (49.2% versus 30.8%, p < 0.001), and activity impairment (48.0% versus 32.6%, p < 0.001) compared to healthy controls.Reference Balp, Krieger and Przybysz 45 The same study attributed these deficits in work efficiency to compromised mental component scores relative to healthy individuals (39.2 versus 45.2, p < 0.001).Reference Balp, Krieger and Przybysz 45 Further important details of these studies are presented in Table 2.
Pathophysiological nexus between NAFLD and psychiatric disorders
Shared pathophysiological domains of NAFLD and psychiatric illnesses are multifold, complex, and likely to follow a multiple-hit model (i.e., cumulative effect of physical and psychological stress over time in presence of a baseline vulnerability).Reference Rostama, Beauchemin and Bouchard 3 , Reference Arab, Arrese and Trauner 46
Hypothalamic–pituitary–adrenal axis dysregulation and subsyndromal inflammation
The human brain is one of the main centers that mounts homeostatic responses to internal and external changes. The main loop that handles both internal and external stressors is the hypothalamic–pituitary–adrenal (HPA) axis.Reference Papadimitriou and Priftis 47 The detailed description of the HPA axis is beyond the scope of this review; however, a brief overview is provided herein. As the name suggests, this feedback loop comprises three major components: hypothalamus, pituitary gland, and the adrenal glands.Reference Papadimitriou and Priftis 47 The hypothalamic component comprises chiefly the paraventricular nucleus (PVN), which has neuroconnections with all brain regions but primarily with the amygdala, hippocampus, and forebrain.Reference Smith and Vale 48 These afferent projections provide the hypothalamus with input regarding psychological, humoral, endocrine, and visceral information. The hypothalamus connects to the pituitary gland through corticotropin-releasing hormone, which further exerts its action on adrenal secretion of glucocorticoids (namely cortisol) through adrenocorticotropic hormone.Reference Smith and Vale 48 The hormone cortisol modulates the stress response, immune function, glucose metabolism, and executive cognitive functions.Reference Shields, Bonner and Moons 49 , Reference Thau, Gandhi and Sharma 50 Because of this complex interconnected nature of the HPA axis, any dysregulation across it can affect psychological, humoral, endocrine, and visceral domains of the human body.Reference Smith and Vale 48 , Reference Baumeister, Lightman and Pariante 51 .
The HPA axis is also described as a resilience network against routine stressorsReference Daskalakis, Bagot and Parker 52 . The axis is also malleable and keeps on transforming across one’s lifespan. The highest malleability is during the initial years of life (including the prenatal life).Reference Heim and Nemeroff 53 Any early life adverse event (either physical in the form of an illness or psychological in the form of abuse, etc.) can trigger a series of events both in the brain and the body that can lead to a dysfunctional HPA axis later on in life.Reference Heim and Nemeroff 53 , Reference Lecic-Tosevski, Vukovic and Stepanovic 54 This dysfunctional HPA axis can often lead to mounting excessive responses to a threat (that might be just a perceived one at times). The dysregulated HPA response leads to an increased release of cortisol hormone that can lead to further manifestations down the line (e.g., mood alterations, obesity, inflammation, metabolic syndrome, NAFLD, etc.).Reference Dienes, Hazel and Hammen 55 Interestingly, early-life adverse effects and malnourishment are positively associated with NAFLD occurrences later in life, providing further support to this proposition.Reference Shea, Lionis and Kite 56 , Reference Wang, Chen and Ning 57 In accordance with multiple hit theory of pathogenesis, it is important to understand that this whole pathophysiology depends on the genetic vulnerability and persistent surrounding stressors gradually accumulating overtime. Moreover, significant chronic unpredictable stress may also offset the HPA axis later in adulthood.Reference Baumeister, Lightman and Pariante 51 , Reference Kang, Zhao and Ryu 58
We hypothesize that dysregulation in the HPA axis drives the link between NAFLD and psychiatric disorders. It has been observed repeatedly that psychiatric disorders (i.e., schizophrenia, MDD, or BD) often lead to an overly responsive HPA axis resulting in elevated/dysregulated cortisol release.Reference Baumeister, Lightman and Pariante 51 , Reference Daskalakis, Bagot and Parker 52 , Reference Brenner, Liu and Laplante 59 , Reference Steudte, Kolassa and Stalder 60 Not only can an overly active HPA axis cause some symptoms of mood/psychotic disorders, it also leads to significant endocrine and visceral adverse effects. Dysregulated cortisol release has been associated with obesity, inflammation, insulin resistance, and fat dysmetabolism.Reference Baumeister, Lightman and Pariante 51 It is also associated with an increased risk of stroke and cardiovascular disease.Reference Steptoe and Kivimäki 61 All these endocrine effects are closely related to the pathogenesis of NAFLD and its progression to NASH and liver cirrhosis.Reference Brunt, Wong and Nobili 1 , Reference Rostama, Beauchemin and Bouchard 3 , Reference Lonardo 4 This hypothesis is also in line with multiple clinical and bench studies that have found increased cortisol levels among patients with NAFLD repeatedly.Reference Ahmed, Rabbitt and Brady 62 – Reference Von-Hafe, Borges-Canha and Vale 64
As previously mentioned, NAFLD is triggered by the build-up of triglycerides in hepatocytes. These triglycerides are often cleared by kupffer cells (resident macrophages in the liver); however, in presence of fat dysmetabolism the triglyceride load exceeds the clearance capacity of kupffer cells. This leads to build up of triglycerides within the hepatocytes and with time, creates a pro-inflammatory environment that can lead to steatohepatitis (i.e., NASH) and cell necrosis (that furthers the inflammation).Reference Brunt, Wong and Nobili 1 , Reference Rostama, Beauchemin and Bouchard 3 Usually the balance between pro-inflammatory and anti-inflammatory factors is also maintained by the HPA axis.Reference Steptoe and Kivimäki 61 It has been found that persistently elevated or dysregulated cortisol secretion can lead to subsyndromal inflammation in the body.Reference Dantzer, O’Connor and Freund 65 This subsyndromal inflammation can exacerbate the local inflammation within the hepatocytes, catalyzing the whole pathogenetic cycle. This hypothesis predicts that psychiatric disorders might affect a person long before they develop NAFLD or NASH. Moreover, in presence of a psychiatric disorder and related HPA axis dysregulation, the longitudinal risk of developing NAFLD can get significantly higher than in the general population.Reference Soto-Angona, Anmella and Valdés-Florido 66 This also suggests that a psychiatric disorder might not be a co-morbidity in select patients of NAFLD, rather both the psychiatric disorder and the NAFLD might be components of a broader HPA-metabolic syndrome.Reference Mansur, Brietzke and McIntyre 7
Metabolic syndrome and obesity
Metabolic syndrome is a complex phenomenon with polygenic inheritance and multiple causal links. It is strongly associated with NAFLD as well as mood disorders.Reference Soczynska, Kennedy and Woldeyohannes 11 , Reference Papadimitriou and Priftis 47 , Reference Nolan and Prentki 67 Although there are multiple working definitions, metabolic syndrome is usually diagnosed with ≥3 abnormalities in the glucose, high-density lipoproteins, cholesterol, triglycerides, obesity, and blood pressure measurements.Reference Alberti, Eckel and Grundy 68 During the past decade, multiple landmark studies have shown that insulin resistance (as a part of metabolic syndrome) is associated with mood disorders and cognitive dysfunctions.Reference Mansur, Brietzke and McIntyre 7 , Reference Pan, Park and Brietzke 39 , Reference Cetinkalp, Simsir and Ertek 69 Moreover, metabolic syndrome (in particular insulin resistance) has shown the strongest association with NAFLD.Reference Watt, Miotto and De Nardo 70 This could be predictive of insulin resistance as a downstream common pathway to both mood disorders and NAFLD in select number of patients. This pathway makes more sense if we involve the obesity component of metabolic syndrome. In the previous section, we have delineated that obesity plus depression was significantly more associated with NAFLD when compared to depression or obesity alone. This can be explained further with one of our group’s previous work where we suggested the presence of a metabolic-mood syndrome.Reference Mansur, Brietzke and McIntyre 7 In that review, we suggested that patients with mood disorders and obesity are harder to treat, and have more severe symptoms and cognitive deficits. The paradigmatic component behind mood-metabolic syndrome remains dysregulated inflammation.Reference Soczynska, Kennedy and Woldeyohannes 11
Similar to our explanation of the HPA axis that gradually favored pro-inflammatory cytokines (e.g. interleukin (IL)-1ß, IL-6) over anti-inflammatory cytokines, we believe that added obesity acts in a very similar fashion.Reference Ferrante 71 , Reference Liu, Ho and Mak 72 Furthermore, inflammatory markers including C-reactive protein (CRP), tumor necrosis factor-alpha (TNF-α), IL-6, adiponectin, and leptin, which are present in visceral fat potentially aggravate the whole pathogenetic picture (including insulin resistance).Reference Ho, Zhang and Mak 73 Insulin resistance is also associated with less glucose uptake and more lipolysis across peripheral tissues of the body. This leads to excessive circulating free fatty acids (FFA) both within and outside the peripheral adipocytes.Reference Rodríguez-Hernández, Simental-Mendía and Rodríguez-Ramírez 74 In particular, within adipose cells FFA binds to toll-like receptors (similar receptors that bind to microbial antigens) which triggers the inflammatory cascade. Interestingly, this inflammation further increases insulin resistance which in turn increases inflammation leading to a vicious self-sustaining cycle.Reference Chait and Kim 75 Obesity either due to metabolic, physical, or psychological causes leads to further availability of FFA that catalyzes the whole pathogenetic cycle. This proinflammatory environment leads to the progression of NAFLD to NASH and possibly liver cirrhosis. Moreover, the inflammatory hypothesis for depression has been well described in the literature with a significant positive association between the severity of depression and inflammatory markers.Reference Soczynska, Kennedy and Woldeyohannes 11 The potentiating inflammatory phenomenon provides an explanatory framework linking the observation that depression and obesity might be greater risk factors for NAFLD as compared to either of these pathologies alone.
As such, the elevated risk of NAFLD in patients with psychiatric disorders is hypothesized to be due to well-established risk factors that are overrepresented in this population. For example, obesity, diabetes, along with metabolic syndrome differentially affect patients with psychiatric disorders. Factors more specific to the mood disorder population that possibly mediate risk are the higher rate of inflammatory alterations in the mood disorder population and possibly exposure to psychotropic agents that affect weight and/or metabolism.Reference McIntyre, Danilewitz and Liauw 76 – Reference Rosenblat, Cha and Mansur 79
Chronic psychosocial stress and personal factors
It is important to note that the pathogenesis of psychiatric illnesses is very often tightly linked with the environment one lives in.Reference Otte, Gold and Penninx 10 Only recently have we started to take the concept of psychache (i.e., perceived mental pain, unease, or stress due to emotional, personal or social unmet needs) seriously into consideration.Reference Tossani 80 It is important to understand the word ‘perceived’ here making stress a subjective experience completely.Reference Lecic-Tosevski, Vukovic and Stepanovic 54 Although perceived stress is important to adapt and grow as a person in life, chronic unpredictable stress is almost always detrimental to pertinent health quality measures.Reference Hammen 81 A previous animal study revealed that chronic unpredictable stress led to obesity due to leptin resistance.Reference Wang, Yang and Xu 82 Furthermore, the combination of a high-fat diet and stress-induced depression in a similar study was associated with abnormal serum lipid levels and increased inflammatory cytokines in the brain that were possibly mediated by Toll-like receptors - Nuclear factor kappa B (NF-κB) signaling.Reference Wang, Yang and Xu 82 This animal study can predict the direction of stress, obesity, and possibly obesity-related disorders (including NAFLD).Reference Wang, Yang and Xu 82
Amongst people who were apparently healthy, higher perceived stress was significantly associated with the presence of NAFLD (OR = 1.17, 95% CI = 1.11–1.22, p < 0.001). The association became stronger when obesity and perceived stress were taken together (OR = 1.26, 95% CI = 1.19–1.33, p = 0.01).Reference Kang, Zhao and Ryu 58 It has been already known that chronic stress is an independent risk factor for cardiovascular and diabetic morbidity which often overlaps with the pathogenesis of NAFLD.Reference Steptoe and Kivimäki 61 ,83–85 Hence, although the direct cross-sectional studies investigating the association of perceived stress with NAFLD are limited, we can extend the hypothesis of the association of stress with NAFLD in the presence of evidence supporting its significant shared links with cardiovascular and endocrinal pathogenesis.Reference Shea, Lionis and Kite 56
The foregoing possibly connects to our recurring hypothesis of HPA dysregulation among patients with NAFLD. Even in the absence of a formal diagnosis of a psychiatric ailment, HPA axis is involved in the analysis and diffusion of stress on a daily basis.Reference Lecic-Tosevski, Vukovic and Stepanovic 54 A dysregulated HPA axis would point towards a higher degree of perceived stress with apparently normal daily cues.Reference Baumeister, Lightman and Pariante 51 It also leads to compromised resilience at a personal level.Reference Daskalakis, Bagot and Parker 52 The positive associations of chronic stress with NAFLD in one large cross-sectional study and with related metabolic and cardiovascular diseases in a multitude of other studies signify the potential relation of environmental factors, perception of those factors, and resultant morbidities.Reference Kang, Zhao and Ryu 58 , Reference Steptoe and Kivimäki 61 , Reference Lloyd, Smith and Weinger 83 – Reference Tilg, Moschen and Roden 85
Other personal factors that can play shared roles are the levels of physical activity, dietary intake, and long-term medications. More often than not, patients with depression are more likely to eat unhealthy food and have a sedentary lifestyle.Reference McIntyre, Lee and Mansur 86 , Reference Stubbs, Vancampfort and Smith 87 Moreover, patients with mood disorders are also more related to consumption of junk foods that can drive obesity and obesity-related disorders (e.g., NAFLD, insulin resistance, cardiovascular disorders, etc.).Reference Tarantino, Cataldi and Citro 88 This pattern closely relates to the psychomotor retardation, anhedonia, and fatigue that are very common findings with depression.Reference McIntyre, Lee and Mansur 86 Along with dietary habits, this lack of physical activity raises the risk of further metabolic disturbances that may later be predictive of NAFLD.Reference Rodriguez, Torres and Harrison 89 In addition to dietary and activity parameters, patients with schizophrenia face significant morbidity due to persistent negative symptoms (i.e., avolition, anhedonia, flat affect or alogia, etc.) and recurrent positive symptoms (i.e., hallucinations, disordered speech, delusions or disorganized behavior, etc.).Reference Kahn, Sommer and Murray 90 To manage the symptoms, patients with schizophrenia often have to take antipsychotics throughout their lives, with some of them significantly associated with metabolic disruption and metabolic syndrome (i.e., olanzapine, clozapine, or quetiapine, etc.).Reference Pillinger, McCutcheon and Vano 29 , Reference Buchanan, Kreyenbuhl and Kelly 91 The same case can be made for patients with BD who have to take maintenance treatment throughout their lives, often with weight-causing agents.Reference McIntyre, Berk and Brietzke 9 Furthermore, reward pathway deficits are often observed in patients with psychiatric ailments that remain even after the remission of their symptoms. These alterations in reward pathways are predictive of greater substance use (e.g., alcohol, cocaine, and other illicit substances, etc.) that can lead to liver injury and exacerbation of any existing liver inflammation.Reference Tarantino, Cataldi and Citro 88 Hence, on a personal level, psychiatric ailments increase the propensity of an individual toward the aforementioned behavioral risk factors that have been strongly associated with the development of NAFLD.Reference Roshanaei-Moghaddam, Katon and Russo 92
Discussion
Our review underlies the clinical burden of NAFLD among patients with psychiatric disorders. Moreover, the present literature suggests that patients with obesity and a psychiatric disorder (i.e., MDD, BD, or schizophrenia) are significantly further associated with NAFLD compared with either of the conditions alone. We also hypothesize that a dysregulated HPA axis and downstream insulin resistance could be a shared pathophysiological link in a select number of patients presenting with a phenotype comprising both mood disorders and NAFLD (Figure 1).
Pathophysiological flow of shared pathogenesis of mood disorders and NAFLD. (Shared and adapted with the permission of Mansur et al.Reference Mansur, Brietzke and McIntyre 7 )
Patient-specific care
It is important to manage patients with mood or psychotic disorders holistically (i.e., pertaining to physical, metabolic, and emotional parameters) to reach optimum patient-reported clinical endpoints.Reference McIntyre, Lee and Mansur 86 Multiple internal (e.g., vulnerability to metabolic dysregulation) and external (e.g., long-term treatment with metabolically disruptive agents) factors require patients with psychiatric illnesses to undergo strict observation and follow-up for their weight and metabolic parameters.Reference McIntyre, Berk and Brietzke 9 , Reference Soczynska, Kennedy and Woldeyohannes 11 Herein, it is important to select treatments that are safer to use metabolically over a long period of time whenever it is therapeutically possible.
Although many antidepressants (i.e., selective serotonin reuptake inhibitors or serotonin–norepinephrine reuptake inhibitors) have been marketed as weight-neutral agents, they appear to cause weight gain in real-world data in subsets of patients.Reference Gafoor, Booth and Gulliford 93 , Reference Lee, Paz-Filho and Mastronardi 94 However, these data have been criticized to be associative and not causal in nature.Reference Gafoor, Booth and Gulliford 93 Nevertheless, with the usual presence of metabolic disruption among patients with mood and psychotic disorders, it is essential to include metabolic parameters (i.e., cholesterol level, hemoglobin A1c level, fasting glucose, waist and body mass index measurements) in our treatment targets even with weight neutral antidepressants and other therapeutic agents.Reference McIntyre, Lee and Mansur 86 , Reference Gafoor, Booth and Gulliford 93 Not only do these metabolic parameters predict morbidity across various physical organs but they also ascertain the response and remission rates of mood or psychotic disorders, respectively.Reference Luppino, de Wit and Bouvy 95 , Reference Manu, Dima and Shulman 96 The strong association of NAFLD with psychiatric disorders urges us to take metabolic disruption more seriously and to tailor patient-specific treatments that can lead to maximum benefit with minimum to no harm whatsoever.Reference Shea, Lionis and Kite 56
Screening for NAFLD
Currently, there is no screening protocol for NAFLD and most of the time, the diagnosis of NAFLD follows the incidental discovery of deranged liver function tests (i.e., increased aspartate aminotransferase or alanine aminotransferase) in primary care settings. The stage of incidental discovery is often associated with a progressive stage of disease with limited treatments at hand.Reference Caussy 97 Although there is a debate about which patient population might benefit the most from standardized screening guidelines, it remains a unified perspective that screening will help in managing this disease better.Reference Caussy 97 Although the diagnostic criteria for NAFLD is liver biopsy, the screening can be reliably done with a liver ultrasound and measurement of liver function tests taken together.Reference Pandyarajan, Gish and Alkhouri 98 Our review highlights patients with psychiatric illnesses to be at high risk for NAFLD and calls for including them in the patient population who can benefit the most from regular screenings for NAFLD.
Moreover, in the presence of obesity, longer duration of psychiatric illness, higher severity, and family history of NAFLD, ultrasound screening becomes an unmet clinical need rather than a preventive measure.Reference Caussy 97 Further work remains to be done to narrow down patients within the psychiatric cohort who would benefit the most from any standardized screening guidelines.
Future research vistas and potential therapeutics
A few questions still remain unanswered in the current literature. Due to the limited number of longitudinal studies, it is hard to determine a causal link, the duration it takes for NAFLD to develop, and any concomitant aggravating or relieving factors within patients with psychiatric disorders. Moreover, literature is limited with respect to the exact role of medications that are being administered. Furthermore, the specific patient characteristics with metabolic disorders and psychiatric illnesses who might be disproportionately more prone to the development of NAFLD still remain to be determined. Future researchers should design longitudinal studies for a sufficient duration of time to answer these key questions, and further elaborate on the association we strongly discern between NAFLD and psychiatric disorders, especially domains of psychopathology such as anhedonia and cognitive impairment.Reference McIntyre, Cha and Soczynska 40 , Reference Hamer, Testani and Mansur 99 Furthermore, future studies should compare treatment-naive and under treatment patients to see if there is any pattern associated with strict management of affective or psychotic symptoms, choice of medication, and development of NAFLD.
As far as potential therapeutics are concerned, no medication has been approved by the United States Food and Drug Administration to treat NAFLD.Reference Brunt, Wong and Nobili 1 The first line of treatment for NAFLD remains weight loss and multimodal lifestyle modification to decrease insulin resistance (e.g., avoidance of red meat, trans-fat, or highly refined carbohydrates, etc.).Reference Mundi, Velapati and Patel 100 Quite interestingly the drugs that are being used off-label to treat metabolic syndrome and insulin resistance (such as glucagon-like peptide-1 receptor agonists, metformin, and thiazolidinediones) have also shown some efficacy in ameliorating cognitive symptoms and depressive features.Reference Hamer, Testani and Mansur 99 , Reference Mundi, Velapati and Patel 100 Further research is required to assess the effects of select psychotropic medications that demonstrate anti-inflammatory effects (e.g., fluoxetine or fluvoxamine), as well as anti-cytokine agents, and potential therapeutic effects on reducing the severity of NAFLD.Reference Lee, Subramaniapillai and Brietzke 101 , Reference Lu, Ho and Liu 102
Furthermore, it will be beneficial to the field to investigate predictive genetic biomarkers that can aid in identifying subsets of psychiatric patients who are at an increased risk of developing NAFLD. Currently, there is a scarcity of research dedicated to identifying shared genetic pathways between psychiatric disorders and NAFLD.Reference Soto-Angona, Anmella and Valdés-Florido 66 One study in patients with BD reported that a recessive variant genotype (MM versus II/IM) patatin-like phospholipase domain-containing protein 3 (PNPLA3), which is involved in triglyceride metabolism, could increase the odds of developing NAFLD (OR = 4.579; CI = 1.607–13.043; p = 0.0044) in patients with BD.Reference Kenneson and Funderburk 103 MicroRNA (miRNAs) and altered mitochondrial metabolic genes are two other promising shared genetic components that have shown to play a role in energy metabolism perturbations in patients with NAFLD and psychiatric disorders (ie. depression, BD, and schizophrenia) and hence, need to be studied further.Reference Soto-Angona, Anmella and Valdés-Florido 66
Limitations
Our review remains limited despite the strong association between NAFLD and psychiatric disorders due to the cross-sectional nature of most of the studies. Although a few included longitudinal studies strengthened the propensity of a mood or psychotic disorder to have a bidirectional pathogenetic link with NAFLD, further research is needed to test the hypothesis behind the potential cause and effect in disparate psychiatric disorders.
Conclusion
NAFLD disproportionately affects individuals with psychiatric disorders, with risk factors for NAFLD being primarily overrepresented in people with mood disorders. The concurrent diagnosis of NAFLD and a mood disorder predisposes to increased morbidity and significantly decreased quality of life. Future research should attempt to parse mechanistic substrates that underlie NAFLD and mood disorders. In the meantime, clinicians should be vigilant in their assessment, psychoeducation, and treatment of patients with mood disorders and/or NAFLD.
Author Contributions
Conceptualization: M.Y.J., R.B.M., J.D.R., R.S.M.; Data curation: M.Y.J., S.M., A.T., A.M., J.D.D.V., Z.G., N.B.M., L.P., F.C., A.T.H.K., R.R., G.H.L.; Formal analysis: M.Y.J., T.G.R.; R.S.M.; Investigation: M.Y.J.; Methodology: M.Y.J., R.S.M.; Project administration: M.Y.J.; Resources: M.Y.J.; Software: M.Y.J.; Supervision: M.Y.J., T.G.R., R.S.M.; Validation: M.Y.J.; Visualization: M.Y.J.; Writing—Original draft: M.Y.J., T.G.R.; Writing—Review and editing: M.Y.J., S.M., A.T., A.M., J.D.D.V., Z.G., N.B.M., L.P., F.C., A.T.H.K., R.R., G.H.L., R.B.M., J.D.R., R.H.
Disclosures
R.B.M. has received research grant support from the Canadian Institutes of Health Research (CIHR), the Physicians’ Services Incorporated (PSI) Foundation and the Baszucki Brain Research Fund; and support from an Academic Scholars Award from the Department of Psychiatry, University of Toronto. J.D.R. is the medical director of the Braxia Health (formally known as the Canadian Rapid Treatment Center of Excellence and is a fully owned subsidiary of Braxia Scientific Corp) which provides ketamine and esketamine treatment for depression; he has received research grant support from the American Psychiatric Association, the American Society of Psychopharmacology, the Canadian Cancer Society, the Canadian Psychiatric Association, the Joseph M. West Family Memorial Fund, the Timeposters Fellowship, the University Health Network Centre for Mental Health, and the University of Toronto and speaking, consultation, or research fees from Allergan, COMPASS, Janssen, Lundbeck, and Sunovion. was supported in part by the National Institute on Aging (NIA) through Yale School of Medicine (#T32AG019134) in the past 3 years. T.G.R. has also been funded by the NIA and National Institute of Mental Health (#R21MH117438 and R21AG070666) and Institute for Collaboration on Health, Intervention, and Policy (InCHIP) of the University of Connecticut. R.M. has received research grant support from CIHR/GACD/National Natural Science Foundation of China (NSFC) and the Milken Institute; speaker/consultation fees from Lundbeck, Janssen, Alkermes, Neumora Therapeutics, Boehringer Ingelheim, Sage, Biogen, Mitsubishi Tanabe, Purdue, Pfizer, Otsuka, Takeda, Neurocrine, Sunovion, Bausch Health, Axsome, Novo Nordisk, Kris, Sanofi, Eisai, Intra-Cellular, NewBridge Pharmaceuticals,Viatris, Abbvie, Atai Life Sciences. Dr. Roger McIntyre is a CEO of Braxia Scientific Corp.
