Alanine aminotransferase-old biomarker and new concept: a review.

Measurement of serum alanine aminotransferase (ALT) is a common, readily available, and inexpensive laboratory assay in clinical practice. ALT activity is not only measured to detect liver disease, but also to monitor overall health. ALT activity is influenced by various factors, including viral hepatitis, alcohol consumption, and medication. Recently, the impact of metabolic abnormalities on ALT variation has raised concern due to the worldwide obesity epidemic. The normal ranges for ALT have been updated and validated considering the metabolic covariates in the various ethnic districts. The interaction between metabolic and demographic factors on ALT variation has also been discussed in previous studies. In addition, an extremely low ALT value might reflect the process of aging, and frailty in older adults has been raised as another clinically significant feature of this enzyme, to be followed with additional epidemiologic investigation. Timely updated, comprehensive, and systematic introduction of ALT activity is necessary to aid clinicians make better use of this enzyme.

1. Introduction

Serum alanine aminotransferase (ALT) is a readily available, inexpensive, and routine biochemical assay used in clinical practice 1. Initially, many concerns were voiced regarding the clinical significance of ALT activity on viral and toxic hepatitis, muscular dystrophy, and other muscular diseases 2, which might cause a substantial increase in the ALT level 3 in a relatively low percentage of the overall population. However, more and more metabolic disorders, such as obesity, hyperlipidemia, and diabetes mellitus (DM) have been observed independently associated with mild-to-moderate ALT elevation 4. The series of metabolic disorders were referred to metabolic syndrome (MetS) 5, and featured insulin resistance and obesity. The clinical implication of ALT elevation in representing MetS has caused worldwide concern 4, 6, 7 in Western and Eastern countries with the rapidly increasing prevalence of obesity 8. Currently, ALT measurement is not only widely used in detecting the incidence, development, and prognosis of liver disease with obvious clinical symptoms, but also provides reference on screening the overall health status during health check-ups 1, 9.

Some demographic factors, such as gender and age, might also interfere with the ALT level in the general population 10-13. Even the laboratory method and the diurnal variation might interfere with the ALT level, which has been discussed in previous studies 14-16. For a comprehensive association linked to many disease processes, co-morbidities, and concurrent diseases, the clinicians should carefully evaluate and explain the clinical implication underlying the abnormal ALT level when faced with an individual patient. The relative confounders should be adjusted. Thus, the overview provided in this paper was intended to summarize the clinical factors that influence the ALT level based on previous reports and provide better knowledge of this liver enzyme as a tool in clinical estimation.

2. Chemical Formation and Physiology of ALT

ALT is an enzyme that is mainly aggregated in the cytosol of the hepatocyte 17. ALT consists of 496 amino acids, has a half-life of 47±10 hours 18, 19, and is coded by the ALT gene, which is on the long arm of chromosome 8 19. Physically, the ALT enzyme catalyzes the transfer of amino groups from L-alanine to α-ketoglutarate, and the converted products are L-glutamate and pyruvate (Figure 1) in the liver, which is a critical process of the tricarboxylic acid (TCA) cycle. In this process, the coenzyme, pyridoxal phosphate, is required 17, 20.

Figure 1

The transaminationreaction catalyzed by alanine aminotransferase.

ALT is mainly aggregated in the cytosol of the hepatocyte. ALT activity in hepatic cells is approximately 3000 times higher than serum ALT activity 17. When liver injury occurs, ALT is released from injured liver cells and causes a significant elevation in serum ALT activity. ALT also exists in muscles, adipose tissues, intestines, colon, prostate, and brain 21, 22; however, the concentration of ALT in these organs is much lower than the liver 1, 23, 24.

3. Physiologic Factors Associate with the Serum ALT Level

The impact of some physiological factors has been shown to be associated with the serum ALT level. Extreme physical exertion can induce a short-term, reversible elevation in ALT. In a study focused on Thai boxers, the ALT level was 2-2.25-fold higher than the baseline value after a fight in 20 male adolescent Thai boxers 25. In addition, an elevation in ALT was also observed in 9 runners who took part in a 1600 km ultra-marathon 26; specifically, a 4-fold increase in ALT activity was observed on the 4th day of running compared to the baseline value. Many authors attribute the elevation in ALT after physical exertion to muscle injury 27-29. In contrast, modest physical activity is helpful in normalizing ALT levels 30, 31. In addition, the serum ALT values in those who exercise at customary levels were lower than those who do not exercise or exercise more strenuously than customary 3, 32. ALT activity has a diurnal variation in a given population. Co'rdoba et al 14 observed a circadian variation in ALT activity, with the nadir value at 04:00h, and the peak value at 16:00h in 12 patients with chronic liver disease. All physiologic confounders should be taken into consideration and the outside environment should be standardized when comparison of ALT activity is performed.

4. Demographic Characteristics Influence the Serum ALT Level

Several demographic covariates may contribute to disparities in the ALT level. Specifically, age has been shown to be associated with the ALT level in prior cross-sectional and prospective studies 11-13. Elinav et al 12 reported an inverted curve with peak ALT levels at 40-55 years in a group of Israeli participants. Further, Dong et al 11, 13 reported a significant decrease in a large sample of community participants through cross-sectional and prospective studies.

In addition, gender plays a crucial role in modulating discrepancies in ALT activity. In a study focused on blood donors without a diagnosis of hepatitis from London, the ALT level was much higher in males than females 33. Many population-based health examinations have also shown significant discrepancies between genders under similar living conditions 4, 34-36. Some scholars have attributed the gender-based differences in ALT levels to the hormonal differences between males and females 10.

Ethnicity difference in ALT levels has also been observed. In the Third National Health and Nutrition Examination Survey (NHANES III), Mexican-Americans had a higher prevalence of ALT elevation compared to other ethnicities 4, 37. Indeed, the increased prevalence in ALT elevation might be correlated with the higher prevalence of metabolic syndrome (MetS) 37, 38, which is considered to be a major cause of ALT elevation in Mexican-Americans 6, 39, 40.

5. Clinical Factors Associated with Serum ALT level

5.1 Hepatic-related causes

5.1.1 Viral hepatitis (mainly hepatitis B virus [HBV] and hepatitis C virus [HCV] infections)

Viral hepatitis infection is the leading or secondary cause of ALT elevation in populations worldwide 4, 35, 36. ALT activity is an indicator of liver injury in patients with acute and chronic viral hepatitis 41.

With respect to HBV infection, ALT elevation is often observed in the process of the cytolytic immune response (acute phase) and the following ineffective HBV clearance (chronic phase) 42. Liaw et al 42 have described a fluctuation in ALT activity during the process of HBV infection. ALT activity is a crucial reference indicator in treatment selection and the evaluation of prognosis in patients infected with HBV 42-44. Nevertheless, controversy exists and Lai et al 45 reported significant fibrosis and inflammation in 37% of patients infected with HBV and persistently normal ALT levels 45. Unlike HBV infections, the ALT level is less meaningful for diagnosis and prognosis of HCV treatment. More patients infected with HCV progress to chronic hepatitis with persistent hepatocyte injury 41. Greater than 6 in 10 of common HCV carriers have normal ALT levels or mildly elevated ALT levels (< 2 times the upper limit of normal [ULN]) with rare hepatic histologic lesions confirmed by liver biopsy 46. Recently, Ruhl et al 47 suggested that lowering the ULN of the ALT level (29 IU/L for men and 22 IU/L for women) was the best cut-off value to identify HCV infectors in the US population with a high prevalence of HCV infection. Otherwise, the HCV RNA titer is closely linked to the ALT elevation. Cathy et al 48 reported that approximately 68% of patients with positive HCV-RNA levels have ALT elevations in asymptomatic blood donors who tested positive for antibodies to the HCV (anti-HCV).

5.1.2 Alcohol intake

Excessive alcohol intake is another cause of ALT elevation in the general population. In an Italian population, 45.6% of altered liver tests have been attributed to excessive alcohol intake (≥28 g/day) 35, while in a US national population survey, excessive alcohol (>1 time/day) is a crucial cause of ALT elevation, second only to HCV infection 4, however, alcohol intake might be a time-and dose-dependent covariate that influences ALT activity. Short-term and light alcohol consumption was not shown to induce significant ALT elevation in adults 49, 50, however, ethnicity differences exist regarding the biological consequences of alcohol abuse 51, 52. In a UK study based on alcohol abusers from different areas worldwide, adults from South Asia were shown to be more susceptible to alcohol-related liver damage and the ALT levels were higher than in European alcohol abusers 51. Indeed, the effect of mild alcohol intake on ALT activity can be distinguished from binge drinking. Moderate alcohol intake does not contribute to significant ALT elevation, especially in a normal weight population, due to its potential effect on improvement of insulin sensitivity 53-55.

5.1.3 Some medications

Pratt et al 9 listed the medications that might cause ALT elevations. A randomized controlled trial (RCT) indicated that the estimated odds ratios (ORs) of ALT elevation in active treatment groups (including acetaminophen, hydromorphone+acetaminophen, morphine+acetaminophen, and oxycodone+acetaminophen) were 2.57-3.08 compared to the placebo group involving 343 healthy participants, even at the recommended dose 56. Another commonly used medication, statins, also causes mild ALT elevation 57, 58. The mechanism underlyingstatin-associated ALT elevation is still unclear. Some scholars have suggested that the ALT elevation in statin users is attributed to cholesterol reduction in hepatocytes and co-morbid conditions, rather than liver damage or dysfunction 59, 60. Therefore, the long-term medications should be carefully considered when faced with an unexplained ALT elevation. Pratt et al 9 also indicated that cessation of drug treatment is the best way to confirm the relationship between a drug and ALT elevation.

5.1.4 Coffee consumption

Of note, coffee intake might be a protective factor against ALT elevation. In NHANES III, there was a 50% and 70% decrease in ALT elevation amongst participants who consumed >2 cups of coffee/day or ≥373 mg of caffeine, respectively, compared to participants who did not consume coffee 61. Lee et al 62 attributed the protective effects of caffeine to antioxidant activity.

5.1.5 Non-alcoholic fatty liver disease (NAFLD)

NAFLD is a spectrum of clinical and pathologic changes, from fatty liver alone to steatohepatitis 63. NAFLD is common in asymptomatic patients, and the prevalence ranges from 10 to 24% worldwide 64. Considered as a manifestation of the MetS in liver 65, 66, NAFLD has been strongly associated with ALT activity in previous studies 4, 35, 36, 67. NAFLD is the common cause of unexplained mild ALT elevation 3, 68. NAFLD in asymptomatic patients is often serendipitously detected by liver biochemistry testing during routine health check-ups 69, 70. Similar to the increasing prevalence of obesity 8, NAFLD is increasing and becoming a major health burden 71. In spite of the non-linear correlation between the degree of ALT elevation and the histologic severity of NAFLD 72, 73, a mild ALT elevation is largely attributed to NAFLD. When faced with an unexplained ALT elevation (without viral hepatitis or a history of excessive alcohol intake), NAFLD should be considered in the differential diagnosis.

5.1.6 Autoimmune hepatitis

Autoimmune hepatitis is a less common liver disease than NAFLD 74, and the mechanism underlying autoimmune hepatitis is still unknown 75. ALT elevation is an available auxiliary measurement in the diagnosis of autoimmune hepatitis 76.

Moreover, ALT activity is a crucial indicator in detecting the effect of immunosuppressive treatment, prognosis, and long-term survival in patients with autoimmune hepatitis. In a study based on 84 Japanese autoimmune hepatitis patients, a persistently low ALT level (≤40 U/l) was the threshold value that was associated with improved prognosis 77. Another study also showed an association between persistent ALT elevation and poor survival in 69 autoimmune hepatitis patients 78. ALT is considered to be a crucial non-invasive marker of inflammation in patients with autoimmune hepatitis 79.

5.2 Non-hepatic cause

5.2.1 Metabolic covariates

Except for apparent causes, such as viral hepatitis, alcohol intake, and some medications, the so-called unexplained causes of ALT elevation in some previous studies have mainly been attributed to MetS 4, 68. Similar to the pandemic of obesity 8, MetS presents as a series of metabolic disorders, including glucose intolerance, central obesity, dyslipidaemia, and hypertension, has caused worldwide concern in the most recent decades.

Compared to the obvious cause of ALT elevation referred above, ALT elevation caused by MetS is mild and neglected. In a cross-sectional study, the ALT level in MetS patients, as defined by the National Cholesterol Education Programme Adult Treatment Panel III (NCEP-ATP-III) criteria, was approximately 30% higher than participants without MetS in a male population from south China 7. This impact of MetS on ALT elevation, however, is progressive and cumulative with a linear trend 6, 34, 39. Even within ULN values, which did not consider the impact of MetS and are higher than the updated values, the increasing prevalence of MetS is still correlated with the increasing ALT level in the general population. In a community-based Korean population, the ORs for MetS in the highest quintiles of ALT were 7.1-fold higher than the reference quintile in men and 2.1-fold higher in women 80. All of the enrolled participants were selected within the ULN values (the ULN value is 30 U/l for males and 19 U/l for females) 80. Another Korean national health survey also showed a significantly increased prevalence of MetS components, as defined by NCEP-ATP-III criteria, in the subgroup with high-normal ALT levels 81. With respect to the lipoproteins, ALT was shown to be stably and significantly associated with intermediate-density lipoprotein (IDL) and apolipoprotein B (ApoB) after adjusting various covariates in different models 82. These associations were commonly attributed to the stable and independent effects of insulin resistance and fatty liver disease in subjects with ALT elevations 6, 49.

In addition, the impact of MetS components on ALT activities varies to some extent. These distinguishing effects emerged after logistic regression using all of the MetS components as covariates. The impact of the individual MetS components on ALT elevation was disproportionate 7. The body mass index (BMI) and waist circumstance (WC), representing the central obesity component of MetS, were more closely linked to the ALT elevation 4, 7, 36, 67, 83, 84, although, the underlying mechanism has not been elucidated. The possible explanation is that obesity, especially abdominal fat, is potentially involved in the visceral adipose deposition that causes hepatotoxic fatty acids 85. In a US national population-based study 86, BMI lost significance when evaluating the association between obesity and ALT abnormities after adjusting for leptin, insulin, and triglyceride concentrations, rather than the WC, which was representative of the visceral adipose deposition, indicating that BMI might be an intrinsic association between obesity and ALT abnormalities. Another viewpoint was that the obesity-ALT elevation association was modulated by insulin resistance (IR). In a national health survey from a Korean adolescent population, the prevalence of IR status was positively correlated with the degree of obesity 87. The OR of obesity-induced ALT elevation was significantly decreased after adjusting the homeostasis model (HOMA-IR) as an index of IR status 88. Despite the ill-defined intrinsic mechanism, central obesity is the crucial MetS component that most influences the ALT level in general population.

5.2.2 Celiac disease and muscle injury

Chronic ALT elevation is also found in several non-hepatic disorders, such as celiac disease and muscle injury 27, 89, 90. The intrinsic mechanism between celiac disease is not known. Approximately 40-57% of patients with celiac disease have abnormal liver tests 91-95. Celiac disease patients with elevated ALT levels should be treated with gluten-free diet; doing so will restore ALT levels to normal in 75-95% of patients within 6 months 92, 94, 95. It is well known that ALT elevation is often observed in patients with muscle necrosis 96 and the ALT elevation without evidence of liver disease should be considered due to muscle injury 27. Otherwise, the increased creatine kinaseand lactate dehydrogenase activities that occurred following muscle injury should be assayed to identify the cause of ALT elevation 9, 27.

5.2.3 Hemochromatosis

As an ethnic specific disease mainly occurring in individuals of Nordic descent, hemochromatosis is a less common cause of ALT elevation. HFE gene mutations are the major cause of hereditary hemochromatosis, and iron overload is the main cause of hepatic injury. Measurements of serum ferritin, total iron-binding capacity, and the HEF mutation test can help diagnose hemochromatosis 97. Liver biopsy might be necessary if the aforementioned tests are negative in patients highly suspected to have hemochromatosis to evaluate the status of liver injury from iron overload.

6. Interaction on ALT Level between Distinguished Covariates

Except for the independent univariate impact, ALT fluctuation has also been shown to be influenced by multivariate interaction in several studies 54, 55, 84. Among 13,580 US participants, after excluding the patient with hepatitis B or C infection, or iron overload, obesity significantly increased the risk of alcohol-related abnormal aminotransferase activity. The prevalence of abnormal aminotransferase activity was increased from <5% in the normal weight group to near 30% in the obesity subgroup 54. Similarly, these effects were also observed in the Finnish population 55. Piton et al 84 also described the interaction of ALT activity by neural network and recommended the distinguishing ULN of the ALT level to be classified by BMI and gender. In addition, the age and gender interaction was also observed and expressed in mathematical formulas 12, 98. In contrast, our prior cross-sectional study did not show a synergistic effect between MetS and HBV infection, which were also considered as critical covariates associated with ALT elevation 7.

7. Several Concerns that Clinicians should Raise

7.1 Age as a critical covariate that should be emphasized

As a covariate influencing ALT activity, the association between age and ALT activity has been described in previous studies 11-13. The impact of age on ALT fluctuation is not only present based on quantitative discrimination, but also on the diagnostic value in predicting all-cause and disease-specific mortality. A previous authoritative review 1 indicated that elevated ALT might predict higher mortality in a general population; however, this opinion might be questionable, especially in old adults. Inverse relationships between ALT activity and mortality were observed in several studies focused on older population 57, 99, 100, and this opinion was confirmed by subsequent meta-analysis 101. Dong et al 11, 13 raised the concept that ALT activity might be influenced by accelerated aging and frailty in older adults independent of its traditional role in screening liver function. Although age lost significance in many multi-covariate analyses 7, 35, 36, clinicians should carefully explain the extremely low ALT level, especially in older population owing to its potential implication on increasing mortality.

7.2 ULN of ALT as a hot scientific topic

Earlier ULNs for the ALT level were defined in blood donors with non-B, non-C hepatitis, and ranged from 40 to 50 U/l 102-105. These definitions did not consider that metabolic covariates caused liver injury with slight-to-moderate ALT elevation 3, 106. Because of the increasing prevalence of patients with metabolic disorders, the clinical significance of ALT values have been recognized. Metabolic covariates have been enrolled when evaluating the ULNs of ALT activity by many scholars in recent decades. Prati et al 107 first raised the concept that subjects with metabolic abnormalities should be excluded in evaluating the ULN of ALT level for the potential risk to the general health. In agreement with Prati et al 107, many scholars have re-evaluated the ULN of ALT values in specific populations, including adults and adolescents 10, 47, 108-123.

The detailed ULNs of ALT levels are distinct for the intrinsic differences in ethnicity, gender, and age distribution. In addition, the difference in definition of so-called “healthy subjects” and statistical methods might also contribute to the variations in the ULNs. Pacifico et al 15 had summarized the reported ULNs of serum ALT levels in published studies 10, 84, 107-110, 113-118, 121. This review is incomplete due to omitting some additional references 47, 111, 112, 119, 120, 122, 123 and indirect comparisons. The etiology of ALT elevation can be attributed to viral hepatitis, excessive alcohol consumption, and metabolic disorders, including fatty liver disease 4, 35, 36, 83, in population-based studies. Subjects with the above-mentioned problems should be excluded when evaluating the ULNs of ALT values in a given population. Therefore, we re-summarized the reported ULNs of ALT values by extracting the key exclusion information in various districts as a supplement and presented the information by category (Table 1, Figure 2). As shown in Figure 2, the ULNs of ALT levels in studies excluding the subjects with metabolic abnormalities were greater than those without (41.0±10.8 vs. 27.1±7.0 U/l). About one third decrease on ULN of ALT definition when excluding the subjects with metabolic disorders. The adoption of an updated ULN of ALT activity in a subsequent investigation raised the sensitivity of the diagnosis of potential liver disease with acceptable specificity decrease 47, 107, 110, 114, 117.

Table 1

Updated Upper Limit of Normal Serum Alanine Aminotransferase Value in Reported Studies.

Authors,country,year[reference]	Number of enrolled participants(M/F)	Age of enrolled participants [year, mean±SD or mean (age range)]	Exclusion of viral hepatitis (No/Yes)	Exclusion of excessive alcohol consumption (No/Yes)	Exclusion of medication (No/Yes)	Exclusion of metabolic abnormity (No/Yes)	Exclusion of fatty liver disease by imaging tools (No/Yes)	Liver biopsy (No/Yes)	Statistical methods	ULN of ALT value (U/l)
Piton et al,France, 1998, 84	487/546	30±0.36	Yes	No	No	No	No	No	95th percentile one sided	BMIa≤23: 42 for men31 for womenBMI>23: 66 for men44 for women
Prati et al, Italy, 2002, 107	3865/2970	29.8±9.5	Yes	No	Yes	Yes	No	No	95th percentile one sided	30 for men19 for women
Kariv R et al, Israel, 2006, 109	6124/11374	31.91±17.07	Yes	No	Yes	Yes	No	No	95th percentile one sided	37.5 for all44.9 for men31.8 for women
Poorten et al, Australia,2007, 110	206/0	16.8±1.4	Yes	Yes	Yes	Yes	No	No	95th percentile one sided	28
Jamali et al, Iran,2008, 111	628/1300	40.7±14.7	Yes	Yes	No	No	No	No	95th percentile one sided	BMI<25 and non-diabetics:36.1 for all37.5 for men,36 for womenBMI>25:51 for all59 for men45.25 for women
Kibaya et al, Kenya,2008, 112	1020/521	30(18-55)	No	No	No	No	No	No	97.5th percentile one sided	52 for all53.9 for men47 for women
Lee et al, Korea, 2010, 108	643/462 for pathological normality346/313 met for Prati criteria	29.1±9.0 for all27.2±8.4 for men31.6±9.3 for women	Yes	Yes	No	Yes	No	Yes	97.5th percentile one sided	Normal histological liver donor:35 for men26 for womenSubjects met the Prati criteria:33 for men25 for women
Schwimmer et al, USA, 2010, 114	548/434	14.5±1.8 for boys15.1±1.8 for girls	Yes	Yes	Yes	Yes	No	No	95th percentile one sided	25.8 for boys22.1 for girls
Kang et al, Korea,2011, 113	704/1041	41.8±12.5	Yes	Yes	Yes	Yes	No	No	95th percentile one sided	28 for all31 for men23 for women
Poustchi et al, IRAN, 2011, 10	186/185	12.87±3.13 for all12.58±3 for boys13.17±3.41 for girls	Yes	Yes	No	Yes	No	No	95th percentile one sided	30 for boys21 for girls
Volzke et al, Germany, 2011, 115	1180/1423	20-79	Yes	Yes	No	No	Yes	No	95th percentile one sided	60 for men45 for women20-49 yr: 65 for men and 42 for women50-79 yr: 42 for men and 35 for women
Park et al, Korea,2012,116	880/836	14.4 (10-19)	Yes	Yes	No	Yes	No	No	97.5th percentile one sided	33 for boys25 for girls
Ruhl et al, USA, 2012, 47	1607/2140	42.0±16.9	Yes	Yes	Yes	Yes	No	No	95th percentile one sided	24 for men18 for women
Wu et al, Taiwan, 2012, 117	1237/1657	52.4±13.1	Yes	No	No	Yes	Yes	No	95th percentile one sided	21 for men17 for women
Zheng et al, China, 2012, 118	4765/8872	35.3(19-44)	Yes	Yes	Yes	Yes	Yes	No	95th percentile one sided	35 for men23 for women
Alhamoudi et al, Saudi Arabia,2013, 119	41/93 for pathological normality52/11 for Prati criteria	28.9±7.3 for all28.5±7.3 for men30.0±7.1 for women	Yes	Yes	Yes	Yes	No	Yes	95th percentile one sided	Normal histological liver donor:42 for men31.1 for womenSubjects met the Prati criteria:33 for men22 for women
Kabir et al, Iran,2013, 120	688/621	61.5 ± 7.9 for men61.4 ± 7.9 for women	Yes	Yes	Yes	Yes	No	No	95th percentile one sided	21.4 for men18.8 for women
Park et al, Korea,2013, 121	1355/1961	36.3(≥20)	Yes	Yes	Yes	Yes	No	No	95th percentile one sided	42 for men25 for women
Sohn et al, Korea,2013, 122	297 663/113 577	24.6±6.4 for men22.2±5.2 for women	Yes	Yes	Yes	No	No	No	95th percentile one sided	34 for men24 for women15-19 yr: 27 for men, 20 for women20-24 yr: 31 for men, 21 for women25-29 yr: 35 for men, 22 for women30-34 yr: 36 for men, 23 for women35-39 yr: 36 for men, 24 for women40-44 yr: 34 for men, 28 for women45-49 yr: 37 for men, 24 for women≥50 yr:33 for men, 29 for women
Tanaka et al, Japan, 2013, 123	1462/2046	50.6±19.8 for men48.8±19.4 for women	Yes	No	Yes	Yes	Yes	No	95th percentile one sided	36 for men27 for women

Abbreviation: ALT: alanine aminotransferase; BMI: body mass index; F: female; M: male; SD: standard deviation; ULN: upper limit of normal; U/l: unit per liter; yr: year.

a: the unit of BMI is kg/m2.

Figure 2

Pooled reported ULN of ALT values by categories. a the others is represented as Australian and African. b Group A includes the studies without exclusion of subjects with metabolic abnormity when evaluating the ULN of ALT value. c Group B includes the studies with exclusion of subjects with metabolic abnormity when evaluating the ULN of ALT value. P-value is based on the comparison of ALT ULN classified by different categories (by Mann-Whitney U test).

Although controversy exists in the definition of the ALT normal range for the debate on the focus of risk-benefit or cost-effectiveness 124, the metabolic disorders should be considered when defining the normal range of the ALT level. And the ULN of ALT level should be re-defined individually with more specific ethnicity to make best use of ALT in related disease detection. In addition, these updated thresholds in the general population should be validated in the follow-up studies to make the best balance between sensitivity and specificity.

7.3 ALT-cardiovascular disease (CVD) association as a plausible issue

Some scholars have summarized the previous studies and referred to the existing controversy on the association between ALT elevation and CVD incidence 125-127. Some scholars have found a positive link between ALT elevation and CVD-related incidence in their own studies 128-130, while others did not 131-134. Wang et al 135 attributed the unobserved significant ALT-CVD association to the presence of viral hepatitis infections and alcohol abusers amongst the enrolled participants, in agreement with the opinion of Stefano et al 125. Otherwise, age, gender distribution, and ethnicity may also contribute to the heterogeneity of the ALT-CVD association in specific populations 99, 127, 136.

Inferior to gamma glutamyltransferase (GGT) 127, the ALT level is not the best indicator to screen for the incidence of CVD events in the general population. More evidence-based studies focused on ALT-CVD association are needed to disclose their inner relationship.

7.4 Key limitation of serum ALT assay in health check-ups

Although available in detecting underlying disease status, some limitation for the ALT assay should be noted. First, the histologic severity of NAFLD does not correlate with the ALT elevation. Patients with non-alcoholic steatohepatitis (NASH) were also observed in patients with normal ALT levels 63. No significant difference on histologic severity was found between NAFLD patients with or without ALT elevation after matching gender and age of respective subgroups 137. Second, a significant percentage of liver lesions were observed in patients with viral hepatitis and persistently normal ALT (PNALT). The percentage with corresponding liver fibrosis amongst patients infected with HCV/HBV with PNALT was 16% and 37%, respectively, while the percentage with corresponding cirrhosis was higher in patients infected with HBV (27%) 45,138. Third, the ALT assay is a continuous variable and even the fluctuation in the normal range also indicates the potential risk for metabolic disorders or cardiovascular disease in a given population 139, 140. Therefore, clinicians should interpret the sole ALT abnormality carefully, and multi-biomarker evaluation might enhance the diagnostic efficiency further 141.

7.5 What should clinicians do when faced with an ALT elevation

In spite of the limitations referred earlier, what appropriate measure should be taken for a clinician when faced with subjects who have ALT abnormalities? A flow diagram has been created to aid clinicians in treating adults with ALT elevations (Figure 3). When the ALT value is in normal range, annual routing test is recommended. When the ALT level is decreased to an extremely low level, especially in older adults, the accelerated process of aging and frailty, followed by reduced liver size and lowered liver blood flow, is suggested. Therefore, imaging measurements should include ultrasonography and CT scan to evaluate metabolic function of the liver. When the ALT level is elevated and exceeds the normal range, the common cause of ALT elevation, including viral hepatitis indicators, metabolic covariates, alcohol abuse, long-term medication history, and other liver functional indicators, should be investigated. Less common causes of ALT elevation, including hemochromatosis, autoimmune hepatitis, celiac disease, and muscle injury, should be identified when the inspection results referred before are negative. A liver biopsy might be the final instrument when the ALT abnormality cannot be explained. However, the clinicians should carefully evaluate and balance the patients' effect and risk they should afford when a liver biopsy is performed.

Figure 3

Strategy diagram of checking the subjects with ALT assay (on the ideal status). a the extremely low ALT value is represented as lower than the median value. b the Older subjects are indicative of the subjects aged > 70 years. c the Younger subjects are indicative of the subjects aged≤70 years. d Viral biomarkers includes the hepatitis B virus surface antigen, and the anti-HCV. e Metabolic covariates includes the BMI, triglyceride, high density lipoprotein cholesterol, glucose, and blood pressure et al expressed as a series of metabolic status.