The Effect of Overweight and Obesity on Liver Biochemical Markers in Children and Adolescents

Magnus J. Johansen; Julie Gade; Stefan Stender; Christine Frithioff-Bøjsøe; Morten A. V. Lund; Elizaveta Chabanova; Henrik S. Thomsen; Oluf Pedersen; Cilius E. Fonvig; Torben Hansen; Jens-Christian Holm

Disclosures

J Clin Endocrinol Metab. 2020;105(2) 

In This Article

Results

Study Population

From the population-based cohort, 1858 (1096 girls) normal weight children and adolescents with a median age of 11.6 years and 10.6 years for girls and boys respectively, were included in the study. From the cohort with overweight or obesity, 2155 (1158 girls) children and adolescents with a median age of 11.6 years and 11.8 years for girls and boys, respectively, were included in the study. The number of girls and boys in each 1-year age stratum in each cohort are shown in Supplementary Table S1[33] (All supplementary material and figures are located in a digital research materials repository[33]), and the distribution of BMI-SDS stratified by age, is shown in Supplemental Figure 1.[33] Age- and sexspecific percentiles for ALT, AST, LDH, GGT, bilirubin, and ALP for the population-based cohort are shown in Supplemental Tables S2 and S3.[33] A small number of participants had missing values for each biochemical parameter (range < 3%-17%). Children and adolescents with overweight or obesity (n = 463) or underweight (n = 134) in the population-based cohort were excluded from the main analysis, to ensure that the reference values were based solely on healthy normal-weight participants. However, including these participants did not substantially change any of the reference values reported in Supplemental Tables S2 and S3[33] (data not shown).

Anthropometric Parameters and Percentile Curves of Liver Biochemical Markers

Median and interquartile ranges for the 6 biochemical markers, stratified by age group, sex, and cohort, as well as anthropometric parameters of the cohort are shown in Table 1.

Population-based Cohort. The percentile curves for ALT, AST, LDH, GGT, bilirubin, and ALP are shown stratified by age and sex for the population-based cohort in the left panels (A) of Figures 1–4. The ALT percentile curves showed a small decrease with age in girls, but not in boys. The median ALT concentration in girls aged 6–9, 10–13 and 14–18 years were 21 U/L, 18 U/L, and 18 U/L, respectively, and 21 U/L, 19 U/L, and 20 U/L in boys, respectively. Increasing BMI-SDS was associated with increasing ALT in boys (P < 0.001) but not in girls (P = 0.885) from the population based cohort. The AST percentile curves showed significant decreases with age in both girls and boys. Median AST concentration in girls aged 6–9, 10–13, and 14–18 years were 31 U/L, 25 U/L, and 20 U/L, respectively, and 32 U/L, 27 U/L, and 25 U/L in boys, respectively. The percentile curves for LDH concentrations in girls and boys showed a decrease with higher age. The median LDH concentration in girls aged 6–9, 10–13, and 14–18 years were 228 U/L, 201 U/L, and 161 U/L, respectively. The corresponding values in boys were 233 U/L, 212 U/L, and 184 U/L, respectively. The percentile curves for GGT concentrations showed no changes with age in neither girls nor boys. Median concentrations in girls and boys aged 6–9, 10–13, and 14–18 years were 17 U/L, 16 U/L, and 16 U/L, and 17 U/L, 17 U/L, and 18 U/L, respectively. The bilirubin percentile curves showed no change with age in girls, but a modest increase with age in boys. Median concentrations in girls aged 6–9, 10–13, and 14–18 years were 7 μmol/L, 7 μmol/L, and 8 μmol/L. The corresponding values in boys were 6 μmol/L, 7 μmol/L, and 9 μmol/L. The percentile curves for ALP showed a complex pattern in the population cohort, with a modest increase until reaching a peak at ages 11–12 years in girls and 13–14 years in boys, followed by a sharp decline in both sexes. Median concentrations in girls and boys aged 6–9, 10–13 and, 14–18 years were 262 U/L, 275 U/L, and 101 U/L and 247 U/L, 271 U/L, and 233 U/L, respectively.

Figure 1.

Percentile curves for alanine aminotransferase (ALT), aspartate aminotransferase (AST), and lactate dehydrogenase (LDH), in girls from the population-based cohort (left panels) and the cohort with overweight or obesity (right panels). Upper and lower black lines represent the 2.5th and 97.5th percentile, dark grey lines are medians, and grey lines are the 5th, 25th, 75th, and 95th percentiles.

Figure 2.

Percentile curves for gamma-glutamyltransferase (GGT), bilirubin, and alkaline phosphatase (ALP) in girls from the population-based cohort (left panels) and and the cohort with overweight or obesity (right panels). Upper and lower black lines represent the 2.5th and 97.5th percentile, dark grey lines are medians, and grey lines are the 5th, 25th, 75th, and 95th percentiles.

Figure 3.

Percentile curves for alanine aminotransferase (ALT), aspartate aminotransferase (AST), and lactate dehydrogenase (LDH) in boys from the population-based cohort (left panels) and the cohort with overweight or obesity (right panels). Upper and lower black lines represent the 2.5th and 97.5th percentile, dark grey lines are medians, and grey lines are the 5th, 25th, 75th, and 95th percentiles.

Figure 4.

Percentile curves for gamma-glutamyltransferase (GGT), bilirubin, and alkaline phosphatase (ALP) in boys from the population-based cohort (left panels) and the cohort with overweight or obesity (right panels). Upper and lower black lines represent the 2.5th and 97.5th percentile, dark grey lines are medians, and grey lines are the 5th, 25th, 75th, and 95th percentiles.

We calculated standard deviation scores for ALT and AST for each child from the population-based cohort. There were 27 girls and 16 boys with ALT-SDS above the 97.5th percentile, and 26 girls and 14 boys with AST-SDS above the 97.5th percentile. Of these, 8 girls and 6 boys had both ALT-SDS and AST-SDS over the 97.5th percentile (Supplemental Table S4).[33] The 27 girls with high ALT alone and the 26 girls with high AST alone were similar to the rest of the girls in the population-based cohort, whereas the girls with both high ALT and AST had slightly lower BMI-SDS. Compared with the rest of the boys, the 16 boys with high ALT and the 14 boys with high AST, as well as the 6 boys with high ALT and AST were older, taller, had larger waist and hip circumferences and had a higher BMI-SDS.

Cohort With Overweight or Obesity. The percentile curves for ALT, AST, LDH, GGT, bilirubin, and ALP are shown stratified by age and sex for the cohort with overweight or obesity in the right panels (B) of Figures 1–4.

Compared with the population-based cohort, participants in the cohort with overweight and obesity exhibited higher concentrations of ALT in all age groups and in both sexes (6–9, 10–13, and 14–18 years of age; P < 0.001 for all comparisons, Table 1). In contrast to the pattern observed in the population-based cohort, there was a marked increase in ALT with increasing age among children and adolescents with overweight or obesity, which was most pronounced in boys. At 6–9, 10–13, and 14–18 years of age, boys with overweight or obesity exhibited median ALT concentrations of 22 U/L, 23 U/L, and 29 U/L, respectively. We examined whether differences in abdominal adiposity might play a role in the seemingly stronger effect seen in boys. There was a strong direct correlation between age and waist/hip ratio (a proxy of abdominal obesity) in boys (P < 0.003), but an inverse association in girls (P < 0.001), regardless of adjustment for BMI-SDS. Increasing BMI-SDS was strongly associated with increasing ALT in both girls and boys with overweight or obesity (P < 0.001, both). The percentile curve patterns for AST in the population-based cohort and the cohort with overweight or obesity were overall comparable. However, in girls and boys younger than 10 years of age, there was a modest decrease in AST in the cohort with overweight or obesity compared with the population-based cohort (P < 0.001). In contrast, girls with overweight or obesity older than 13 years of age exhibited a small increase in AST compared with girls of similar age from the population-based cohort (P = 0.02). Concentrations of LDH were overall higher in the cohort with overweight and obesity than in the populationbased cohort, for all ages, and in both sexes (P < 0.001). The LDH percentile curves showed a robust decrease with age in the cohort with overweight and obesity, comparable with what was observed in the population-based cohort. Compared with the population-based cohort, there was a modest age-associated increase in GGT in girls and boys with overweight or obesity, which was most pronounced in boys. Girls and boys with overweight or obesity who were older than 13 years had significantly higher median concentrations of GGT (17 U/L and 20 U/L, respectively) compared with participants of the same age in the population-based cohort. The percentile curves for bilirubin and ALP were not significantly different in the population-based and the cohort with overweight or obesity, except for ALP in the youngest age group for both girls and boys (P < 0.001)

Effects of Puberty

Data on puberty stage were available in 78.5% and 58.0% of girls and boys, respectively, in the population-based cohort, and in 85.5% and 74.7% of girls and boys in the cohort with overweight or obesity. Compared with prepuberty, puberty was associated with 2 U/L lower ALT (P = 0.005), 2 U/L lower AST (P = 0.003), 1.6 U/L lower GGT(P = 0.001), and 96 U/L higher ALP (P < 0.001) in girls from the population-based cohort. Only ALP was influenced by puberty in boys from the population-based cohort (28 U/L higher ALP, P = 0.017). In the cohort with overweight or obesity, puberty was associated with 4 U/L lower ALT (P = 0.024), 10 U/L lower LDH (P = 0.007), 3 U/L lower GGT (P = 0.001), and 28 U/L higher ALP (P < 0.001) in girls. In boys with overweight and obesity, puberty was associated with 4 U/L lower ALT (P = 0.036), 3 U/L lower GGT (P = 0.008), and 30 U/L higher ALP (P < 0.001). All associations were adjusted for age and BMI-SDS.

Hepatic Steatosis Measured by 1H-MRS

Combined from both cohorts, hepatic 1H-MRS data within 30 days of blood sampling were available for 458 children and adolescents (248 girls). We defined hepatic steatosis as a liver fat content of > 5%.[34,35] Using this definition, 25 girls (9.9%) and 44 boys (21%) exhibited hepatic steatosis. Children and adolescents with hepatic steatosis had a higher degree of adiposity than those without steatosis (Supplemental Table S5).[33] We used ROC-curves and AUC to assess the ability of elevated plasma ALT to identify children and adolescents with hepatic steatosis (Figure 5). For both girls and boys combined, the optimal ALT cutoff was 31.5 U/L, which yielded an AUC of 76.6%, a sensitivity of 80.7% and a specificity of 65.2%, for correctly identifying those with hepatic steatosis. The corresponding cutoff in girls was 24.5 U/L (sensitivity, 55.6%; specificity, 84.0%; AUC, 71.8 %), and 34.5 U/L in boys (sensitivity, 83.7%; specificity, 68.2%; AUC, 79.1 %).

Figure 5.

Receiver operating characteristics (ROC) curves for alanine aminotransferase (ALT) as a diagnostic tool for hepatic steatosis of >5% in children and adolescents. (A) Boys and girls combined. The area under the curve (AUC) was 76.6 %, and the optimal cutoff of ALT 31.5 U/L identified hepatic steatosis with a sensitivity of 80.7 % and a specificity of 65.2 %. (B) Boys alone. The AUC was 79.1 %, the optimal cutoff of ALT of 34.5 U/L identified hepatic steatosis with a sensitivity of 83.7 % and a specificity of 68.2 %. (C) Girls alone. The AUC was 71.8 %, and an optimal cutoff of ALT of 24.5 U/L identified hepatic steatosis with a sensitivity of 55.6 % and a specificity of 84.0 %.

We also assessed the ability of ALT to identify hepatic steatosis as defined by liver fat content above 1.5% (Supplemental Figure S2),[33] a cutoff we recently found to more accurately identify steatosis in children.[31] Using the 1.5% definition, 68 girls (27.4%) and 83 boys (39.5%) exhibited hepatic steatosis. Here we found the optimal ALT cutoff to be 25.5 U/L with an AUC of 72.2%, a sensitivity of 64.2%, and a specificity of 70.9% for both boys and girls combined. The corresponding cutoff in girls was 24.5 U/L (sensitivity, 61.7%; specificity, 75.0%; AUC, 70.3 %), and 31.5 U/L in boys (sensitivity, 83.5%; specificity, 54.2%; AUC, 73.0 %).

processing....