Intrahepatic cholestasis of pregnancy and associated hepatobiliary disease: A population-based cohort study


  • Hanns-Ulrich Marschall,

    Corresponding author
    • Sahlgrenska Academy, Institute of Medicine, Department of Molecular and Clinical Medicine, University of Gothenburg, Gothenburg, Sweden
    Search for more papers by this author
  • Elisabeth Wikström Shemer,

    1. Department of Obstetrics and Gynaecology and Department of Clinical Sciences, Karolinska Institutet, Danderyd Hospital, Stockholm, Sweden
    Search for more papers by this author
  • Jonas F. Ludvigsson,

    1. Clinical Epidemiology Unit, Department of Medicine, Solna, Karolinska University Hospital and Institutet, Stockholm
    2. Department of Pediatrics, Örebro University Hospital, Örebro, Sweden
    Search for more papers by this author
  • Olof Stephansson

    1. Clinical Epidemiology Unit, Department of Medicine, Solna, Karolinska University Hospital and Institutet, Stockholm
    2. Department of Women's and Children's Health, Karolinska University Hospital and Institutet, Stockholm, Sweden
    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.

  • Supported by the Swedish Society of Medicine (to H.U.M., J.F.L., and O.S.), the Swedish Research Council – Medicine (K2005-72X-04793-30A, to H.U.M.; 522-2A09-195, to J.F.L.), the Fulbright Commission (to J.F.L.), and the regional research councils – ALF (to H.U.M. and O.S.).

Address reprint requests to: Hanns-Ulrich Marschall, M.D., Ph.D., Professor of Clinical Hepatology, Sahlgrenska Academy, Institute of Medicine, Department of Molecular and Clinical Medicine, University of Gothenburg, S-41345 Gothenburg, Sweden. E-mail:; fax: +46-31-827458.


Intrahepatic cholestasis of pregnancy (ICP) is the most common liver disease in pregnancy. We aimed to estimate the risk of developing hepatobiliary disease in women with ICP and the odds of developing ICP in women with prevalent hepatobiliary disease. We analyzed data of women with births between 1973 and 2009 and registered in the Swedish Medical Birth Register. By linkage with the Swedish Patient Register, we identified 11,388 women with ICP who were matched to 113,893 women without this diagnosis. Diagnoses of preexisting or later hepatobiliary disease were obtained from the Patient Register. Main outcome measures were hazard ratios (HRs) for later hepatobiliary disease in women with ICP and odds ratios (ORs) for developing ICP in preexisting hepatobiliary disease. Risk estimates were calculated through Cox and logistic regression analyses. Women with ICP were more often diagnosed with later hepatobiliary disease (HR 2.62; 95% confidence interval [CI] 2.47-2.77; increment at 1% per year), hepatitis C or chronic hepatitis (HR 4.16; 3.14-5.51 and 5.96; 3.43-10.33, respectively), fibrosis/cirrhosis (HR 5.11; 3.29-7.96), gallstone disease or cholangitis (HR 2.72; 2.55-2.91, and 4.22; 3.13-5.69, respectively) as compared to women without ICP (P < 0.001 for all HRs). Later ICP was more common in women with prepregnancy hepatitis C (OR 5.76; 1.30-25.44; P = 0.021), chronic hepatitis (OR 8.66; 1.05-71.48; P = 0.045), and gallstone disease (OR 3.29; 2.02-5.36; P < 0.0001). Conclusion: Women with ICP have substantially increased risk for later hepatobiliary disease. Beyond gallstone-related morbidity we found a strong positive association between ICP and hepatitis C both before and after ICP diagnosis. Thus, we advocate testing for hepatitis C in women with ICP, in particular, since this potentially life-threatening infection can be treated successfully in the majority of patients. (Hepatology 2013;58:1385–1391)


confidence interval


hazard ratio


international classification of diseases


intrahepatic cholestasis of pregnancy


odds ratio.

Intrahepatic cholestasis of pregnancy (ICP) is the most common liver disease during pregnancy,[1] with reported prevalence rates of between 0.4% and 1.5%.[2, 3] ICP is characterized by otherwise unexplained pruritus with elevated bile acids and/or transaminases in the late second and third trimester of pregnancy.[2, 3] Genetic predisposition (variants of hepatobiliary transport proteins), impaired turnover of reproductive hormones, and environmental factors play key roles in the etiology and pathogenesis of ICP.[2, 3]

By definition, ICP is a diagnosis of exclusion. Pruritus in ICP spontaneously resolves and deranged liver function tests typically normalize within 4 weeks after delivery. Thus, ICP is usually a self-limiting, benign condition for the mother, although the fetus is at increased risk of adverse outcome, in particular preterm delivery.[4, 5] In a small number of women with ICP, liver function does not return to normal, suggesting an underlying hepatobiliary disease that needs further diagnostic testing.[6] These women might have been misclassified as suffering from ICP only.

It is well established that ICP is associated with gallstone disease,[3, 4] in part explained by variants in the ABCB4 gene encoding the canalicular phospholipid floppase associated with both cholesterol gallstone disease and ICP.[7-9] ICP has also been linked to hepatitis C infection.[10, 11] A large population-based study from Finland comparing women with and without an ICP diagnosis in 1972-2000 found a significant risk of later hepatobiliary disease for women diagnosed with ICP, in particular gallstone-related morbidity and hepatitis C infection.[12]

The current population-based study examined the association between ICP and hepatobiliary disease both before and after ICP diagnosis using a detailed classification of hepatitis. For this purpose we linked data from the Swedish Medical Birth Register on women with first and consecutive births between 1973 and 2009 with data from the Swedish Patient Register. We identified 11,388 women with ICP and matched them to 113,893 women without this diagnosis.

Patients and Methods

All women with their first and consecutive births recorded in the Swedish Medical Birth Register between 1973 and 2009 were selected for this study. The Birth Register includes information on about 98% of births in Sweden, with information on the mother as well as the pregnancy, delivery, and neonatal period.[13] By linkage between the Birth and Patient registers, we identified a study population of women with ICP and controls without ICP. This study was approved by the local Ethics Review Board.

Data on ICP were collected prospectively starting at the woman's first antenatal visit. Women with ICP were identified according to the relevant International Classification of Diseases (ICD). Three ICD versions were used to indicate ICP in Sweden between 1973 and 2009, as follows: ICD-8 (1968 to 1986): 639.00, pruritus; 639.01, icterus gravis; 639.09, necrosis acuta et subacuta hepatis; ICD-9 (1987 to 1996): 6467A, hepatosis gravidarum; 6467X, hepatopathia alia gravidarum; ICD-10 (1997 to 2010): O26.6, intrahepatic cholestasis of pregnancy.

Starting with first-time births recorded in the Birth Register between 1973 and 2009, we matched each woman with ICP to 10 women without ICP based on maternal age and calendar year of delivery. After all women with ICP in the first parity had been matched with 10 control women, the matching procedure continued with the second birth, and so on. If a control woman was matched to a woman with ICP, she could not be matched to another woman with ICP at a subsequent parity. Thus, each woman could only appear once in the ICP group and once in the control group. The matching procedure was performed using the “Match” function within the package “Matching” in R.[14] Detailed demographics of ICP and control groups are shown in Table 1.

Table 1. Demographics
  Control (N = 113,893)ICP (N = 11,388)All (N = 125,281)
Age of motherMean (SD)29.01 (5.34)29.01 (5.34)29.01 (5.34)
Range15 to 4515 to 4515 to 45
Sex of childBoy n (%)58,702 (51.54)6,199 (54.44)64,901 (51.81)
Girl n (%)55,187 (48.46)5,187 (45.56)60,374 (48.19)
Birth weight gMean (SD)3,504.13 (576.23)3 436.87 (594.71)3,498.03 (578.27)
Range307 to 6,454435 to 5,670307 to 6,454
Parity n (%)165,201 (57.25)6,519 (57.24)71,720 (57.25)
231,469 (27.63)3,146 (27.63)34,615 (27.63)
312,774 (11.22)1,278 (11.20)14,058 (11.22)
43,158 (2.77)316 (2.77)3,474 (2.77)
5943 (0.83)92 (0.81)1,035 (0.83)
6239 (0.21)27 (0.24)266 (0.21)
762 (0.05)3 (0.03)65 (0.05)
828 (0.02)6 (0.05)34 (0.03)
915 (0.01)2 (0.02)17 (0.01)
103 (0.00)0 (0.00)3 (0.00)
121 (0.00)0 (0.00)1 (0.00)
140 (0.00)1 (0.01)1 (0.00)
Twins n (%)No112,723 (98.97)11,029 (96.85)123,752 (98.78)
Yes1,170 (1.03)359 (3.15)1,529 (1.22)
Prior liver diagnosisNo107,185 (94.11)10,067 (88.40)117,252 (93.59)
Yes6,708 (5.89)1,321 (11.60)8,029 (6.41)

The ICD codes 7-10 used for the identification of hepatobiliary disease in the patient register between 1964 and 2010 are shown in Table 2. In addition, we also identified individuals with symptoms of hepatobiliary disease as well as with relevant operation codes indicating symptomatic gallstone disease.[15] Of note, ICD-7 to 9 did not contain specific diagnostic codes for viral hepatitis A, B, and C but instead listed “any infectious hepatitis.” The specified ICD-10 subcode K83.0A for primary sclerosing cholangitis is as yet not established and K83 is still commonly used.

Table 2. Hepatobiliary Disease as Defined in International Classification of Diseases (ICD) Lists 7-10 Between 1964 and 2010
Disease GroupICD-7 (1964-1967)ICD-8 (1968-1986)ICD-9 (1987-1996)ICD-10 (1997-2010)
Infectious hepatitis09207070B15-19
Hepatitis A   B15
Hepatitis B   B16 B18.0 B18.1
Hepatitis C   B17.1 B18.2
Bilirubin disorders 273277EE80
Wilson 273.30275BE83.0
Portal vein thrombosis583.00452.99452I81
Budd-Chiari  453AI82
Liver alcoholic581.10 583.10571.00 571.01571A-DK70
Liver toxic  573DK71
Liver failure580.19570.02 573.02570 572CK70.4 K71.1 K72
Liver chronic hepatitis583.00570.01 573.00571EK73
Liver fibrosis / cirrhosis581.00571.90 573.01 573.02 573.03571 571FK74
Liver other inflammatory582.00 582.10 583.29572.99572K75.4 K75.5 K75.9
Liver other583.98 583.9957308 573.09573K76
Liver not classified 571.98 571.99571W 571XK77
Primary biliary cirrhosis  571GK74.3
Primary sclerosing cholangitis   K83.0A
Gallstone disease584574574K80
Gallbladder other586576576K82
Gallbladder biliary pancreas other   K87
Postcholecystectomy syndrome   K91.5
Icterus 785782ER17
Varices 456.00456A-CI85
Operation codes indicating gallstone disease5350-53595350-53595350-5359JKA20 JKA21 JKB11 JKB30
 5394 5388 90145388 5394 90145388 5394 9014JKE00 JKE02 JKE06 JKE12
 534153415341JKE15 JKE18 JKF10

The primary objective was to estimate the relative risk (hazard ratio, HR) of developing hepatobiliary disease after ICP. For the primary analysis all women with a hepatobiliary disease diagnosis prior to the birth of interest were excluded from the time to event analyses. A secondary objective was to estimate the odds ratio (OR) for ICP in women with a previous diagnosis of hepatobiliary disease.

Time to first hepatobiliary disease diagnosis was analyzed using a stratified Cox proportional hazards model where each woman with ICP was compared separately with her matched controls before a summary estimate was calculated. The observation time was censored at the date of migration, date of death, or end of study (December 31, 2010), unless a positive event had occurred. For the control group the time was also censored at an ICP delivery. The proportionality assumption was tested using scaled Schoenfeld residuals with time since ICP diagnosis (or corresponding birth in matched controls) as the principal time scale. In the initial analyses we stratified for maternal age, cigarette smoking, and parity, and in a separate analysis we also included adjustment for maternal country of birth. The case-control analysis was performed using a mixed effects logistic regression with the matched pairs and individuals as random factors. In this article we report unadjusted estimates with 95% confidence intervals (CI).


The study cohort consisted of 125,281 women of whom 11,388 were diagnosed with ICP. ICP women and their matched controls had a similar age distribution and their offspring did not differ in birth weight (Table 1). A larger proportion of boys were born and there were more twin pregnancies in the ICP group (Table 1). The higher twin rate was expected, since multiple pregnancy is an established risk factor for ICP.[3]

ICP and Risk of Later Hepatobiliary Disease

The HR for later hepatobiliary disease after first ICP diagnosis was 2.62 (95% confidence interval [CI] 2.47 to 2.77) with a cumulative annual increase of ∼1% (Fig. 1). The highest risk estimates (P < 0.0001 for all) were found for chronic hepatitis (HR 5.96, 95% CI 3.43 to 10.33), liver fibrosis/cirrhosis (HR 5.11, 95% CI 3.29 to 10.33), hepatitis C (HR 4.16, 95% CI 3.14 to 5.51), and cholangitis (HR 4.22, 95% CI 3.13 to 5.69) (Table 3). Although gallstone disease was the most frequently recorded hepatobiliary disease in the ICP group (n = 1169, 11.6%), it was only associated with a moderately increased risk (HR 2.72, 95% CI 2.55 to 2.91). Hepatitis A, B, and C were diagnosed in a total of 84 women with ICP, and the remaining 46 out of 130 “any infectious hepatitis” cases were considered as non-A non-B hepatitis. HRs for alcoholic and toxic liver disease were 1.99 (95% CI 1.02 to 3.88) and 2.17 (95% CI 1.08 to 4.33), respectively. Adjusting for maternal country of birth did not influence the risk estimates (e.g., HR for later hepatobiliary disease, 2.62, 95% CI 2.49 to 2.80; data not shown).

Figure 1.

Time after ICP to first hepatobiliary disease. For this analysis all women with a prior liver disease were excluded as well as all matched controls for ICP women with a prior liver disease.

Table 3. Risk of Developing Hepatobiliary Disease in Women With ICP
Disease GroupControl (N = 94,863)ICP (N = 10,067)Hazard Ratio95% Confidence IntervalP ValueHR <1yrP ValueHR 1-5 yrP ValueHR >5yrP Value
  1. Results from the Cox regression analysis of the complete ICD dataset from 1973 on.

  2. a

    Subanalysis for hepatitis A, B, C and primary sclerosing cholangitis only for women with the specific birth after 1996 (controls, N = 66,651; ICP, N = 6,667).

Any5,937 (6.3%)1,514 (15%)2.62(2.47-2.77)<0.00014.53<0.00012.67<0.00012.26<0.0001
Infectious hepatitis311 (0.3%)130 (1.3%)3.91(3.18-4.81)<0.00013.410.00024.30<0.00014.01<0.0001
Hepatitis Aa7 (0%)1 (0%)1.31(0.16-10.67)0.79971.580.6701
Hepatitis Ba78 (0.1%)11 (0.1%)1.34(0.71-2.52)0.36992.360.18422.240.10500.630.4413
Hepatitis Ca161 (0.2%)72 (0.7%)4.16(3.14-5.51)<0.00015.660.00084.710.00014.10<0.0001
Bilirubin disorders32 (0%)9 (0.1%)2.80(1.33-5.89)0.006937.720.00122.900.06310.530.5364
Hemochromatosis29 (0%)0 (0%)
Wilson2 (0%)0 (0%)
Portal vein thrombosis14 (0%)0 (0%)
Budd-Chiari177 (0.2%)17 (0.2%)0.93(0.56-1.53)0.76585.660.01770.430.40630.830.5243
Liver alcoholic53 (0.1%)11 (0.1%)1.99(1.02-3.88)0.04232.030.0326
Liver toxic42 (0%)10 (0.1%)2.17(1.08-4.33)0.028714.140.00371.880.56431.640.2646
Liver failure41 (0%)13 (0.1%)3.29(1.74-6.22)0.00039.400.00612.020.0712
Liver chronic hepatitis35 (0%)21 (0.2%)5.96(3.43-10.33)<0.00012.690.21724.78<0.0001
Liver fibrosis / cirrhosis61 (0.1%)31 (0.3%)5.11(3.29-7.96)<0.000147.150.00044.710.07374.12<0.0001
Liver other inflammatory23 (0%)10 (0.1%)4.30(2.03-9.15)0.000118.850.01653.130.16193.590.0076
Liver other182 (0.2%)98 (1%)5.26(4.10-6.76)<0.000128.32<0.00018.50<0.00012.79<0.0001
Liver not classified12 (0%)7 (0.1%)5.26(2.07-13.36)0.00054.710.20602.860.1110
Primary biliary cirrhosis5 (0%)3 (0%)5.54(1.32-23.20)0.01919.410.11294.720.0730
Cholangitis146 (0.2%)63 (0.6%)4.22(3.13-5.69)<0.000122.00<0.00012.880.00224.04<0.0001
Primary sclerosing cholangitisa4 (0%)3 (0%)6.58(1.47-29.49)0.01384.710.20604.790.2010
Gallstone disease4,320 (4.6%)1,169 (11.6%)2.72(2.55-2.91)<0.00014.11<0.00012.96<0.00012.39<0.0001
Cholecystitis281 (0.3%)64 (0.6%)2.18(1.66-2.86)<0.00013.000.01142.010.01132.13<0.0001
Gallbladder other140 (0.1%)52 (0.5%)3.55(2.57-4.89)<0.00017.91<0.00013.66<0.00012.83<0.0001
Gallbladder biliary pancreas other11 (0%)2 (0%)1.62(0.36-7.32)0.53083.140.3218
Postcholecystectomy syndrome5 (0%)2 (0%)4.83(0.88-26.36)0.06924.750.0719
Ikterus852 (0.9%)147 (1.5%)1.68(1.41-2.00)<0.00013.58<0.00011.280.11521.520.0014
Hepatosplenomegaly15 (0%)4 (0%)2.40(0.79-7.34)0.1236 3.140.16162.140.3310
Varices31 (0%)14 (0.1%)4.82(2.51-9.23)<0.00015.00<0.0001
Ascites95 (0.1%)21 (0.2%)2.13(1.32-3.43)0.0019 2.040.0049

In a sensitivity analysis we restricted the analysis of births from 1997 and onwards using ICD-10 for ICP diagnosis. In the restricted analysis, the HR for any liver disease was 2.97 (95% CI 2.66-3.22) (Supporting Table 1). HRs were generally somewhat higher as compared to the HRs of the original analysis (Table 3). The risk of change findings is higher in the restricted analyses, since these were based on fewer observations than the original analyses. On the contrary, for diseases with longer onset the restricted analyses will not be able to assess risk due to the shorter time of follow-up.

Risk of later hepatobiliary disease was not related to smoking status in early pregnancy, age at first ICP diagnosis, or number of earlier pregnancies (Supporting Table 2).

Hepatobiliary Disease and Risk of Later ICP

Preexisting hepatobiliary disease was associated with a 2-fold increased risk of later ICP (OR 1.98, 95% CI 1.45 to 2.70). Specifically, there was a positive association between gallstone disease (OR 3.29, 95% CI 2.02 to 5.36), hepatitis C (OR 5.76, 95% CI 1.30 to 25.44), and chronic hepatitis (OR 8.66, 95% CI 1.05 to 71.48) and later ICP (Table 4).

Table 4. Risk of Developing ICP After a Hepatobiliary Disease
Disease GroupControl (N = 113,893)ICP (N = 11,388)Odds Ratio95% Confidence IntervalP Value
  1. Results from the mixed logistic regression analysis.

  2. a

    Subanalysis for hepatitis A, B, C and primary sclerosing cholangitis only for women with the specific birth after 1996 (controls, N = 66,651; ICP, N = 6,667).

Infectious hepatitis468 (0.4%)145 (1.3%)3.22(1.30-7.97)0.0115
Hepatitis Aa2 (0%)2 (0%)10.02(0.00-43,953.34)0.5901
Hepatitis Ba149 (0.2%)20 (0.3%)0.87(0.06-13.55)0.9226
Hepatitis Ca121 (0.2%)66 (1%)5.76(1.30-25.44)0.0210
Bilirubin disorders44 (0%)12 (0.1%)2.73(0.12-64.80)0.5334
Hemochromatosis3 (0%)0 (0%)
Portal vein thrombosis3 (0%)0 (0%)
Budd-Chiari37 (0%)7 (0.1%)1.89(0.03-112.20)0.7595
Liver alcoholic9 (0%)3 (0%)3.34(0.01-1,993.94)0.7118
Liver toxic15 (0%)17 (0.1%)11.47(0.80-164.61)0.0726
Liver failure5 (0%)5 (0%)10.04(0.05-2,018.76)0.3940
Liver chronic hepatitis33 (0%)28 (0.2%)8.66(1.05-71.48)0.0451
Liver fibrosis cirrhosis5 (0%)6 (0.1%)12.06(0.20-712.56)0.2315
Liver other inflammatory8 (0%)14 (0.1%)17.68(1.02-306.34)0.0484
Liver other65 (0.1%)78 (0.7%)12.76(3.57-45.58)0.0001
Liver not classified5 (0%)10 (0.1%)
Primary biliary cirrhosis0 (0%)1 (0%)
Cholangitis37 (0%)19 (0.2%)5.19(0.38-69.99)0.2150
Primary sclerosing cholangitisa3 (0%)3 (0%)10.04(0.01-9,445.52)0.5092
Gallstone disease1,486 (1.3%)441 (3.9%)3.29(2.02-5.36)0.0000
Cholecystitis152 (0.1%)47 (0.4%)3.14(0.69-14.22)0.1384
Gallbladder other95 (0.1%)43 (0.4%)4.61(0.89-24.01)0.0694
Gallbladder biliary pancreas other3 (0%)0 (0%)
Postcholecystectomy syndrome1 (0%)0 (0%)
Ikterus4,643 (4.1%)643 (5.6%)1.18(0.77-1.80)0.4498
Hepatosplenomegaly9 (0%)0 (0%)
Varices2 (0%)1 (0%)5.003(0.00-413,376.67)0.7805
Ascites17 (0%)5 (0%)2.947(0.06-139.18)0.5827


In the present population-based study an ICP diagnosis was associated with an increased risk of later hepatobiliary disease with an increment of ∼1% per year. Similarly, preexisting liver disease increased the odds of subsequently being diagnosed with ICP. The major finding of our study is the strong association between ICP and hepatitis C infection, not only after but also before ICP diagnosis.

Our observation of future liver disease in women once given the diagnosis of ICP is in agreement with earlier data.[10-12] Our study adds precision given that it included 10 times as many controls without ICP as the population-based study from Finland that compared 10,504 women with and without an ICP diagnosis in 1972-2000.12 In particular, we provide more precise estimates of hepatitis C infection rates since a dedicated diagnosis code for hepatitis C was not available before 1996.12 The Finnish study was also restricted to inpatient care data,[12] which might bias risk estimates for liver disease since a large share of hepatobiliary disease is diagnosed and managed in an outpatient setting. In addition, our study is the first to present ORs for later ICP in women with prior hepatobiliary disease.

The main strengths of our study are the large number of participants, the long follow-up, and our use of nationwide register data to ascertain both ICP and hepatobiliary disease. In Sweden, public health care is almost free of charge and almost 100% of pregnant women attend antenatal health care check-ups.[4] Every Swedish resident is assigned a unique personal identity number,[16] allowing for extensive registry linkage also with emigration and cause of death data, thereby assuring a complete and correct follow-up of study participants.

We did not have access to data on liver biopsy or serology and were therefore unable to validate the hepatobiliary disease diagnoses. Earlier research, however, has shown a positive predictive value of about 85%-95% in chronic diseases recorded in the Swedish patient register.[17] Although the medical birth and patient registers are nationwide and are thus likely to have a high sensitivity for specific diagnoses such as ICP, we cannot rule out that a small number of women with ICP were not diagnosed with ICP. Given that ICP in Sweden has an overall prevalence of 1.5%,[5] the existence of a small number of false-negative ICP women among our controls, however, is unlikely to have affected our risk estimates more than marginally and if so should have pushed the risk estimates towards 1.

The positive association between ICP and other hepatobiliary disease was independent of temporality. Some part of the association is likely due to shared risk factors such as variants in the ABCB4 gene that are associated both with ICP and gallstone disease[7-9] and probably also drug-induced cholestasis,[18] as reflected by an increased risk of developing toxic liver disease. Through matching we were able to minimize the influence of some potential confounders (e.g., shared genetic background and environment), but we lacked data on levels of trace elements such as selenium[19] or vitamin D,[20] which have been linked to both ICP and hepatobiliary disease.

The high prevalence of hepatitis C infection in women with ICP may be due to an enhanced susceptibility to hepatitis C infection in ICP and vice versa. Prevalent hepatitis C infection may also worsen cholestatic pruritus under high estrogen and progesterone loads of late pregnancy[21] when potentially itch-related enzymes such as autotaxin are also markedly elevated.[22, 23] Hepatitis C has been shown to decrease the expression of MRP2, which is a putative transport protein for sulfated pregnancy hormones in human liver.[24] Also, bile acids have been found to affect the response to hepatitis C treatment in patients who carry a common polymorphism of the gene ABCB11 encoding the bile salt export pump.[25] This ABCB11 polymorphism has also been associated with ICP.[26]

A more simple explanation for the high prevalence of hepatitis C infection in women with ICP may be more plausible. Once a woman presents with pruritus in late pregnancy, elevated serum bile acids and perhaps also transaminases are considered to be sufficient for the diagnosis of ICP, although it is a diagnosis that should only be given after exclusion of preexisting liver disease. Also, current EASL (European Association for the Study of the Liver) guidelines are somewhat indistinct in this respect by recommending hepatitis C testing only after delivery in women with persistently abnormal liver function tests.[6] Interestingly, a recent Dutch study evaluating pregnancy outcomes associated with viral hepatitis did not find an association between hepatitis C and cholestasis, possibly due to underreporting,[27] but an increased risk for gestational diabetes[27] that we recently found to be strongly associated with ICP.[5]

The substantially increased risk of developing liver fibrosis and cirrhosis in women with ICP is also likely to be caused by hepatitis C infection. Of note, HRs for all of these diagnoses were higher during the first year after a diagnosis of ICP than at a later timepoint, which might indicate clinical reevaluation of ICP after delivery. Compared with controls, women with ICP were less likely to have a diagnosis of alcoholic cirrhosis. One may debate the validity of this diagnosis since it is largely dependent on the patient's self-report data. However, in Sweden women will receive a diagnosis of substantial abuse of alcohol at consumption levels of more than 20 g per day. Also, fatty liver disease may not be an important contributor to fibrosis and cirrhosis in ICP since our previous research has shown a lower mean body weight in women with ICP than in control women.[5]

In conclusion, women with ICP have a substantially increased risk of later hepatobiliary disease. The prevalence and later diagnosis of chronic hepatitis C and its complications are of particular importance since this potentially life-threatening infection can now be treated successfully in the majority of patients. Thus, we strongly advocate testing for hepatitis C in women with signs of ICP.


We thank Rosie Perkins for editing the article.