Non-alcoholic fatty liver disease (NAFLD) as a neglected metabolic companion of psychiatric disorders: common pathways and future approaches

Psychiatric patients suffer from a substantial reduction (ranging between 7 and 20 years) in life expectancy, with high rates of all-cause mortality [1]. It is estimated that around 60% of this excess of mortality is due to medical comorbidities, especially cardiovascular diseases [2]. In this regard, metabolic syndrome (MetS), defined as the combination of abdominal obesity, high blood pressure, low high-density lipoprotein cholesterol, elevated triglycerides, and hyperglycemia, plays an important role, since it is a major risk factor for the development of diabetes and cardiovascular disease [3]. A significant association exists between major psychiatric disorders, such as bipolar disorder, depressive disorder or schizophrenia, and MetS, partly due to psychotropic medication use (e.g., atypical antipsychotics) and an unhealthy lifestyle [4]. In a worldwide survey conducted among more than 47.000 individuals across 17 countries, specific mental disorders were associated with 1.5–13.3% of physical condition onsets [5]. Recently, in a Danish cohort of 5.9 million people, there was a median hazard ratio of 1.37 of developing a medical condition if a psychiatric disorder was present [6].

In this context, non-alcoholic fatty liver disease (NAFLD) is an entity characterized by excessive hepatic fat accumulation. This is defined as the presence of steatosis in > 5% of hepatocytes according to histological analysis [7, 8]. It is associated with insulin resistance and it was considered to be the hepatic manifestation of the MetS. Recent evidence suggests a complex, bidirectional relationship between NAFLD and cardiometabolic diseases [9].

The distinction of NASH from simple steatosis is key. Steatosis due to secondary causes, such as viruses, autoimmune responses, metabolic or hereditary factors, and drugs or toxins, should be ruled out (Table 1). Classically, the diagnosis of NAFLD required the exclusion of alcohol abuse (daily alcohol consumption greater than 30 g for men and 20 g for women) along with secondary causes for NAFLD (viral pathologies, steatogenic medications, or other monogenic hereditary disorders). However, the rising prevalence of NAFLD makes its coexistence with other chronic liver diseases quite possible. Therefore, a positive diagnosis rather than a diagnosis based on exclusion of concomitant diseases has been proposed [11], and a dual etiology for fatty liver disease is considered possible and even frequent [12].

Ultrasonography plays a key role in the determination of steatosis, and a step-by-step approach is preferred [13]. The precise global incidence of NAFLD is unknown, due to the complex screening required, wide variations in the enrollment of differing populations, racial groups, exclusion criteria for alcohol consumption, study design, and diagnosis methods [14]. However, its prevalence is thought to be on the rise, ranging from 20–30% in Western countries to 8–20% in Africa [15, 16]. There are many common medical comorbidities among NAFLD and most psychiatric disorders, including cardiovascular and endocrine-related diseases such as MetS, diabetes mellitus, dyslipidemia, arterial hypertension, obesity, hypothyroidism, hypogonadism, polycystic ovary syndrome, osteoporosis, obstructive sleep apnea syndrome (OSAS), chronic kidney disease, psoriasis, or cancer [17‐20] (Fig. 1). Taking into account its disease burden and growing prevalence, and the neglected relationship of NAFLD with MetS and psychiatric disorders, there is a need to better understand this disease and its commonalities with psychiatric illnesses, in order to develop proper and tailored assessment and test interventions for this prevalent but neglected comorbidity.

To gather data on the subject, a search on PubMed with the terms “([Title/Abstract]NAFLD OR [Title/Abstract]NASH) AND psych*” was performed. We also conducted a ScienceDirect search with the terms “NASH, NAFLD, psychiatry.” A total of 92 results were obtained, of whom only those pertaining the specific link of NAFLD, or NASH, and psychiatric disorders, were included, leaving a total of 11 publications. The small number illustrates the sparsity of the literature describing NAFLD in psychiatry. We also explored secondary references found in the initial bibliography. A secondary search of reviews and data on the physiopathology of NAFLD, and on relevant subjects such as general morbidity of psychiatric disorders, was also performed to give the reader an overview on the underlying mechanisms of this disease and to explore possible commonalities not yet investigated.

NAFLD includes two distinct conditions with different prognoses: non-alcoholic fatty liver (NAFL) and non-alcoholic steatohepatitis (NASH). NAFL is defined by steatosis, meaning the presence of lipidic vacuoles in hepatocytes. NASH involves steatosis plus inflammation, with signs of hepatic disease, such as ballooning, and might involve fibrosis [21]. NASH has a higher risk of cirrhosis and hepatocellular carcinoma [22]. NASH carries a high risk of both liver related morbidity and mortality as well as metabolic comorbidities (ten times higher than the general population), cardiovascular disease and mortality (twice higher than the general population), and cancer (particularly bowel and breast cancer) [17]. Around 20% of patients suffering from NASH will develop cirrhosis [16], although it is important to note that hepatocellular carcinoma can occur in the absence of cirrhosis [23], making NASH the most rapidly growing indication for liver transplantation in patients with hepatocellular carcinoma in the USA [24].

In a recent meta-analysis, the pooled population estimated prevalence of NAFLD was 24–25%, although it varied across countries. Currently, it is thought to occur in around 24% in the general population of Europe and the USA, 27% in Asia, 30% in South America, and 32% in the Middle East. In Africa, prevalence estimates are lower, at around 13%. It is worth noting that this data refers to diagnosis made by imaging, mainly through ultrasound measuring hepatic fat, as studies using blood tests (measuring ALT, AST, and platelet count) report notably lower prevalences [7, 8, 16]. The regional NASH prevalence estimates among NAFLD patients with an indication for biopsy were around 63% for Asia, 69% for Europe, and 61% for North America. On the other hand, NASH prevalence estimates among NAFLD patients without an indication for biopsy were 7% in Asia and 30% in North America [16].

Both NAFLD and NASH are more prevalent among males [19]. Gender differences also exist in most psychiatric disorders, such as schizophrenia, which is more common in males, and depression, which is more common in females [25‐27]. On the other hand, racial and ethnic variations exist in NAFLD and NASH; in the USA, the prevalence of NAFLD is highest among Hispanic and lowest among Afro-American populations [28].

While the total prevalence of liver disease in patients with psychiatric illness is not fully known [29], the prevalence of MetS in schizophrenic or bipolar patients is remarkably high, ranging from 22 to 42% [30], compared to non-psychiatric control populations (with an estimated prevalence of 15–20%). A recent meta-analysis found patients suffering from severe mental illnesses to have a pooled RR of 1.58 (1.35–1.86) of developing MetS [2]. Hence, people suffering from mental health disorders are thought to have an increased incidence of NAFLD, although data are scant [18]. In this regard, several studies and publications have aimed to evaluate its comorbidity with specific psychiatric disorders.

Most research in NAFLD in psychiatry focuses on affective disorders. In subjects suffering from NASH, Elwing et al. observed a higher rate of lifetime major depression (MDD) and generalized anxiety disorders (GAD) as compared to the control population. Comorbid anxiety in patients is linked to more advanced liver histological abnormalities [31]. In patients with NAFLD, there is also evidence pointing for more severe hepatic ballooning when they have MDD [32]. It is important to note that these studies examined affective disorders in liver disease patients. The other way around, examining depressed patients, Weinstein et al. found higher prevalence rates and severity of NASH compared to other chronic liver diseases [33]. Two studies among bipolar patients also indicated a higher prevalence of chart-reported NAFLD and NASH [34, 35], pointing to risks and mechanisms beyond lifestyle or drug use, although no sustained explanatory theory has yet been suggested. Unhealthy lifestyle habits are disproportionally present in affective disorders, where inflammatory and metabolic disorders are similarly frequently found.

A higher prevalence of NAFLD in schizophrenia is likely [35, 36]. Yan and colleagues, in a large cross-sectional study comparing young males with schizophrenia with general young males, found a significantly higher prevalence of NAFLD of up to 49.5% in the study group compared to 20.1% in the control group. They found the factors most correlated with NAFLD were triglyceride levels, BMI, medication combinations, drug dosage, and negative factor scores on the PANSS [37].

Anorexia nervosa is also associated with significant liver complications, as fasting can lead to steatosis [38]. However, we found no study assessing the specific prevalence of NAFLD in these patients. Regarding other psychiatric disorder populations, such as those with anxiety, OCD, or PTSD, we are not aware of specific studies examining the comorbidity with NAFLD/NASH.

NAFLD, diabetes mellitus type 2, and obesity are thought to share similar risk factors and pathophysiological pathways (Fig. 2). Since the liver is involved in the majority of metabolic routes, NAFLD can be considered a consequence of disturbed metabolism, observed with insulin-resistance and metabolic inflammation [22]. NAFLD develops through the induction of hepatic lipogenesis which, in turn, stimulates dysregulated secretion of proinflammatory cytokines and adipokines; hence, lipid storage in the liver is increased. A complex interaction between genetic, environmental, and epigenetic factors interplays in the genesis of this disease. Importantly, these factors share interesting commonalities with psychiatric diseases. They are not only important to the onset of NAFLD, but also to its progression to NASH, and have an influence in its natural history and severity [22]. A multiple-hit model has been proposed, implying multiple insults acting together on genetically predisposed subjects to induce NAFLD [39]. Similarly, explanatory models for psychiatric disorders are moving towards this conception as well [40].

A number of harmful life habits such as a sedentary life style or an excessive calorie intake have a strong correlation with the development of this disease, as well as an excess intake of saturated fatty acids, polyunsaturated omega 6 fatty acids, and industrial fructose [22]. Low levels of physical activity, and poor diet quality including increased emotional eating (eating in response to negative emotions), have been linked to depressive symptoms [41]. However, unhealthy lifestyles and obesity are not the sole factors in NAFLD genesis, as a relevant percentage (25–30%) of individuals with obesity might present with a lower risk of these alterations (the so-called healthy obesity), whereas there is a subset of non-obese individuals that can suffer from MetS [42]. However, it is relevant to note that healthy obesity still involves a higher risk of developing NAFLD than normal weight [43].

In psychiatric patients, the effects of medication also need to be taken into account, as psychiatric treatments may have specific side effects on the liver. They also contribute greatly to weight gain, and the MetS itself [29]. Moreover, there seems to be an inherent risk of developing MetS in relation to psychiatric illness that is also present in non-treated individuals [2]. Diabetes and insulin resistance have also been associated with an increased risk of depression and anxiety [44]. Recent publications point to depression as an independent risk factor for obesity [45], and this relationship might be bi-directional [46, 47]. Alterations in cortisol, which are more frequent in psychiatric patients, can also promote insulin resistance, and pro-inflammatory cytokines like tumor necrosis factor-α and interleukin-6 are involved in depressive disorders and NAFLD/NASH [48].

The best and most effective treatment for NAFLD and NASH remains a change in lifestyle, promoting a better diet, weight loss, and physical activity [7] (Table 3). In the early stages, a weight loss of 5–8% and a healthy diet might be sufficient. In a recent systematic review, NAFLD/NASH was even proposed as a cognitive-behavioral disease, as lifestyle changes are its most effective management strategy [100]. However, in more advanced stages of liver disease, pharmacological treatment, mainly insulin sensitizers, and more aggressive approaches such as bariatric surgery, might be needed [86]. Exercise, healthy diet, and weight loss also substantially improve the outcome of several psychiatric illnesses, improving factors such as cognitive functioning, negative symptoms, depression or anxiety, and have a beneficial impact on common pathways like inflammatory regulation [99, 101].

Among people with psychiatric disorders, even though having NAFLD does not necessarily mean impairment of hepatic function, it is important to bear in mind that a vast number of psychiatric drugs are hepatically metabolized, and their half-life, secondary effects profile, and metabolism should be accounted for [29]. As noted before, some medications are much more associated with the MetS than others.

Some therapeutic approaches are proposed to target the metabolic imbalance that underlies NAFLD. They might also prove beneficial for psychiatric disorders, based on the common mechanisms mentioned before. For example, new insulin sensitizers targeting mitochondrial pathways are being tested for this purpose. They could improve the metabolic pathways leading to diabetes mellitus type 2, inflammation, and oxidative stress and hence could be of help in both entities [102]. Some insulin sensitizers may have value for psychiatric disorders as well, although data remain preliminary [103]. To target oxidative stress, vitamin E has recently been tested as a treatment with promising results [104]. Based on the correlation between poor life style and depressive symptoms, as mentioned before, antidepressants might also be considered treatment for NAFLD, although the potential secondary side effects affecting metabolism of some agents should be taken into account [105, 106].Statins are also interesting drugs for treating both NAFLD and psychiatric illnesses. They have a role in diminishing free cholesterol and have been found to protect against histological damage [107]. As they have anti-inflammatory and anti-oxidative effects, they have also been proposed as adjuvant therapies for a number of psychiatric disorders [108]. Another interesting example of common pathways and treatment can be found in the role of cannabinoid 1 receptor, which modulates the hepatic energy state and food intake, and could play a role in the development of hepatic steatosis [109]. The role of cannabinoid receptors in the development of psychiatric disorders is being widely studied as well, although the evidence remains scant and caution is advised [110].

As mentioned before, NAFLD and NASH have a high prevalence in the general population, but this is even higher in psychiatric patients. There is a need for NAFLD and NASH to be taken into account when assessing such patients, and their specific impact needs to be further investigated. However, as NAFLD is frequently asymptomatic, and is not detected with current psychiatric medical screening protocols, it is likely grossly underdiagnosed and disregarded. On the other hand, the current criteria are based on ruling out other possible causes, whereas, based on the knowledge on the pathophysiology of the disease, new nomenclature based on specific causes has been proposed, aiming for better and more precise definition and diagnosis [111]. The notion of metabolic dysfunction-associated fatty liver disease (MAFLD) has recently been suggested, involving positive criteria that relate to the probable metabolic-inflammation impairment that underlies this pathology [11].

Diagnosis is currently based on imaging and histology. A widely used surrogate index, the fatty liver index (FLI) [112], might be useful in detecting steatosic changes, but cannot be used for the diagnosis of NASH. To consider a diagnosis of NAFLD, and refer a person for imaging (liver ultrasound), the patient should present with known risk factors, symptoms of fatigue, abdominal pain, or show an abnormal liver function in the liver test. NAFLD diagnosis requires either imaging or histologic demonstration of more than 5% hepatic steatosis in the absence of excessive alcohol consumption. Other possible causes of steatosis should also be ruled out, such as hepatitis C infection or hypothyroidism [105]. In contrast, a NASH diagnosis requires a liver biopsy with histologic examination, which is not feasible, cost-effective, or necessary in every patient with NAFLD. These methods can be invasive and are difficult to implement in the general population for screening purposes. In this regard, recent advances in ultrasonography techniques might allow for rapid, non-invasive, and cost-effective alternative methods, and some instruments such as the ultrasound fatty liver indicator (US-FLI) have proven to yield a high predictive value for NASH diagnosis [113]. Currently, biological markers such as the FLI are only considered appropriate for large epidemiological studies [11].

There is also a lack of reliable biomarkers [17], although there are some efforts made in this direction, based on mechanisms mentioned before, like inflammatory profiles [114] or miRNAs [60], that could help in aiding an early diagnosis and more closely monitoring the progression of the disease. Some algorithms based on non-invasive methods, and aiming for wider screening, are proposed, as the global awareness of this disease is increasing [19]. Non-invasive scoring systems can estimate the degree of fibrosis without biopsy, helping to assess the risk of progression to NASH and cirrhosis. The Fibrosis-4 (FIB-4) Index, incorporating platelet count, age, AST, and ALT, is the most commonly used and has been validated for NAFLD and NASH [115, 116].

Some questionnaires are also being validated for monitoring impact and progression, like the Liver Disease Questionnaire [117]. Nevertheless, greater awareness and suspicion of the disorder in at-risk cohorts should prompt further investigation. In research, adequate and standardized measures for non-invasive diagnosis and measuring of progression should be implemented, in order to increase the external validity of the results and enhance our knowledge of the real incidence and prevalence of these diseases.

In Fig. 3, we propose a practical diagnostic algorithm for NAFLD and NASH tailored for psychiatric populations, based on clinical experience and the American Association for the Study of Liver Diseases (AASLD) guidelines [8, 19].

Prevalence studies show higher rates of NAFLD and NASH among people suffering from psychiatric disorders, suggesting common risk factors and pathophysiological links between these entities, with the MetS at their core. Although few studies have been performed to assess this relationship, common factors such as genetics, mitochondrial dysregulation, and dysbiosis have been found, strengthening this theory. NAFLD might be part of a complex system of psychiatric and non-communicable medical disorders with a common pathogenesis based on dysregulation of inflammation, redox pathways, and mitochondrial biogenesis (Fig. 4).

There is a knowledge gap that needs to be filled in order to proper understand the relationship of psychiatric illnesses with NAFLD and its impact. Future research should target the common underlying risk factors and biological pathways. Although some treatments have been found and seem promising, their compound effectiveness is yet to be evaluated. Also, new and more efficient screening tools, tailored for psychiatric patients, who seem to be at a higher risk of suffering this disease, need to be developed in order to assess the real impact of this disease. Clinicians should be aware of this entity, as its presentation might be asymptomatic at onset, but an early diagnosis could yield benefits in terms of prognosis and quality of life.