Changes in hepatitis A and B vaccination rates in adult patients with chronic liver diseases and diabetes in the U.S. population


  • Zobair M. Younossi,

    Corresponding author
    1. Center for Liver Diseases and Department of Medicine, Inova Fairfax Hospital, Falls Church, VA
    2. Betty and Guy Beatty Center for Integrated Research, Inova Health System, Falls Church, VA
    • Betty and Guy Beatty Center for Integrated Research, Claude Moore Health Education and Research Building, 3300 Gallows Road, Falls Church, VA 22042
    Search for more papers by this author
    • fax: 703-776-4386

  • Maria Stepanova

    1. Center for Liver Diseases and Department of Medicine, Inova Fairfax Hospital, Falls Church, VA
    2. Betty and Guy Beatty Center for Integrated Research, Inova Health System, Falls Church, VA
    Search for more papers by this author

  • Potential conflict of interest: Dr. Younossi is a consultant for and advises Salix. He also advises Vertex and Tibotec.

  • This study has been supported, in part, by the Liver Disease Outcomes Fund of the Center for Liver Diseases at Inova Fairfax Hospital, Inova Health System, Falls Church, VA.


Professional societies recommend hepatitis A and hepatitis B immunization for individuals with chronic liver disease (CLD), but the degree of implementation is unknown. Data were obtained from the National Health and Nutrition Examination Surveys (NHANES) conducted in 1999-2008. For the entire study population and for those with CLD and diabetes, we determined the rates and independent predictors of history of hepatitis A and hepatitis B (HepA and HepB) vaccinations, of their effectiveness, and of seroprevalence of hepatitis A antibody and anti-HB surface antibody. In total, 24,871 participants from NHANES were included: 14,886 (1999-2004) and 9,985 (2005-2008). Of these individuals, 14.0% had CLD and 8.6% had diabetes. During the study period, HepA vaccination in CLD increased from 13.3% ± 1.0% to 20.0% ± 1.5%, HepB vaccination increased from 23.4% ± 1.2% to 32.1% ± 1.5%. Of subtypes of CLD, HepA vaccination rates increased only in nonalcoholic fatty liver disease (NAFLD), whereas HepB vaccination increased for patients with hepatitis C and nonalcoholic fatty liver disease. In the diabetic cohort, HepA vaccination rates increased from 9.3% ± 1.1% to 15.4% ± 1.7% and HepB rates increased from 15.2% ± 1.5% to 22.4% ± 1.7%. All changes were similar to those observed in the general population. The quality measure (QM) for HepA in the general population decreased from 44.4% ± 1.2% in 1999-2004 to 41.7% ± 1.9% in 2005-2008, and similar changes were noted for all subcohorts. On the other hand, QM for HepB increased from 31.7% ± 0.9% to 40.7% ± 1.0% in the population, whereas no changes in QM were noted in any diagnostic cohort except for NAFLD. Conclusions: Although vaccination rates in CLD and diabetic cohorts are increasing, they remain low. Given the public health implications of acute hepatitis A and hepatitis B in patients with CLD, better implementation of the vaccination recommendations for these populations is warranted. (HEPATOLOGY 2011)

The Centers for Disease Control and Prevention estimates that liver disease is currently the 12th leading cause of death in the United States.1 Liver-related mortality usually results from complications of chronic liver disease, including advanced cirrhosis and hepatocelllar carcinoma (HCC). Despite a recent decline in many other cancers, the incidence of HCC continues to increase, especially in men.2-4 Furthermore, chronic liver disease (CLD) and related complications are associated with increased mortality, severely impaired quality of life, and substantial resource utilization.5-8 Despite a decline in the incidence of hepatitis C, other liver diseases, such as diabetes and obesity-related nonalcoholic fatty liver disease (NAFLD), are increasing.9-11

Increasing evidence suggests that patients with preexisting CLD are at risk for a severe liver disease after acute infection with hepatitis A and/or hepatitis B viruses.12-15 This superinfection in patients with preexisting CLD may have a rapidly progressive course, leading to liver failure and death.16, 17 Additionally, severe acute hepatitis B infection has also been reported in patients with type II diabetes (diabetes mellitus [DM]).18 Given the high prevalence of NAFLD in patients with DM, many diabetics may have underlying CLD related to NAFLD.9-11, 19, 20 In fact, patients with DM not only have high prevalence of NAFLD, but DM is an independent risk factor for the progressive form of NAFLD or nonalcoholic steatohepatitis (NASH), hepatic fibrosis, and liver-related mortality.9, 11, 19, 20 Therefore, underlying NASH and fibrosis could potentially predispose patients with DM to severe liver disease related to acute hepatitis A and/or B infection. Furthermore, outbreaks of acute B infection, most likely related to shared glucose-monitoring equipment, have been reported in diabetics.18 Based on these data, there may be a rationale to consider patients with DM who are at a high risk for having underlying CLD for HepB vaccination.

Several vaccines against hepatitis A and hepatitis B viruses are currently available.21-24 Although the effectiveness of HepA and HepB vaccination may decline in patients with more advanced liver diseases or in the post-transplant setting,25, 26 they are believed to be safe and effective for most patients with CLD.27, 28 However, despite the risk of acute hepatitis A and hepatitis B infection in patients with CLD and the reported protective immunogenicity of the current HepA and HepB vaccines, recommendations for vaccinating CLD patients against hepatitis A and hepatitis B are nonuniform. The Centers for Disease Control and Prevention (CDC) recommends HepA vaccination for post-transplant patients and those with documented CLD.29, 30 The American Association for the Study of Liver Diseases (AASLD) recommends HepA vaccination for all hepatitis B– and hepatitis C–infected patients, regardless of the stage of their liver disease.31, 32 Furthermore, the National Institutes of Health (NIH; National Institutes of Health) and AASLD all recommend HepB vaccination only for chronic hepatitis C patients who are at risk for hepatitis B infection, whereas the CDC recommends HepB vaccine to all individuals with CLD.29-36 On the other hand, there is still no recommendation for vaccination against viral hepatitis for individuals with DM, despite consistent evidence supporting the high prevalence of CLD in diabetics.

This study used recent U.S. population data to assess HepA and HepB vaccination rates in patients with CLD.


ALD, alcoholic liver disease; ALT, alanine transaminase; AST, aspartate aminotransferase; BMI, body mass index; CH-B, chronic hepatitis B; CH-C, chronic hepatitis C; CLD, chronic liver disease; ELISA, enzyme-linked immunosorbent assay; HAV, hepatitis A virus; anti-HAV, hepatitis A antibody; HBV, hepatitis B virus; anti-HBs, hepatitis B surface antibody; anti-HBc, hepatitis B core antibody; HBsAg, hepatitis B surface antigen; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; anti-HCV, hepatitis C infection; HDL, high-density lipoprotein; HepA vaccine, hepatitis A vaccine; Hep B vaccine, hepatitis B vaccine; HIV, human immunodeficiency virus; HOMA, homeostasis model assessment; IgG, immunoglobulin G; IR, insulin resistance; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; NHANES, National Health and Nutrition Examination Survey; LDL, low-density lipoprotein; PCR, polymerase chain reaction; QM, quality measure.

Patients and Methods

Study Population.

Data obtained from five continuous cycles of the National Health and Nutrition Examination Survey (NHANES), conducted between 1999 and 2008, were combined into two larger periods: 1999-2004 and 2005-2008. The NHANES is a nationwide survey representing the health and nutritional status of the of the noninstitutionalized civilian U.S. population. These data were collected by the U.S. National Center for Health Statistics (NCHS) of the CDC via household interviews, physical examinations, and laboratory data, including blood and urine samples collected in designated examination centers. The surveys included similar questionnaires and methods for serum and blood assays. Demographic, clinical, and laboratory parameters were transformed, according to the provided guidelines, to make the data comparable between the cycles.37 The distribution of participants was representative of the U.S. population after weighting on the basis of age, gender, level of education, and race or ethnic background.37

Inclusion criteria were the following: age of 18 years or older and availability of complete demographic, social history, clinical data, socioeconomic status, medical conditions, and vaccination questionnaires. Body mass index (BMI), waist circumference, and blood pressure were measured for all NHANES participants at time of examination. Additional laboratory tests included the following: fasting serum glucose and insulin, triglycerides, high-density lipoprotein (HDL), low-density lipoprotein (LDL) and total cholesterol, aspartate aminotransferase (AST), alanine transaminase (ALT), and transferrin saturation. Hepatitis A total immunoglobulin G (IgG) antibody (anti-HAV), hepatitis B core antibody (anti-HBc), hepatitis B surface antibody (anti-HBs), and hepatitis C antibody (anti-HCV) were determined with the enzyme-linked immunosorbent assay (ELISA), and hepatitis B surface antigen (HBsAg) was tested in duplicate using AUSZYME Monoclonal test (Abbott Diagnostics, Abbott Park, IL). HCV RNA was determined by polymerase chain reaction (PCR).

Participants with data insufficient for ruling in or ruling out CLD, diabetes, or without vaccination questionnaires were excluded. Excluded participants were not different from included participants in any other way.


Medical definitions used throughout the study are summarized in Table 1. Comorbidities that could be associated with the prevalence and effectiveness of HepA and HepB vaccination were assessed using the Medical Conditions questionnaires completed by NHANES participants. Human immunodeficiency virus (HIV) positivity was determined with the ELISA anti-HIV test (run for individuals of age under 50 years). The health insurance status of the participants was determined using the NHANES Health Insurance Questionnaire. A modified insurance questionnaire was used in the 2005-2008 study cycle; so, the data were transformed as recommended.38 The primary source of healthcare was determined using the Hospital Utilization and Access to Care questionnaire.

Table 1. Definitions of Medical Conditions Used in the Study
 DefinitionNHANES Source
Excessive alcohol consumption>20 g per day for men and >10 g per day for women within a year before the completion of data collectionAlcohol Use questionnaire; calculated using self-reported amount and frequency of alcohol consumption
Injection drug useIntravenous injection of illicit drugsDrug Use questionnaire
Positive smoking historyOngoing smoking or more than 100 cigarettes in the past 5 yearsSmoking and Tobacco Use questionnaire
ObesityBody mass index (BMI) ≥30Body Measurements examination
Visceral obesityWaist circumference of 88 cm or more for women and 102 cm or more for menBody Measurements examination
Diabetes mellitus (DM)Fasting glucose equal or greater than 126 mg/dL or the history of oral hypoglycemics or insulin use or bothDiabetes questionnaire; plasma fasting glucose and insulin laboratory data
Insulin resistance (IR)Homeostasis model assessment score (HOMA) score of 3.0 or greaterPlasma fasting glucose and insulin laboratory data
HypertensionSystolic blood pressure (BP) of ≥140 mm Hg, diastolic BP of ≥90 mm Hg, or the use of oral antihypertensive medicationsBlood pressure examination; Blood Pressure and Cholesterol questionnaire
HypercholesterolemiaTotal cholesterol ≥200 mg/dL, or LDL ≥ 139 mg/dL, or HDL less than 40 mg/dL for men and less than 50 mg/dL for women, or the use of prescribed medicationsTotal cholesterol, HDL cholesterol, triglycerides and LDL cholesterol laboratory data; Blood Pressure and Cholesterol questionnaire
Ischemic heart disease (IHD)History of coronary artery disease, angina, or heart attackMedical Conditions questionnaire
Chronic obstructive pulmonary disease (COPD)History of chronic bronchitis or emphysemaMedical Conditions questionnaire
CancerHistory of any cancerMedical Conditions questionnaire
Elevated serum aminotransferasesALT >40 U/L or AST >37 U/L in men and ALT or AST >31 U/L in women(Standard) biochemistry profile

Etiology of CLD.

Five major categories of CLD were considered: alcohol-related liver disease (ALD), chronic hepatitis B (CH-B), chronic hepatitis C (CH-C), NAFLD, and iron overload (defined at the transferrin saturation level of 50% or higher). Alcohol-related liver disease was presumed in subjects who reported excessive alcohol consumption and had elevated serum aminotransferases (both defined in Table 1). CH-B was defined as being positive for HBsAg, regardless of the anti-HBc or anti-HBs status. It is important to note that the length time for HBsAg-positivity was not available in the NHANES data collection, so we could not reliably distinguish between those with recent and chronic HepB infection. Furthermore, individuals with positive serology for anti-HCV were defined as hepatitis C antibody positive (HCV+). Of the HCV+ patients with available data, approximately 75% were HCV RNA positive and were considered to have CH-C. Finally, subjects were presumed to have NAFLD if they had elevated serum aminotransferases in the absence of any other evidence of CLD, such as alcohol use, iron overload, or a positive hepatitis B or hepatitis C serologic test.

Together, individuals with ALD, CH-B, HCV, iron overload, and NAFLD were included in the cohort of CLD. Subjects without evidence of CLD, according to the inclusion criteria described here, were considered to have no CLD and were used as controls. Given the lack of additional iron studies needed to establish the diagnosis of hemochromatosis, we elected not to study iron overload separately.

Because NHANES does not collect information on the date of diagnosis for diabetes, we were unable to distinguish between type I and II diabetes. However, because we have limited our assessment of diabetes to adults only, we believe that these patients with diabetes mostly have type II diabetes.

Vaccination Status.

A self-reported history of receiving hepatitis A and B vaccinations was collected from the NHANES Immunization questionnaires. Vaccination against hepatitis A (or hepatitis B) was defined as receiving at least one dose of the vaccine. If fewer than two doses of HepA vaccine (fewer than three for HepB vaccine) were reported, the series was presumed incomplete. Effective vaccination was defined as any number of doses of HepA (or HepB) vaccine and positive anti-HAV (positive anti-HBs without anti-HBc). The quality measure (QM) for HepA vaccination (or HepB vaccination) was defined for all study participants using the Medicare definition for QM for patients with hepatitis C: “percentage of individuals aged 18 years and older with a diagnosis of hepatitis C who received at least one injection of hepatitis A (or B) vaccine, or who had documented immunity to hepatitis A (or B)”.39, 40 When defining QM, in addition to anti-HBs immunity, natural immunity for hepatitis B was defined as positive anti-HBc (regardless of anti-HBs) in the absence of HBsAg. Finally, ineffective vaccination against hepatitis A was defined as the absence of anti-HAV in individuals with a positive history of HepA vaccination by questionnaire. Similarly, ineffective HepB vaccination was the absence of any positive HepB serology in individuals who reported a positive history of HepB vaccination.

Four parameters were used to assess changes in vaccination or immunity rates against hepatitis A and hepatitis B: (1) self-reported history of vaccination against hepatitis A and hepatitis B; (2) effective vaccination against hepatitis A and hepatitis B; (3) seroprevalence rates for anti-HAV and anti-HBs; and (4) QM for HepA and HepB vaccines. For each NHANES cycle, we calculated each parameter for both HepA and HepB vaccination, together with their independent predictors in the general U.S. population and in individuals with CLD (and subtypes of CLD) as well as those with diabetes.

Statistical Analyses.

Sample weights were used to account for nonresponse and unequal selection probabilities for certain categories of the population, and stratum and sampling units accounted for the survey design effects by using Taylor series linearization. When merging NHANES study cycles, appropriate selection of sampling weights and adjustment coefficients were applied according to the NHANES Analytic and Reporting Guidelines.37 The prevalence of demographic, social, and clinical parameters was compared across the two study cycles by using the stratum-specific chi-square test for independence. Logistic regression was used to identify independent predictors of being vaccinated, meeting QM and predictors of ineffective vaccination for hepatitis A and hepatitis B separately. The list of potential predictors included all studied demographic and socioeconomic parameters, together with medical conditions. Odds ratios with P values below 0.05 were considered significant. All analyses were run with SAS 9.1 and SUDAAN 10.0 (SAS Institute Inc., Cary, NC). The study was approved by Inova Institutional Review Board.


Study Population.

Of the initial study population (51,623 participants from NHANES 1999-2008), 24,871 were considered eligible (14,886 in the 1999-2004 cycle and 9,985 in the 2005-2008 cycle). Of those, 3,233 individuals (14.04% ± 0.34%) fulfilled our definition of CLD and 2,772 (8.57% ± 0.29%) had diabetes.

The most relevant clinicodemographic differences between the two study cycles are noted in Table 2. As expected, in the later cycle, individuals became more obese, and the percentage of “baby boomers” (i.e., age 45-65) increased over time. The rates of most of other socioeconomic and clinical parameters did not change over the past decade, although we noted a different pattern of healthcare facility use (from clinics and outpatient hospital visits to doctors' offices and hospital emergency rooms) and an increase in the number of government-sponsored health insurance plans (Table 2).

Table 2. General Characteristics of the U.S. Population Between 1999 and 2008*
Study cycle1999-20042005-2008P Value1999-2008
  • *

    NHANES data; parameters are reported as mean ± standard error.

  • Abbreviations: NAFLD, nonalcoholic fatty liver disease; HIV, human immunodeficiency virus.

Demographic summary    
 Caucasian (%)71.88 ± 2.2470.37 ± 2.780.671571.28 ± 1.75
 African American (%)11.21 ± 1.2511.85 ± 1.680.761711.47 ± 1.01
 Hispanic (%)12.97 ± 1.6012.24 ± 1.700.755212.68 ± 1.18
 Other (%)3.93 ± 0.425.54 ± 0.600.02824.58 ± 0.35
 Male (%)47.88 ± 0.3748.63 ± 0.460.200448.18 ± 0.29
 Age <45 years (%)52.76 ± 0.9750.26 ± 1.320.134851.76 ± 0.80
 45 ≤ age < 55 (%)19.93 ± 0.6321.53 ± 0.660.084320.57 ± 0.47
 55 ≤ age < 65 (%)12.08 ± 0.4714.26 ± 0.640.007412.95 ± 0.39
 Age ≥ 65 (%)15.23 ± 0.5013.94 ± 0.890.202914.72 ± 0.47
 Military service (%)13.72 ± 0.4611.27 ± 0.660.002612.74 ± 0.40
 College degree (%)24.14 ± 1.1326.30 ± 1.510.251125.01 ± 0.91
 Married (%)56.43 ± 1.0857.01 ± 1.150.715356.67 ± 0.80
Social behavior    
 Excessive alcohol use (%)8.85 ± 0.429.58 ± 0.540.28779.14 ± 0.33
 Smoking (%)33.38 ± 0.9931.50 ± 1.270.242832.62 ± 0.79
 Injected drug use (%)3.91 ± 0.422.28 ± 0.280.00142.89 ± 0.24
Access to healthcare    
 Insured (%)82.11 ± 0.9780.82 ± 1.070.366181.59 ± 0.72
 Private insurance (%)67.31 ± 1.3265.21 ± 1.330.263966.46 ± 0.96
 Medicare/Medicaid (%)19.79 ± 0.6618.35 ± 1.070.250319.20 ± 0.58
 Government insurance (%)4.60 ± 0.417.23 ± 0.650.00125.66 ± 0.38
 Single service insurance (%)3.65 ± 0.414.68 ± 0.570.14454.07 ± 0.34
 Uninsured for 3+ years (%)10.04 ± 0.74NANA10.04 ± 0.74
Primary source of healthcare: none (%)15.29 ± 0.7314.66 ± 0.870.576915.04 ± 0.56
 Clinic or health center (%)17.44 ± 1.6414.25 ± 1.120.111016.16 ± 1.09
 Doctor's office or HMO (%)62.49 ± 1.6466.32 ± 1.510.092364.02 ± 1.17
 Hospital ER (%)1.69 ± 0.162.17 ± 0.240.09571.88 ± 0.14
 Hospital outpatient department (%)1.94 ± 0.231.23 ± 0.150.01001.66 ± 0.15
Chronic liver disease    
 Alcoholic liver disease (%)2.20 ± 0.182.09 ± 0.220.69522.16 ± 0.14
 Iron overload (%)2.76 ± 0.201.63 ± 0.400.02092.59 ± 0.18
 Chronic hepatitis B (%)0.33 ± 0.060.34 ± 0.080.91710.34 ± 0.05
 HCV+ (%)1.91 ± 0.191.61 ± 0.150.21501.79 ± 0.13
 Chronic hepatitis C (%)1.44 ±+ 0.191.21 ± 0.150.22751.35 ± 0.13
 NAFLD (%)9.81 ± 0.3310.91 ± 0.540.081410.25 ± 0.29
 Any chronic liver disease (%)14.48 ± 0.3813.37 ± 0.600.127414.04 ± 0.34
Other metabolic disorders    
 Insulin resistance (%)33.13 ± 0.9536.11 ± 1.410.082234.32 ± 0.82
 Diabetes (%)8.14 ± 0.339.20 ± 0.500.08258.57 ± 0.29
 Hypercholesterolemia (%)72.12 ± 0.6870.28 ± 0.860.091271.38 ± 0.54
 Hypertension (%)27.83 ± 0.7628.55 ± 1.020.572128.12 ± 0.61
 Obesity (%)30.69 ± 0.6433.75 ± 1.110.021431.93 ± 0.61
 Visceral obesity (%)48.67 ± 0.8251.81 ± 1.280.045749.92 ± 0.73
Other comorbidities    
 Asthma (%)12.34 ± 0.4114.35 ± 0.560.005213.15 ± 0.35
 Congestive heart failure (%)2.15 ± 0.152.17 ± 0.190.92312.16 ± 0.12
 Ischemic heart disease (%)6.16 ± 0.335.24 ± 0.310.04285.79 ± 0.24
 Stroke (%)2.45 ± 0.162.63 ± 0.220.52442.52 ± 0.13
 Chronic obstructive pulmonary disease (%)7.84 ± 0.397.57 ± 0.510.67797.73 ± 0.31
 Self-reported chronic liver disease (%)1.60 ± 0.152.09 ± 0.180.03911.79 ± 0.12
 Cancer (%)8.07 ± 0.358.12 ± 0.480.93278.09 ± 0.28
 Kidney failure (%)2.11 ± 0.162.12 ± 0.190.95302.11 ± 0.12
 HIV (%)0.28 ± 0.060.28 ± 0.070.95390.28 ± 0.04

Our analysis reveals that over the past decade, the overall seroprevalence of anti-HAV antibody in the adult U.S. population decreased, but more individuals reported a history of hepatitis A vaccination and the rate of effective vaccination also increased (Table 3). Although both self-reported and effective HepA vaccination rates increased between the two cycles, both rates remained quite low. Furthermore, the QM rate for hepatitis A did not change over the study period because of an increase in the HepA vaccination rates, counterbalanced by a decrease in the incidence of hepatitis A infection (Table 3).

Table 3. Vaccination and Immunity Parameters in Individuals With Chronic Liver Disease and Diabetes*
  Hepatitis AHepatitis B
CohortYearsSelf-Reported Vaccination (%)Effective Vaccination (%)Seropositivity for Anti-HAV (%)QM (%)Self-Reported Vaccination (%)Effective Vaccination (%)Seropositivity for Anti-HBs (%)QM (%)
  • Parameters are reported as mean ± standard error.

  • *

    NHANES 1999-2008.

  • Different (P < 0.05) from controls in the same study cycle.

  • Abbreviations: anti-HAV, hepatitis A antibody; QM, quality measure; anti-HBs, hepatitis B surface antibody; HCV+, hepatitis C virus; NAFLD, nonalcoholic fatty liver disease.

U.S. population1999-200411.93 ± 0.545.92 ± 0.3738.42 ± 1.1644.44 ± 1.1524.08 ± 0.7810.38 ± 0.5213.44 ± 0.5631.69 ± 0.85
 2005-200820.52 ± 0.949.74 ± 1.0430.92 ± 1.6941.67 ± 1.8534.32 ± 0.8116.11 ± 0.5719.84 ± 0.6240.67 ± 0.97
Chronic liver1999-200413.33 ± 0.986.74 ± 0.7639.95 ± 1.8546.37 ± 1.7323.39 ± 1.208.23 ± 0.7811.50 ± 0.9034.86 ± 1.41
disease2005-200820.01 ± 1.519.54 ± 1.5030.06 ± 2.6940.74 ± 2.5032.11 ± 1.5012.80 ± 1.0915.78 ± 1.0540.51 ± 1.58
Alcoholic1999-200416.90 ± 3.148.80 ± 2.2135.27 ± 4.0342.97 ± 4.3018.86 ± 3.125.71 ± 2.077.47 ± 2.1735.39 ± 4.22
liver disease2005-200812.75 ± 3.453.30 ± 2.5718.56 ± 5.7023.51 ± 6.9426.41 ± 4.2511.36 ± 3.6912.39 ± 3.5332.25 ± 4.35
HCV+1999-200416.20 ± 3.526.53 ± 1.9644.39 ± 4.0653.18 ± 4.6320.38 ± 3.572.56 ± 1.365.75 ± 1.8657.80 ± 4.76
 2005-200824.37 ± 4.5013.06 ± 4.6731.86 ± 6.5343.00 ± 7.1135.13 ± 4.769.48 ± 2.7611.31 ± 2.9256.49 ± 5.78
NAFLD1999-200412.51 ± 1.256.60 ± 0.8639.40 ± 2.1445.48 ± 1.9522.78 ± 1.528.82 ± 0.9812.12 ± 1.1730.05 ± 1.74
 2005-200819.62 ± 1.778.62 ± 1.8730.72 ± 3.2842.77 ± 3.1731.72 ± 1.8813.14 ± 1.2316.73 ± 1.2339.24 ± 1.97
Diabetes1999-20049.34 ± 1.055.55 ± 0.8356.30 ± 2.0060.20 ± 2.1215.16 ± 1.454.23 ± 0.746.79 ± 0.8825.36 ± 1.76
 2005-200815.35 ± 1.678.34 ± 1.8243.39 ± 3.9151.11 ± 3.7922.42 ± 1.724.42 ± 0.775.98 ± 0.8528.80 ± 2.10

Similarly, seroprevalence for anti-HBs in the U.S. population increased between the study cycles. Self-reported hepatitis B vaccination as well as effective vaccination and QM for hepatitis B also increased simultaneously (Table 3). These findings are consistent with an increase in HepB vaccination rate, accompanied by the stable prevalence of both chronic hepatitis B and natural immunity against HBV.

A summary of independent predictors of HepA and HepB vaccination parameters in the entire U.S. population appears in Table 4. All clinical, demographic, and social parameters presented in Table 2 were included in the initial list of potential predictors, but only those with significant odds ratios for association with an outcome are listed.

Table 4. Independent Predictors of Self-Reported Vaccination and QM for Hepatitis A and B in the U.S. Between 1999 and 2008* (NHANES 1999-2008)
 Hepatitis AHepatitis B
 Self-Reported VaccinationQMSelf-Reported VaccinationQM
  • *

    Odds ratio (95% CI).

  • Abbreviations: QM, quality measure; ER, emergency room; HCV+, hepatitis C virus positive; HIV, human immunodeficiency virus.

Caucasian0.65 (0.49-0.86)0.57 (0.47-0.69)0.16 (0.13-0.19)0.15 (0.11-0.22)0.92 (0.75-1.12)0.83 (0.7-0.98)0.73 (0.61-0.87)0.63 (0.53-0.75)
AfricanAmerican0.73 (0.56-0.97)0.73 (0.56-0.96)0.31 (0.24-0.39)0.28 (0.20-0.39)1.12 (0.88-1.42)0.97 (0.77-1.23)1.26 (1.02-1.55)0.94 (0.76-1.16)
Male1.03 (0.87-1.22)0.89 (0.74-1.07)0.91 (0.82-1.01)0.86 (0.71-1.06)0.61 (0.53-0.70)0.65 (0.56-0.77)0.75 (0.66-0.84)0.71 (0.6-0.83)
Age <45 years1.37 (1.15-1.64)1.66 (1.4-1.95)0.59 (0.51-0.68)0.79 (0.63-1.00)2.26 (1.98-2.57)2.43 (2.07-2.87)1.72 (1.52-1.93)2.11 (1.81-2.45)
Age ≥65 years0.88 (0.63-1.23)0.55 (0.41-0.75)2.48 (1.94-3.17)2.21 (1.40-3.50)0.37 (0.25-0.54)0.38 (0.28-0.53)0.56 (0.42-0.75)0.43 (0.33-0.55)
Military service2.28 (1.77-2.92)1.89 (1.49-2.39)1.39 (1.13-1.71)1.34 (0.98-1.84)1.78 (1.43-2.2)1.55 (1.21-1.97)1.51 (1.24-1.85)1.50 (1.20-1.87)
College degree1.55 (1.28-1.89)1.48 (1.20-1.81)0.97 (0.85-1.1)0.89 (0.67-1.19)1.28 (1.09-1.51)1.49 (1.23-1.81)1.25 (1.07-1.45)1.47 (1.22-1.78)
Married0.73 (0.62-0.86)0.77 (0.65-0.91)1.03 (0.9-1.18)0.95 (0.74-1.21)0.70 (0.61-0.81)0.70 (0.60-0.81)0.74 (0.65-0.84)0.62 (0.54-0.72)
Excessive alcohol use1.01 (0.78-1.31)0.75 (0.57-0.99)0.95 (0.77-1.18)0.74 (0.59-0.93)0.93 (0.74-1.17)0.76 (0.63-0.91)0.92 (0.74-1.14)0.74 (0.62-0.88)
Private insurance1.07 (0.85-1.34)0.98 (0.80-1.20)0.88 (0.74-1.04)0.62 (0.50-0.77)1.15 (0.92-1.44)1.22 (1.01-1.47)1.04 (0.87-1.24)1.14 (0.93-1.40)
Medicare/ Medicaid0.91 (0.67-1.22)0.96 (0.71-1.29)1.25 (1.00-1.56)0.80 (0.51-1.27)0.72 (0.55-0.94)0.83 (0.64-1.06)0.76 (0.60-0.97)0.87 (0.69-1.09)
Government insurance1.42 (1.01-1.99)1.25 (0.97-1.62)0.97 (0.74-1.26)0.87 (0.65-1.17)1.15 (0.81-1.62)1.49 (1.15-1.92)1.13 (0.82-1.55)1.36 (1.06-1.73)
Uninsured ≥3 years0.82 (0.58-1.14)NA1.35 (1.07-1.72)NA0.68 (0.52-0.89)NA0.74 (0.59-0.94)NA
Primary healthcare: clinic1.27 (0.94-1.71)1.04 (0.82-1.31)1.09 (0.86-1.38)0.83 (0.61-1.11)1.31 (1.05-1.64)1.46 (1.09-1.95)1.25 (1.03-1.52)1.38 (1.08-1.77)
Primary healthcare: doctor1.07 (0.84-1.35)0.90 (0.72-1.13)0.80 (0.66-0.97)0.77 (0.59-1.00)1.05 (0.86-1.29)1.29 (1.03-1.61)1.01 (0.84-1.20)1.21 (1.01-1.46)
Primary healthcare: ER1.48 (0.81-2.72)0.92 (0.58-1.46)1.00 (0.64-1.57)0.49 (0.28-0.86)1.32 (0.81-2.16)1.52 (1.05-2.21)1.30 (0.84-2.01)1.46 (1.02-2.09)
HCV+1.41 (0.76-2.63)0.91 (0.49-1.70)1.28 (0.80-2.06)1.01 (0.44-2.27)0.9 (0.54-1.49)0.99 (0.54-1.80)3.29 (1.99-5.42)1.75 (0.99-3.07)
HIV3.54 (1.50-8.36)2.26 (0.53-9.55)1.64 (0.61-4.42)5.56 (1.05-29.58)3.15 (1.19-8.3)3.93 (1.17-13.22)4.8 (1.59-14.50)12.36 (3.08-49.57)

CLD Population.

In the CLD cohort, changes in seroprevalence of anti-HAV, self-reported vaccination, effective vaccination, and QM for hepatitis A were all similar to the general population (Table 3). Of the studied subtypes of CLD, the same pattern was found for the NAFLD cohort. However, no changes in anti-HAV seroprevalence and vaccination rates were noted for both the ALD and HCV cohorts. Moreover, both study cycles showed no difference from controls in the prevalence of hepatitis A vaccination for the CLD cohort as well as for HCV and NAFLD, whereas in the ALD cohort, the HepA vaccination rate decreased significantly.

Moreover, in the past decade, HepB vaccination rates in patients with CLD and most of its subcohorts (except for ALD) (Table 3) increased, together with seropositivity rates, effective vaccination rates, and QM. Nonetheless, neither CLD nor its subtypes were different from controls in terms of HepB vaccination rates, whereas the anti-HBs positivity and effective HepB vaccination rates among individuals with CLD remained lower than controls. Additionally, hepatitis B QM increased in the CLD cohort and the NAFLD subcohort, whereas no changes were noted in the HCV and ALD subcohorts, and no differences were observed in QM versus controls for the entire CLD cohort and all its subtypes, except for HCV where the hepatitis B QM rate was higher than controls. This finding is potentially attributable to a higher rate of natural immunity for hepatitis B in patients with HCV: 43.46% ± 4.39% versus 4.71% ± 0.36% (non-HCV), P < 0.0001, in 1999-2004; 30.49% ± 4.87% versus 3.90% ± 0.37% (non-HCV), P < 0.0001, in 2005-2008.

Diabetic Population.

In the diabetic cohort, trends in all HepA vaccination and immunity parameters over time were similar to those for the general population and to the CLD cohort (Table 3). The actual percentages, however, which partially reflected the age disparity of the diabetic cohort, were significantly different from the controls. Specifically, the HepA vaccination rate for diabetics was consistently lower than the nondiabetic population (9.34% ± 1.05% versus 12.22% ± 0.57% in 1999-2004, P = 0.0152, and 15.35% ± 1.67% versus 21.16% ± 0.98%, P = 0.0020, in 2005-2008). On the other hand, anti-HAV seropositivity and hepatitis A QM in diabetics were significantly higher than in the controls.

Vaccination rates for hepatitis B in the diabetic cohort increased with the rest of the population, but remained consistently lower than in the nondiabetic controls. The same was true for anti-HBs seropositivity, effective HepB vaccination, and QM rates (Table 3).

Independent predictors of vaccination and QM for both hepatitis A and hepatitis B in individuals with CLD and diabetes are summarized in Supporting Table 1 for the two study cycles separately. Additionally, for patients with subtypes of CLD, independent predictors of HepA and HepB vaccination and QM are summarized in Supporting Table 2 for the two study cycles merged together.

Predictors of Ineffective HepA and HepB Vaccination.

Vaccination ineffectiveness was studied in the merged cohort from both study cycles. Vaccination against HepA (or HepB) was presumed ineffective when a reported history of vaccination was not accompanied by the respective positive serology for anti-HAV (or anti-HBs).

For both hepatitis A and hepatitis B, only approximately half of the individuals who reported a history of vaccination also had detectable levels of the respective antibodies. On the other hand, the percentage of individuals who reported incomplete vaccination series ranged from 25% to 32% for hepatitis A and 11% to 22% for hepatitis B in all studied cohorts. We used the parameter of having an incomplete vaccination series as a potential predictor of having ineffective vaccination, together with all demographic, socioeconomic, and medical parameters listed in Table 2. A summary of predictors of ineffective vaccination is given in Table 5.

Table 5. Independent Predictors of Ineffective Vaccination Against Hepatitis A and B Vaccination in All Subjects and in CLD, NAFLD, HCV+, and Diabetes* (NHANES 1999-2008)
 Hepatitis A
 U.S. PopulationCLDHCV+NAFLDDiabetes
Ineffectiveness (%)51.27 ± 1.4850.23 ± 3.3055.32 ± 7.5351.05 ± 4.1145.18 ± 4.21
Incomplete series (%)25.77 ± 1.5529.22 ± 3.1831.75 ± 7.4928.62 ± 3.5924.29 ± 3.67
Independent predictors
 Caucasian4.19 (3.30-5.31)4.34 (2.27-8.27)7.13 (0.87-58.31)6.84 (2.78-16.78)5.42 (1.84-15.99)
 African American3.76 (2.85-4.97)2.30 (0.97-5.42)1.81 (0.20-16.85)5.31 (1.77-15.93)3.99 (1.18-13.52)
 Age ≥65 years0.31 (0.21-0.44)0.40 (0.18-0.88)2.33 (0.27-20.36)0.40 (0.16-1.00)0.26 (0.10-0.64)
 Obesity1.48 (1.21-1.79)1.56 (0.87-2.79)0.12 (0.01-1.47)2.52 (1.22-5.23)1.30 (0.59-2.84)
 Incomplete series1.80 (1.42-2.27)1.75 (1.02-3.01)1.89 (0.42-8.58)1.64 (0.84-3.22)1.18 (0.51-2.76)
 Hepatitis B
 U.S. PopulationCLDHCV+NAFLDDiabetes
  • *

    Odds ratio (95% CI).

  • Different from controls.

  • Abbreviations: CLD, chronic liver disease; NAFLD, nonalcoholic fatty liver disease; HCV+, hepatitis C virus positive; Na, not applicable.

Ineffectiveness (%)50.18 ± 0.8454.10 ± 1.9640.89 ± 5.5657.08 ± 2.5169.10 ± 2.75
Incomplete series (%)11.55 ± 0.6715.65 ± 1.7121.61 ± 5.2914.43 ± 2.0611.89 ± 1.97
Independent predictors
 Caucasian0.76 (0.65- 0.89)0.78 (0.52-1.17)0.84 (0.30-2.33)0.64 (0.40-1.01)0.86 (0.47-1.58)
 Male1.48 (1.24- 1.76)1.37 (0.93-2.03)1.43 (0.45-4.54)1.05 (0.64-1.75)1.04 (0.56-1.94)
 Age ≥65 years1.61 (1.21- 2.13)2.71 (0.97-7.56)15.36 (0.47 503.64)4.88 (1.45-16.35)0.81 (0.37-1.79)
 Diabetes1.97 (1.47- 2.64)2.24 (1.08-4.64)1.90 (0.37-9.69)2.05 (0.71-5.90)Na
 Obesity1.54 (1.33- 1.78)1.85 (1.28-2.68)0.86 (0.29-2.56)1.92 (1.22-3.04)1.47 (0.75-2.89)
 History of cancer1.09 (0.77- 1.54)0.63 (0.27-1.47)0.94 (0.09-9.83)0.50 (0.18-1.40)3.67 (1.16-11.58)
 HIV0.36 (0.17- 0.79)0.41 (0.09-1.80)0.57 (0.05-5.96)0.47 (0.06-3.70)No cases
 Incomplete series2.58 (2.02-3.29)3.83 (2.17-6.73)3.78 (1.11-12.86)4.03 (1.83-8.88)2.32 (0.90-5.94)

For the entire study cohort, age under 65 years, obesity, and receiving an incomplete vaccination series were all independently associated with ineffective HepA vaccination. For the CLD cohort, incomplete vaccination series remained an independent predictor of ineffective HepA vaccination. In the diabetic cohort, only ethnicity was associated with ineffectiveness of HepA vaccination (Table 5).

A different pattern was observed for the ineffectiveness of HepB vaccination. Specifically, NAFLD and diabetic cohorts showed significantly higher rates of ineffective HepB vaccination. Furthermore, in the general population, non-Caucasian race, male gender, age of 65 years or older, and both diabetes and obesity, together with incomplete vaccination series, were all independently associated with higher rates of ineffective HepB vaccination. Similar patterns were observed in the CLD subcohorts. However, for patients with diabetes, only a history of cancer was independently associated with ineffective HepB vaccination (Table 5).


Vaccination guidelines generally recommend immunization for viral hepatitis in children and high-risk populations. Because these guidelines emerged only two decades ago, little is known about the implementation of these guidelines in the targeted high-risk population, such as individuals with CLD. In this article, we report on estimates of the nationwide prevalence of vaccination and immunity against viral hepatitis A and B as well as changes in these estimates since when the guidelines were introduced.

Consistent with the recently reported national aggregates,41 we demonstrated that over the past decade, HepA and HepB vaccination rates in the U.S. population increased by approximately 70% and 40%; respectively. During the same period, respective seroprevalence for anti-HAV decreased, reflecting a decline in the incidence of acute hepatitis A. As a result, the hepatitis A QM rate did not change over time, leaving almost 60% of adult Americans without adequate immunity against hepatitis A virus. On the other hand, the seroprevalence for anti-HBs increased as did the percentage of effective vaccinations. Nevertheless, effective HepB vaccination still does not exceed 20% of the adult U.S. population.

Factors independently associated with lack of vaccination vary with time and different study cohorts, but patients of older age, especially 65 years or older, are consistently undervaccinated. This is particularly disturbing, because acute hepatitis A or B infections can have a severe clinical course in older individuals and can be especially devastating in older patients with preexisting CLD.42 On the other hand, no other demographic or socioeconomic parameter is consistently associated with vaccination or immunity for viral hepatitis (although most show an association in certain diagnostic cohorts), confirming that all U.S. residents, regardless of gender, race, medical history, and social background, should continue being evaluated for vaccination.

Another important result of this analysis demonstrates that despite longstanding recommendations, rates of HepA and HepB vaccination and QM in patients with CLD do not differ from the general population. In fact, the only CLD subtype in which hepatitis B QM is higher than in the rest of the population is HCV+ individuals. Similar percentages were recently reported from the Veteran Affairs data.40 We believe that this is the result of high rates of natural immunity for hepatitis B in patients with HCV, rather than effective implementation of vaccination guidelines in hepatitis C infected population. In support of this hypothesis, our data show that being HCV is independently associated with higher hepatitis B QM, but not with HepB vaccination. Strikingly, our analysis of the U.S. population data shows that among individuals with diabetes, both HepA and HepB vaccination rates are significantly lower than the rest of the population and remain low over time. Given the high prevalence of underlying chronic liver disease (NAFLD) in diabetics, these patients remain vulnerable against acute hepatitis A and B infections.

Our findings also suggest that vaccine ineffectiveness, determined for the aim of the study as the absence of detectable protective antibodies in vaccinated individuals, is approximately 50% in all subcohorts. Although some patients may never develop protective antibody after vaccination (i.e., true ineffective vaccination), some individuals with a history of vaccination who do not show detectable antibody may have lost antibody titer over time. In fact, some of these patients may still be protected.43-45 However, given the limitation of the available data, we were unable to separate those who lost antibody titer over time from those individuals who were unable to develop protective antibody.43 Despite this limitation, our data show that risk factors for ineffective immunization are similar for both hepatitis A and hepatitis B. Not surprisingly, having an incomplete vaccination series was a consistent factor leading to ineffective vaccination. Additionally, we found that diabetes and older age (for hepatitis B only), together with obesity (for both hepatitis A and B), were all associated with vaccine ineffectiveness in the general population as well as in patients with CLD. Given the epidemic of obesity and diabetes, these findings, though preliminary, pose special interest and should be considered by vaccination manufacturers, healthcare providers, public health leaders, and health policy makers.

The limitations of our study include the absence of hepatitis A and hepatitis B antibody titers, which could be associated with “protective antibody.” Furthermore, as noted previously, among successfully vaccinated adults, some individuals may lose detectable antibodies within 10-20 years.43 Despite the loss of detectable antibodies, some individuals may still mount an anamnestic response after exposure to hepatitis B and remain protected.44, 45 In this study, we did not have information on how long before the survey a participant had received vaccination, which could have led to overestimating the rate of true infective vaccination. Additionally, our results may also be potentially biased toward having overestimating national vaccination rates because of the nature of NHANES data collection, which does not include incarcerated, homeless, and hospitalized people.

In conclusion, in this article, we have reported on vaccination and immunity rates for the general U.S. population and for the subpopulations at highest risk for viral hepatitis, such as individuals with CLD. We have shown that despite guidelines recommending hepatitis A and hepatitis B immunization for high-risk cohorts, vaccination rates are still very low and do not differ from the rest of the population. Furthermore, some of the high-risk factors, such as older age, obesity, and diabetes are, in fact, independently associated with no vaccination and insufficient immunity. This situation requires immediate attention, given the public health implication of acute viral hepatitis in older patients,46 those with CLD as well as patients with obesity- and diabetes-related NAFLD. Furthermore, a more uniform set of guidelines for vaccinating patients with CLD is urgently needed.