Liver Failure and Liver Disease
Incidence, natural history, and risk factors for biliary sludge and stones during pregnancy†
Article first published online: 19 JAN 2005
Copyright © 2005 American Association for the Study of Liver Diseases
Volume 41, Issue 2, pages 359–365, February 2005
How to Cite
Ko, C. W., Beresford, S. A. A., Schulte, S. J., Matsumoto, A. M. and Lee, S. P. (2005), Incidence, natural history, and risk factors for biliary sludge and stones during pregnancy. Hepatology, 41: 359–365. doi: 10.1002/hep.20534
Conflict of interest: Nothing to report.
- Issue published online: 19 JAN 2005
- Article first published online: 19 JAN 2005
- Manuscript Accepted: 23 OCT 2004
- Manuscript Received: 21 JUL 2004
- National Institutes of Health. Grant Number: DK46890
Gallstones are strongly associated with higher parity in women. This study prospectively assessed the incidence, natural history, and risk factors for biliary sludge and stones during pregnancy and the postpartum in 3,254 women at an army medical center. Women with a prior cholecystectomy or with stones at their first study ultrasound were excluded. Gallbladder ultrasound and subject questionnaires were obtained in each trimester and at 4 to 6 weeks postpartum. Serum glucose, lipids, insulin, leptin, estradiol, and progesterone were measured at 26 to 28 weeks' gestation. A nested case-control study was done to examine the effects of serum leptin and insulin on incident gallbladder disease. At least two study ultrasounds were available for 3,254 women. Sludge or stones had been found on at least one study ultrasound in 5.1% by the second trimester, 7.9% by the third trimester, and 10.2% by 4 to 6 weeks postpartum. Regression of sludge and stones was common, such that overall 4.2% had new sludge or stones on the postpartum ultrasound. Twenty-eight women (0.8%) underwent cholecystectomy within the first year postpartum. Prepregnancy body mass index was a strong predictor of incident gallbladder disease (P < .001). Serum leptin was independently associated with gallbladder disease (odds ratio per 1 ng/dL increase, 1.05; 95% CI, 1.01, 1.11), even after adjusting for body mass index. In conclusion, incident gallbladder sludge and stones are common in pregnancy and the postpartum, and cholecystectomy is frequently done within the first year postpartum. Prepregnancy obesity and serum leptin are strong risk factors for pregnancy-associated gallbladder disease. (HEPATOLOGY 2005;41:395–365.)
In the United States, gallstone disease is the most common and costly of all digestive diseases, requiring more than 700,000 cholecystectomies annually.1–4 Both the frequency and number of pregnancies are major risk factors for cholesterol gallstones.2, 5–9 European studies have suggested that new biliary sludge (a precursor to gallstones) and gallstones may form in as many as 31% and 2% of pregnant women, respectively.10–12 Gallbladder disease is the most common non-obstetrical cause of maternal hospitalization in the first year postpartum.13
This study prospectively evaluated the incidence and natural history of pregnancy-related gallbladder sludge and stones in the United States and closely examined potential demographic, medical, and behavioral risk factors for their development. Pregnancy may constitute a defined period of metabolic stress in which subclinical tendencies are transiently revealed. For example, pregnant women with gestational diabetes or high blood pressure are at risk of later developing diabetes mellitus or hypertension.14–18 A similar phenomenon may happen with gallbladder sludge and stones. With careful interpretation, our results also may help clarify risk factors for development of gallbladder disease in the general population.
Patients and Methods
Consecutive women attending obstetrics orientation class were approached. Women were excluded because of age younger than 18 years, language comprehension, plans to move away within 3 months, or being more than 20 weeks pregnant. Of 8,929 women approached, 4,897 (55%) were eligible and interested. We excluded an additional 208 women with stones on entry ultrasound and 33 women with a prior cholecystectomy. If sludge was seen on the entry ultrasound, women were followed to see whether sludge progressed to stones. All women gave written informed consent, and the study was approved by the Institutional Review Boards of the participating institutions.
Women were scheduled for an entry ultrasound during their first trimester of pregnancy (9-12 weeks) if possible. Otherwise, their second trimester (16-22 weeks) ultrasound was their entry examination. Additional study ultrasounds were scheduled during the early third trimester (26-32 weeks) and at 4 to 6 weeks postpartum, regardless of the entry ultrasound findings. Serum obtained at 26 to 28 weeks' gestation was tested for fasting glucose and lipid values. Additional serum samples were frozen at −70°C. At entry, subjects completed a questionnaire covering their medical, obstetrical, family, and social histories. Prepregnancy body mass index was calculated from self-reported height and weight immediately before pregnancy. Weight gain during pregnancy was calculated as the difference between predelivery weight and prepregnancy weight. Additional questionnaires were completed in the early third trimester and at 4 to 6 weeks postpartum. Women were invited back at 1 year postpartum for a final study ultrasound and questionnaire. Follow-up data for the first postpartum year were extracted from medical records for all subjects.
Gallbladder ultrasonography was performed with a standard imaging protocol using a 3.5- to 7.0-MHz rotatory sector scanning transducer (ATL Inc., Bothell, WA, or Acuson Corp., Mountain View, CA). All study ultrasounds were performed by sonographers with special training in gallbladder ultrasound and with women fasting or having drunk only sips of water. Findings were recorded by the sonographers, and images and findings were reviewed by one of two designated study radiologists with expertise in gallbladder ultrasound. Sludge was defined as the presence of low-level echoes that shift with position changes and without postacoustic shadowing. Stones were defined as high-amplitude echoes greater than 2 mm in diameter with postacoustic shadowing. In 10,887 scans, there were discrepancies between the radiologist's and the sonographer's readings in 70 regarding the diagnosis of sludge (kappa = .93) and in 25 regarding the diagnosis of stones (kappa = .98). In case of discrepancy, the radiologist's reading was accepted as correct.
Estradiol and leptin levels were measured by radioimmunoassay (ICN Biomedicals, Inc., Costa Mesa, CA, and Linco, St. Charles, MO). Progesterone levels were measured by enzyme-linked immunoassay (Systems Labs, Inc., Webster, TX). Total immunoreactive insulin was measured by a double-antibody radioimmunoassay developed at the Diabetes Endocrinology Research Center at the University of Washington. All assays were performed on stored serum samples obtained at 26 to 28 weeks' gestation.
The cumulative incidence and natural history of gallbladder sludge and stones were determined from the serial gallbladder ultrasounds. Women with new sludge, new stones, or progression of sludge on entry ultrasound to stones were considered to have incident gallbladder disease. The cumulative incidence of sludge or stones was determined, with the denominator representing the total number of study subjects who had at least two ultrasound examinations by that time. We analyzed risk factors for incident gallbladder disease in women with sludge or stones seen on the early postpartum ultrasound. Differences in the risk of developing sludge or stones in various subgroups were compared using t tests or chi-square tests, as appropriate (Stata 8.0, Stata Corp., College Station, TX). Logistic regression models were developed to examine independent risk factors for sludge or stones. Risk factors examined included demographic factors, medical history, body weight, reproductive factors, lipid levels, and behavioral factors including coffee intake, smoking, and alcohol consumption. Two-sided P values less than .05 were considered statistically significant.
Of 4,897 eligible and interested women, 3,254 (66%) with at least two interpretable study ultrasounds were included in this analysis. The number of women with study ultrasound examinations at scheduled times is shown in Table 1.
|No Third Trimester or Postpartum||Third Trimester Only||4–6 Weeks Postpartum Only||Both Third Trimester and 4–6 Weeks Postpartum|
|First trimester, no second trimester (n = 48)||0||26||0||22|
|Second trimester, no first trimester (n = 1,842)||0||684||0||1,158|
|First and second trimester (n = 1,364)||138||349||16||861|
By the second trimester, 3.2% and 1.9% of women had formed new sludge or stones, respectively, and 2% had progression of baseline sludge to stones. By the third trimester, the cumulative incidence of new sludge and stones was 4.5% and 1.8% of women, respectively, with progression of sludge to stones in 1.5%. By the first postpartum examination, 5.1% of women had new sludge, 2.8% had new stones, and 2.3% had progression of baseline sludge to stones. Therefore, by 4 to 6 weeks postpartum, the cumulative incidence of new sludge, new stones, or progression of baseline sludge to stones was 10.2%. Results were similar when we restricted the analysis to only women who had the first-trimester ultrasound (data not shown). Development and regression of sludge and stones was common (Tables 2 and 3). Overall, 4.2% of women had new sludge or stones that persisted to the early postpartum ultrasound.
|First Trimester||Second Trimester||Third Trimester|
|Ultrasound Finding||N||Normal||Sludge||Stones +/− Sludge|
|Normal (n = 1,206)||Normal||1,146||1,067||57||22|
|Stones +/− sludge||21||8||0||13|
|Sludge (n = 4)||Normal||2||0||0||2|
|Stones +/− sludge||2||1||0||1|
|Entry Ultrasound||Final Ultrasound During Pregnancy (Second or Third Trimester)||Postpartum Ultrasound|
|Ultrasound Finding||N||Normal||Sludge||Stones +/− Sludge||Not Done|
|Normal (n = 3,206)||Normal||3,042||1,870||23||30||1,119|
|Stones +/− sludge||52||8||1||24||19|
|Sludge (n = 48)||Normal||42||27||1||0||15|
|Stones +/− sludge||6||1||0||2||3|
Fifty-four women had an ultrasound at 1 year postpartum. Of these, 16 had no prior sludge or stones, 27 had prior sludge, and 10 had prior stones. This ultrasound was normal in 26 of 27 women with prior sludge, but in only 2 of 10 women with prior stones. The remaining 8 women with prior stones had sludge, persistent stones, or a combination of both. New sludge was seen in 1 of the 16 women without prior sludge or stones.
Sludge and stones were asymptomatic in most women. Four women (1.2%) with sludge or stones had symptoms that they attributed to their gallbladder. Four other women (0.1%) without sludge or stones on any ultrasound reported symptoms that they attributed to their gallbladder. Twenty-eight women (0.8%) underwent cholecystectomy for symptomatic gallbladder sludge or stones during the first year postpartum; 5 of these had surgery for severe biliary colic within 6 weeks postpartum. No woman had complications of cholelithiasis, such as cholecystitis, pancreatitis, or cholangitis.
Risk factors for sludge and stones persisting to the early postpartum ultrasound were similar, and results are presented for a combined endpoint of sludge and stones. On univariable analysis, Hispanic women were significantly more likely to develop gallbladder disease than non-Hispanic women, with other racial differences also seen (P < .001) (Table 4). Age, prior hormonal contraception, coffee consumption, and smoking at any time during pregnancy were unrelated to the risk of gallbladder disease. Consumption of alcoholic beverages at any time during pregnancy was associated with a slightly decreased risk of gallbladder disease. The risk of incident gallbladder disease also varied with parity.
|Risk Factor||No Sludge or Stones||New Sludge, New Stones, or Progression of Sludge to Stones||P|
|Age, n (%)||.42|
|<20||335 (92.0)||29 (8.0)|
|20–4||1,169 (90.3)||125 (9.7)|
|25–9||878 (88.3)||116 (11.7)|
|30–4||439 (90.0)||39 (10.0)|
|35–9||118 (90.1)||13 (9.9)|
|≥40||13 (86.7)||2 (13.3)|
|Race, n (%)||.005|
|White||2,022 (89.9)||27 (10.1)|
|Black||377 (91.7)||34 (8.3)|
|Asian/Pacific Islander||221 (93.2)||16 (6.8)|
|Native American||41 (89.1)||5 (10.9)|
|Other/mixed||26 (89.7)||3 (10.3)|
|Hispanic origin, n (%)||<.001|
|Yes||307 (84.1)||58 (15.9)|
|No||2,633 (90.5)||276 (9.5)|
|Any smoking during pregnancy, n (%)||.48|
|Yes||580 (90.3)||62 (9.7)|
|No||2,235 (89.4)||265 (10.6)|
|Any alcohol consumption during pregnancy, n (%)||.02|
|Yes||817 (91.6)||75 (8.4)|
|No||1,998 (88.8)||252 (11.2)|
|Any coffee consumption during pregnancy, n (%)||.82|
|Yes||1,258 (89.7)||144 (10.3)|
|No||1,557 (89.5)||183 (10.5)|
|0||693 (94.9)||37 (5.1)|
|1||634 (97.5)||16 (2.5)|
|≥2||641 (95.2)||32 (4.7)|
|Previous hormonal contraception, n (%)||.83|
|Yes||2,204 (89.9)||247 (10.1)|
|No||728 (89.7)||84 (10.3)|
|Waist–hip ratio at entry||0.84 ± 0.07||0.84 ± 0.07||.53|
|Body mass index prepregnancy, n (%)||<.0001|
|<25||1,288 (97.3)||36 (2.7)|
|25–29.9||443 (95.5)||21 (4.5)|
|≥30||218 (88.3)||29 (11.7)|
|Highest previous body mass index, n (%)||<.0001|
|<25||1,556 (93.1)||116 (6.9)|
|25–29.9||856 (88.7)||109 (11.3)|
|≥30||478 (82.4)||102 (17.6)|
|Weight gain during pregnancy, quartiles, n (%)||.002|
|1||528 (93.1)||39 (6.9)|
|2||561 (96.9)||18 (3.1)|
|3||552 (96.5)||20 (3.5)|
|4||335 (97.4)||9 (2.6)|
|Total cholesterol, mmol/L, mean ± SD||6.20 ± 1.12||6.12 ± 1.26||.58|
|Fasting glucose, mmol/L, mean ± SD||4.11 ± 0.45||4.18 ± 0.55||.20|
|HDL cholesterol, mmol/L, mean ± SD||1.70 ± 0.43||1.62 ± 0.42||.02|
|LDL cholesterol, mmol/L, mean ± SD||3.39 ± 1.06||3.34 ± 1.07||.39|
|Triglycerides, mmol/L, mean ± SD||2.38 ± 0.90||2.59 ± 0.81||.04|
Being overweight (body mass index 25-29.9 kg/m2) or obese (body mass index ≥ 30 kg/m2) immediately prepregnancy was significantly associated with incident gallbladder disease. For example, 2.7% of women with normal prepregnancy body mass index developed incident gallbladder disease, compared with 11.7% of obese women. Weight gain during pregnancy was inversely associated with the risk of incident gallbladder disease.
Because cholesterol gallstones have previously been linked with diabetes and altered lipid metabolism, we examined fasting serum glucose and lipid levels (Table 4). Higher triglyceride levels were associated with an increased risk of sludge or stones, whereas high-density lipoprotein (HDL) cholesterol was inversely related. Total cholesterol, low-density lipoprotein cholesterol, and glucose were not significantly associated with gallbladder disease.
We developed a multivariable model to identify independent risk factors for incident gallbladder disease. Prepregnancy obesity was strongly associated with the risk of incident gallbladder disease (odds ratio, 4.45; 95% CI, 2.59, 7.64) (Table 5). Parity and HDL cholesterol were also inversely associated with incident gallbladder disease.
|Risk Factor||Odds Ratio*||95% CI||P|
|Prepregnancy body mass index, kg/m2||<.001|
|HDL cholesterol (per 1 mmol/dL increase)||0.54||0.30, 0.96||.04|
To examine physiological mechanisms for incident gallbladder disease, we analyzed serum estradiol, progesterone, leptin, and insulin levels measured at 26 to 28 weeks' gestation in a nested case-control study of 50 case and 101 randomly chosen control subjects. Cases and controls had similar serum levels of estradiol (45,803 ± 1,659 vs. 42,808 ± 1,461 pmol/L, P = .29), progesterone (1,221 ± 413 vs. 1,148 ± 305 nmol/L, P = .29), and insulin (153.4 ± 75.1 vs. 158.4 ± 124.3 pmol/L, P = .77). Leptin levels correlated with body mass index (r = 0.68, P < .001). A trend toward higher leptin levels in cases than controls (28.5 ± 12.9 vs. 24.9 ± 11.3 ng/dL; P = .08) was seen. On multivariate analysis, serum leptin and HDL cholesterol were strong predictors of incident gallbladder disease, but serum insulin, estradiol, and progesterone were not (Table 6). The odds of developing gallbladder sludge or stones increased by 5% for each increase of 1 ng/dL in serum leptin level. In further analyses stratified by body mass index, the effect of leptin was strongest in overweight or obese women (odds ratio, 1.13 per 1 ng/dL increase; 95% CI, 1.04, 1.23), whereas no effect was seen in normal weight women (odds ratio, 0.99 per 1 ng/dL increase; 95% CI, 0.93, 1.06). After adjusting for serum leptin, there was no effect of being overweight or obese on the risk of gallbladder disease.
|Risk Factor||Odds Ratio*||95% CI||P|
|Prepregnancy body mass index, kg/m2||.43|
|Serum leptin, per 1 ng/dL increase||1.05||1.01, 1.11||.02|
|HDL cholesterol, per 1 mmol/L increase||0.24||0.09, 0.61||.003|
When gallstones form, changes in bile composition lead to precipitation of bile solutes. These microprecipitates aggregate and grow and are known as biliary sludge (“microlithiasis”). When they are more than 2 mm in size, they produce higher amplitude echoes with post-acoustic shadowing, and are diagnosed as stones. Sludge can spontaneously disappear and re-form over time, or it can evolve to become gallstones. Not all cases of biliary sludge will evolve to gallstones. Conversely, sludge is a necessary precursor to gallstone formation. Changes in bile composition and gallbladder stasis during pregnancy may contribute to sludge and stone formation.19–21
In this study, we examined the incidence, natural history, and risk factors for gallbladder sludge and stones during pregnancy. Over a 7- to 8-month period from the first trimester to the early postpartum, the cumulative incidence of sludge was 5%, whereas an additional 5% developed incident stones or progressed from baseline sludge to stones. Overall, 4.2% had new sludge or stones that were seen on the early postpartum ultrasound. Prior cross-sectional studies of nonpregnant women in the United States have found that 6.5% of women ages 20 to 29 years and 10.2% of women ages 30 to 39 years have prevalent gallstones or a prior cholecystectomy.2 Because regression of stones is relatively uncommon, we thus believe that pregnancy is a high-risk period for gallstone formation. In addition, our results likely underestimate the true incidence of sludge or stones. Transient sludge or stones, not present at the time of any study ultrasound, would not have been detected. Also, the sensitivity of ultrasound for microscopic sludge is only 50% to 60%. Therefore, some women may have had sludge that was not detected by a study ultrasound. Although the clinical significance of transient sludge and stones is not clear, these entities may cause biliary symptoms or complications. In the postpartum period after gallbladder motility is restored, sludge and stones may pass from the gallbladder, potentially causing biliary colic or other complications. Alternatively, postpartum changes in bile composition may favor regression of sludge or stones.22
Most women with gallbladder disease in this study were asymptomatic, but 0.8% underwent cholecystectomy in the first postpartum year. Because more than 4 million women give birth annually,23 we can estimate that more than 32,000 young, otherwise healthy women will require postpartum cholecystectomy each year. Thus, pregnancy-associated gallbladder disease is a significant cause of morbidity in young, otherwise healthy women.
Our results differ from those of Maringhini et al.,11 who found incident sludge in 31% and stones in 2% of pregnant women. Differences in the sonographic definition of sludge and stones may account for part of this discrepancy. Also, our population was slightly younger and less overweight. In contrast to our study, Maringhini et al. found that sludge was inversely related to body weight. However, they did not account for weight gain or loss during pregnancy, which may confound the relationship between prepregnancy weight and sludge. Women who weighed more prepregnancy in our study gained less weight during pregnancy, likely explaining the inverse association between weight gain during pregnancy and gallbladder disease.
The independent risk factors for incident sludge or stones were prepregnancy body mass index, HDL cholesterol, and parity. The association of prepregnancy body mass index is consistent with prior data showing an increasing prevalence of gallstones with increasing weight in the general population.9, 24 Our data further demonstrate that the risk of developing new gallbladder disease is strongly associated with being obese, but only moderately associated with being overweight. As the prevalence of overweight and obesity in young women increases,25 pregnancy-associated gallbladder disease may become an even greater problem.
Similarly, the inverse association of parity with incident gallbladder disease was unexpected. However, this is likely attributable to our selection criteria. Women of higher parity with an underlying tendency to form stones may have already done so in previous pregnancies. These women would have been excluded by our criteria. Women of higher parity and without an underlying tendency toward stone formation would still be eligible for inclusion and would be less likely to form stones during the study pregnancy. As an approximation, our data indicate that approximately 5.1% of women develop gallbladder disease after one pregnancy, 7.6% after two pregnancies, and 12.3% after 3 or more pregnancies. These data are generally similar to those found in other population-based studies.2, 5, 7, 8
HDL cholesterol was inversely related to sludge and stone formation, similar to findings with prevalent stones.26 HDL promotes reverse cholesterol transport by facilitating cholesterol uptake into the liver. Preformed cholesterol in HDL is a major source of biliary cholesterol. Therefore, a direct mechanistic correlation seems unlikely. A simplistic explanation may be that low HDL levels are associated with reduced capacity to solubilize the total body cholesterol pool. The excess cholesterol may be distributed to the biliary system, where it precipitates as sludge or stones. Alternatively, low HDL co-exists with obesity, insulin resistance, and hypertriglyceridemia. The harmful effect of low HDL may be mediated through its association with obesity or insulin resistance.
The only potentially modifiable risk factor for incident gallbladder disease was body mass index. Possible mediators of the association between obesity and gallbladder disease include the hormones insulin and leptin, both of which are associated with body mass index. Insulin levels were associated with prevalent gallbladder disease in prior studies, whereas leptin levels were not.27, 28 Leptin may mediate secretion of cholesterol into bile by the liver,29, 30 potentially promoting cholesterol lithogenesis in obese patients. We found leptin levels were increased relative to those in nonpregnant women, consistent with known changes in leptin levels during pregnancy.31 Although serum leptin level may be a proxy measure for leptin resistance commonly found in obesity, our data show that it is independently associated with incident gallbladder disease. Our data suggest leptin may mediate the association between obesity and gallbladder disease. Further studies to confirm this finding are needed.
In conclusion, more than 4% of pregnant women have incident gallbladder sludge or stones persisting to the early postpartum. Although most women will remain asymptomatic, 0.8% will require cholecystectomy within the first year postpartum. Therefore, gallbladder disease is a significant cause of morbidity for young, otherwise healthy women. Being overweight or obese prepregnancy is a significant, but potentially modifiable, risk factor for developing gallbladder disease during pregnancy. Further interventions aimed at controlling body weight may be beneficial in reducing the burden of this disease in young women.
The authors acknowledge contributions to the study by Andrea K. Herron, Lori J. Green, Dianne Walkup, Lorna Imbruglio, Nancy Allison, and Brett Gates.
- 1Gallstones. Digestive Diseases in the United States: Epidemiology and Impact, 1994. NIH Publication No. 94-1447. Bethesda, MD: National Institutes of Health, 1994..
- 3American Gastroenterological Association. The burden of gastrointestinal diseases. Bethesda, MD: The American Gastroenterological Association, 2001.
- 23Centers for Disease Control and Prevention. National Vital Statistics Report. Volume 2003. 50 (14) ed: Centers for Disease Control and Prevention, 2002.