Skip to main content
NCBI home page
HHS Author Manuscripts logoLink to HHS Author Manuscripts
. Author manuscript; available in PMC: 2024 Apr 8.
Published in final edited form as: Clin Infect Dis. 2010 Jun 1;50(11):1439–1447. doi: 10.1086/652438

Cytomegalovirus Seroprevalence in the United States: The National Health and Nutrition Examination Surveys, 1988–2004

Sheri Lewis Bate 1, Sheila C Dollard 1, Michael J Cannon 1
PMCID: PMC11000537  NIHMSID: NIHMS1979843  PMID: 20426575

Abstract

Background.

Congenital cytomegalovirus (CMV) infection causes permanent disabilities in more than 5500 children each year in the United States. The likelihood of congenital infection and disability is highest for infants whose mothers were CMV seronegative before conception and who acquire infection during pregnancy.

Methods.

To provide a current, nationally representative estimate of the seroprevalence of CMV in the United States and to investigate trends in CMV infection, serum samples from the National Health and Nutrition Examination Survey (NHANES) 1999–2004 were tested for CMV-specific immunoglobulin G antibody, and results were compared with those from NHANES III (1988–1994). Individuals aged 6–49 years (21,639 for NHANES III and 15,310 for NHANES 1999–2004) were included.

Results.

For NHANES 1999–2004, the overall age-adjusted CMV seroprevalence was 50.4%. CMV seroprevalence was higher among non-Hispanic black and Mexican American children compared with non-Hispanic white children and increased more quickly in subsequent age groups. CMV seropositivity was independently associated with older age, female sex, foreign birthplace, low household income, high household crowding, and low household education. Compared with NHANES 1988–1994, the overall age-adjusted CMV seroprevalence for NHANES 1999–2004 was not significantly different.

Conclusions.

Many women of reproductive age in the United States are still at risk of primary CMV infection during pregnancy. There is an urgent need for vaccine development and other interventions to prevent and treat congenital CMV. The substantial disparities in CMV risk among seronegative women suggest that prevention strategies should include an emphasis on reaching racial or ethnic minorities and women of low socioeconomic status.

Cytomegalovirus (CMV), a member of the Herpesviridae family, is endemic throughout the world [1]. Most CMV infections are mild or asymptomatic; however, CMV can cause serious disease in immunocompromised individuals and fetuses. Among newborns, CMV is the leading cause of congenital infection in the developed world [2]. Each year, of an estimated 28,000 children born with congenital CMV infection in the United States, ~150 die, and >5500 have permanent disabilities, such as hearing loss, intellectual disability, psychomotor delay, speech and language disabilities, behavioral disorders, visual impairment, and cerebral palsy [3, 4].

CMV is acquired through contact with CMV-infected body fluids of individuals with symptomatic or asymptomatic CMV infection [1, 5, 6]. Congenital CMV infection is the result of intrauterine transmission of CMV infection from mother to fetus. A fetus is at highest risk of CMV infection when a mother has a primary (ie, first) infection during pregnancy [7, 8]. Compared with a maternal nonprimary infection (ie, reinfection or reactivation), a maternal primary infection is more likely to transmit CMV from mother to fetus (1% vs 32%) and is also more likely to result in severe, long-term sequelae in children born with congenital CMV infection [6, 8, 9]. Women who are CMV seropositive before pregnancy are 69% less likely to give birth to a CMV-infected newborn [10].

The best way to assess the prevalence of CMV infection is through seroprevalence studies of CMV-specific immunoglobulin (Ig) G antibody. A previous nationally representative seroprevalence study of CMV infection in the United States, which used data from the National Health and Nutrition Examination Survey (NHANES) III (1988–1994), indicated that large proportions of women of reproductive age are susceptible to primary infection during pregnancy [11].

The purpose of this study was to provide current estimates of CMV seroprevalence in the United States and to investigate trends in CMV infection by comparing seroprevalence data from NHANES III (1988–1994) with data from NHANES 1999–2004. National seroprevalences, particularly among women of reproductive age, are important for establishing accurate estimates of the risk of congenital CMV infection, for quantifying the potential target population for a CMV vaccine, and for identifying risk groups that should be a high priority to receive behavioral interventions and/or vaccine once one becomes available [12, 13]. National trends in CMV infection have not been examined previously and can provide insight into the rate of change in CMV seroprevalence during the past decade as a function of socioeconomic and demographic factors.

METHODS

Survey sample and design.

NHANES is a series of cross-sectional surveys conducted by the National Center for Health Statistics (NCHS) of the Centers for Disease Control and Prevention (CDC) [14]. Data are collected through household interviews, physical examinations, and blood sampling [14]. To select a nationally representative sample, NHANES uses a complex multistage probability cluster sample design, discussed in greater detail elsewhere [15]. This study was approved by the institutional review board of the CDC, and participants provided written informed consent.

NHANES III was conducted from 1988 to 1994. Since 1999, NHANES has been administered as a continuous annual survey released in 2-year increments. The NHANES 1999–2000, 2001–2002, and 2003–2004 data sets were combined to form NHANES 1999–2004. NHANES III and NHANES 1999–2004 oversampled certain population groups to increase the reliability and precision of estimates for each of the groups [14, 16].

Individuals included in this study were those aged 6–49 years who were interviewed and examined, consented to be a part of additional testing, had serum samples available for testing, and had a nonequivocal CMV test result. The final study sample included 21,639 individuals for NHANES III and 15,310 individuals for NHANES 1999–2004. Serum sample availability was fairly uniform across variable categories except for the 6- to 11-year-old age group, in which greater percentages of persons were without available serum specimens. To address the impact of missing CMV data on the generalizability of the study results, we created adjusted weights by multiplying the original NCHS weights by the weighted proportion of available serum samples for that participant’s sex, age, and race/ethnicity [11]. Seroprevalence analyses using the adjusted weights produced only slightly different point estimates that were within 95% confidence intervals (CIs) of estimates computed with the original NCHS weights; therefore, NCHS weights were used for all subsequent analyses.

Serologic testing.

Laboratory methods for detecting CMV IgG antibody in serum from NHANES 1999–2004 followed the same procedures used previously for testing NHANES III samples [11]. To maximize testing sensitivity, specificity, and through-put, serum samples were screened for CMV-specific IgG antibody with the SeraQuest enzyme immunoassay (Quest International) and the Triturus enzyme-linked immunosorbent assay robot (Grifols USA). Then serum samples with reactivities within a narrow range of the SeraQuest assay cutoff were tested using the VIDAS test (bioMérieux Vitek). Discrepant results between the SeraQuest and VIDAS tests were resolved with an immunofluorescence assay (Bion Enterprises). All testing was conducted by laboratory personnel at the CDC.

Measures.

Variables of interest included sex, age (6–11, 12–19, 20–29, 30–39, and 40–49 years), race or ethnicity (non-Hispanic white, non-Hispanic black, Mexican American), birthplace (born in the 50 states, born in Mexico, born elsewhere), household income level (ratio of household income to the family’s appropriate poverty threshold: low [⩽1.300], middle [1.301–3.500], or high [⩾3.501]), insurance status (covered or not covered by health insurance), household education level (less than high school, high school diploma including Graduate Educational Development diploma, or more than high school), and household crowding index (low [<0.5 person per room], average [0.5–1 person per room], or high [>1 person per room]). Individuals not fitting into 1 of the 3 race/ethnicity groups were classified as “other” in 1999–2004 univariate results but were excluded from 1999–2004 multivariate analyses and comparisons between results from 1988–1994 and 1999–2004. Household income level was based on the poverty income ratio variable, which is the ratio of household income to the family’s appropriate poverty threshold [17]. For the 1999–2004 analyses, household education level represented the highest education level of the household reference person or the reference person’s spouse, if applicable. For comparison with 1988–1994 data, only the reference person’s education level was used to be consistent with the NHANES III household education variable definition.

Statistical analysis.

SUDAAN software, version 10.0 (RTI International), was used for statistical analyses. All prevalence estimates were weighted to represent the civilian, noninstitutionalized US population and to account for the unequal probability of sampling and nonresponse to the household interview and physical examination. To reduce potential confounding by age, age-adjusted estimates were computed using the direct method to the 2000 US Census population [18].

For analyses of the 1999–2004 data, demographic factors were first evaluated with adjustment for age using a general linear contrast procedure. Next, logistic regression models were used to assess the association between CMV seropositivity and the demographic factors while adjusting for multiple covariates. The final logistic model included age, sex, race/ethnicity, household income level, birthplace (country of origin), household education, crowding index, and an age by sex interaction, age by race/ethnicity interaction, and race/ethnicity by sex interaction. The Satterthwaite adjusted F test was used to assess the statistical significance of variables and interactions in the model. Model fit was evaluated using the Hosmer-Lemeshow goodness-of-fit Satterthwaite-adjusted F test. Individuals with missing data on the variables in the multivariate models were excluded.

Because of interactions, logistic models were created for each age, sex, and racial/ethnic subgroup, as well as for combined subgroups for sex and race/ethnicity (eg, non-Hispanic white females) and sex and age (eg, 6- to 11-year-old girls). To aid the interpretation of odds ratios (ORs), predictive margins were computed using the PREDMARG statement in SUDAAN. Predictive margins are akin to adjusted seroprevalences; the predictive margin for a group represents the average predicted response if all individuals in the sample had been in that group, while controlling for all other covariates [19, 20]. The 95% CIs of predictive margins were based on the actual degrees of freedom for each level of each variable.

Methods to compare the 1988–1994 and 1999–2004 data were similar to those used for analyses of the 1999–2004 data, except that a variable representing survey year (1 for NHANES 1998–1994 and 2 for NHANES 1999–2004) was forced into the model. The final logistic model included age, sex, race/ethnicity, household income level, birthplace, household education, crowding index, and an age by sex, age by race/ethnicity, age by household income level, race/ethnicity by sex, and age by race/ethnicity by sex interaction. Because of numerous interactions with age, logistic models were performed for each age group. Additional stratifications were performed because of a significant 3-way interaction among age, race/ethnicity, and sex.

RESULTS

NHANES 1999–2004.

The overall age-adjusted seroprevalence of CMV infection for individuals 6–49 years old was 50.4% (Table 1). In age-adjusted analyses, CMV seropositivity was significantly associated with female sex, non-Hispanic black race and Mexican American ethnicity, older age, foreign birthplace, low household income level, lack of insurance, low household education, and high crowding index (Table 1).

Table 1.

Differences in Age-Adjusted Cytomegalovirus (CMV) Seroprevalence in Individuals Aged 6–49 Years, by Selected Demographic Factors between National Health and Nutrition Examination Survey (NHANES) 1988–1994 and 1999–2004

Demographic factor NHANES 1988–1994 NHANES 1999–2004 Difference, % (95% CI)
Sample size CMV seroprevalence, % (95% CI) Sample size CMV seroprevalence, % (95% CI)
Total 14,538 50.8 (48.7–52.9) 15,310 50.4 (48.0–52.7) −0.4 (−3.6 to 2.8)
Sex
 Female (reference) 7695 56.1 (53.5–58.7) 7882 55.5 (53.3–57.7) −0.6 (−4.0 to 2.8)
 Male 6843 45.5a (43.1–47.8) 7428 45.2a (42.4–48.0) −0.3 (−3.9 to 3.3)
Race/ethnicity
 Non-Hispanic white (reference) 4209 41.7 (39.3–44.2) 5284 39.5 (36.9–42.2) −2.2 (−5.8 to 1.4)
 Non-Hispanic black 4759 70.9a (69.6–72.1) 4227 70.6a (68.5–72.8) −0.2 (−2.6 to 2.2)
 Mexican American 4921 77.6a (75.8–79.4) 4679 76.9a (74.1–79.6) −0.7 (−3.9 to 2.5)
Age, years
 6–11 (reference) 2679 36.3 (32.7–39.9) 2384 37.5 (34.2–40.8) 1.2 (−3.6 to 6)
 12–19 2918 41.7a (38.2–45.2) 6066 42.7a (39.6–45.9) 1.1 (−3.5 to 5.7)
 20–29 3302 49.0a (45.5–52.5) 2391 49.5a (46.1–52.8) 0.5 (−4.3 to 5.3)
 30–39 3156 54.0a (50.2–57.9) 2251 56.7a (53.2–60.1) 2.6 (−2.6 to 7.8)
 40–49 2483 64.3a (60.4–68.1) 2218 58.0a (54.8–61.1) −6.3 (−11.3 to −1.3)b
Birthplace
 United States (reference) 11,573 46.1 (44.0–48.3) 12,387 45.1 (42.7–47.6) −1 (−4.2 to 2.2)
 Mexico 2054 90.7a (88.4–93.0) 1880 89.4a (87.6–91.2) −1.3 (−4.1 to 1.5)
 Other 869 78.8a (74.3–83.4) 1042 76.0a (71.4–80.6) −2.9 (−9.3 to 3.5)
Household income levelc
 Low (≤1.300) 5388 66.3a (63.2–69.5) 5380 66.0a (62.4–69.6) −0.3 (−5.1 to 4.5)
 Middle (1.301–3.500) 5676 52.2a (48.5–55.9) 5209 50.9a (47.9–53.8) −1.3 (−5.9 to 3.3)
 High (≥3.501; reference) 2282 37.7 (34.4–41.1) 3624 38.9 (36.5–41.4) 1.2 (−3.0 to 5.4)
Insurance
 Insured (reference) 10,493 48.4 (46.0–50.8) 11,423 47.0 (44.6–49.4) −1.4 (−4.8 to 2.0)
 Not insured 3215 60.9a (57.1–64.8) 3694 62.8a (58.7–66.9) 1.9 (−3.7 to 7.5)
Household education level
 Less than high school 5557 67.6a (64.4–70.8) 4909 69.2a (66.0–72.4) 1.6 (−2.8 to 6.0)
 High school graduate or GED diploma 4496 52.0a (48.6–55.3) 3620 51.2a (47.7–54.8) −0.8 (−5.6 to 4.0)
 More than high school (reference) 4381 42.5 (39.6–45.4) 6216 42.8 (40.0–45.6) 0.3 (−3.7 to 4.3)
Crowding index
 Low (<0.5 person per room) (reference) 3684 42.2 (39.8–44.6) 4405 41.4 (38.9–44.0) −0.8 (−4.4 to 2.8)
 Average (0.5–1 person per room) 7617 53.1a (50.3–56.0) 8047 54.2a (51.5–56.9) 1.0 (−2.8 to 4.8)
 High (>1 person per room) 3195 73.8a (70.2–77.3) 2673 75.8a (72.1–79.6) 2.1 (−2.9 to 7.1)
Open in a new tab

NOTE. Ages were adjusted to the year 2000 US Census Bureau by the direct method to the age groups 6–11, 12–19, 20–29, 30–39, and 40–49 years. CI, confidence interval; GED, General Educational Development.

a

P≤.05 (associated with the Student t test evaluating pairwise comparisons).

b

P<.05 (associated with the Student t test comparing pairwise differences between prevalence percentages).

c

Household income level was based on the poverty income ratio variable, which is the ratio of household income to the family’s appropriate poverty threshold.

For both females and males, multivariate-adjusted CMV seroprevalences were higher for non-Hispanic blacks and Mexican Americans compared with non-Hispanic whites (Table 2 and Figure 1). Overall CMV seroprevalences were higher for non-Hispanic black and Mexican American 6–11-year-olds (45.7% and 61.7%, respectively) compared with non-Hispanic white children of the same age group (29.0%). Compared with the 6–11-year age group, CMV seroprevalences for the 12–19-year and 20–29-year age groups were significantly higher for non-Hispanic blacks (OR, 1.6 [95% CI, 1.2–2.1] and 4.1 [95% CI, 2.7–6.2], respectively) and Mexican Americans (OR, 1.6 [95% CI, 1.2–2.2] and 3.2 [95% CI, 2.1–4.9], respectively), whereas CMV prevalences were not significantly higher by age group among non-Hispanic whites until the 30–39-year age group (OR, 2.7; 95% CI, 1.9–3.7). Among non-Hispanic blacks, a strong association with CMV seropositivity (OR, 21.3; 95% CI, 11.7–38.9) occurred in the 30–39-year age category (compared with 6–11-year-olds). Examination of estimates by sex and race indicated that this was due to a strong association for women (OR, 26.0; 95% CI, 13.8–49.2) rather than men (OR, 3.6; 95% CI, 2.2–6.1) (Figure 2); CMV seroprevalences were 76.6% and 94.6% for non-Hispanic black women aged 20–29 years and 30–39 years, respectively, whereas seroprevalences were 62.4% and 74.5% for non-Hispanic black men of the corresponding age groups (Figure 1).

Table 2.

Factors Associated with Cytomegalovirus (CMV) Seroprevalence in the United States: Predictive Margins (Multivariate Adjusted CMV Seroprevalences), by Sex and Race or Ethnicity, National Health and Nutrition Examination Survey 1999–2004

Characteristic Non-Hispanic white females Non-Hispanic black females Mexican American females Non-Hispanic white males Non-Hispanic black males Mexican American males
Sample size Predictive margin (95% Cl) Sample size Predictive margin (95% Cl) Sample size Predictive margin (95% Cl) Sample size Predictive margin (95% Cl) Sample size Predictive margin (95% Cl) Sample size Predictive margin (95% Cl)
Subpopulation total 2760 45.2 (42.9–47.6) 2125 77.2 (74.7–79.6) 2394 78.2 (74.6–81.8) 2522 35.1 (31.9–38.3) 2102 61.9 (58.7–65.2) 2272 73.1 (70.2–76.0)
Age, years
 6–11 302 29.6 (24.2–34.9) 394 45.6 (40.5–50.7) 401 61.3 (53.9–68.7) 319 28.3 (21.3–35.3) 405 46.4 (40.5–52.3) 415 62.3 (55.9–68.7)
 12–19 735 33.3 (29.5–33.1) 943 56.5 (51.6–61.3) 1064 70.4 (64.4–76.3) 792 29.6 (24.4–34.3) 1020 52.8 (48.2–57.3) 1053 65.8 (60.7–70.8)
 20–29 599 35.7 (31.4–40.0) 261 76.6 (70.0–83.2) 378 81.3 (75.4–87.2) 438 32.6 (26.6–38.7) 214 62.4 (54.9–69.9) 287 72.9 (67.1–78.6)
 30–39 602 52.1 (47.3–57.0) 256 94.6 (91.3–97.4) 266 37.3 (32.4–93.2) 464 36.2 (30.7–41.7) 206 74.5 (66.9–82.1) 236 78.6 (70.6–86.5)
 40–49 472 57.5 (52.2–62.9) 271 90.3 (86.9–93.6) 285 90.1 (84.4–95.8) 509 42.0 (36.9–47.1) 257 70.2 (61.7–78.7) 281 86.2 (82.7–89.7)
Birthplace
 United States 2635 44.2 (41.7–46.7) 1991 76.5 (74.0–79.0) 1482 72.8 (67.8–77.8) 2397 33.8 (30.5–37.2) 1938 59.3 (55.9–62.7) 1310 62.8 (58.5–67.1)
 Mexico …a … 0 … 912 39.7 (36.1–93.3) …a … 0 … 962 85.2 (82.4–88.0)
 Other 125 68.4 (54.9–81.9) 133 87.5 (79.0–96.0) …b … 125 60.1 (48.4–71.9) 164 90.5 (82.8–98.2) …b …
Household income levelc
 Low (≤1.300) 613 51.6 (45.6–57.6) 944 81.3 (77.9–84.7) 1086 81.2 (76.4–86.1) 495 45.4 (39.6–51.2) 846 63.7 (57.8–69.6) 953 77.4 (73.2–81.7)
 Middle (1.301–3.500) 335 44.2 (39.4–49.0) 693 76.5 (72.0–31.0) 335 77.7 (73.4–32.1) 307 35.0 (30.3–39.7) 716 62.0 (57.5–66.4) 887 73.0 (69.4–76.6)
 High (≥3.501) 1129 43.1 (39.5–46.7) 309 67.4 (58.4–76.5) 288 72.5 (64.4–80.5) 1082 31.6 (28.1–35.1) 356 58.9 (51.9–65.9) 235 62.9 (54.4–71.5)
Insurance
 Insured 2382 44.2 (41.9–46.5) 1765 76.6 (74.0–79.2) 1474 78.5 (75.0–82.1) 2059 34.2 (31.1–37.4) 1624 59.6 (55.8–63.4) 1277 73.2 (70.2–76.1)
 Not insured 361 51.4 (43.7–59.1) 339 30.2 (75.1–35.3) 396 77.5 (71.4–33.6) 427 39.3 (31.0–47.5) 433 68.8 (64.6–73.1) 965 73.1 (68.1–78.0)
Household education level
 Less than high school 263 57.3 (52.2–63.5) 645 31.5 (76.3–36.3) 1157 33.4 (79.7–37.1) 236 36.6 (29.5–43.7) 608 66.3 (60.7–71.8) 1146 75.8 (72.4–79.3)
 High school graduate or GED diploma 612 50.0 (45.4–54.5) 553 80.8 (76.7–84.8) 520 76.7 (72.3–81.2) 561 35.2 (30.4–40.1) 506 64.2 (58.8–69.6) 484 74.3 (69.8–78.8)
 More than high school 1337 42.3 (39.4–45.2) 336 72.7 (69.6–75.9) 641 73.9 (67.6–30.1) 1691 34.9 (31.6–38.3) 916 58.6 (54.7–62.5) 569 69.2 (63.2–75.3)
Crowding index
 Low 1313 42.1 (39.1–45.2) 564 74.7 (70.4–73.9) 252 69.4 (61.2–77.5) 1196 34.2 (30.5–37.9) 554 60.6 (54.1–67.1) 257 69.7 (61.3–78.1)
 Average 1337 48.3 (45.0–51.6) 1238 78.0 (75.4–80.6) 1303 78.6 (75.0–82.2) 1196 36.1 (32.2–39.9) 1187 61.5 (57.2–65.9) 1132 72.8 (69.2–76.4)
 High 90 53.0 (40.3–65.2) 301 30.6 (75.3–35.9) 317 33.6 (79.0–33.3) 96 38.1 (28.6–47.6) 320 67.4 (61.3–73.6) 860 76.1 (71.8–80.5)
Open in a new tab

NOTE. The models for sex by race/ethnicity were adjusted for age, birthplace, household income level, insurance, household education, and crowding index. CI, confidence interval; GED, General Educational Development.

a

The logistic models for non-Hispanic whites excluded 1 female born in Mexico and 1 male born in Mexico.

b

The logistic model for Mexican Americans excluded 6 females born in “other” country and 7 males born in “other” country.

c

Household income level was based on the poverty income ratio variable, which is the ratio of household income to the family’s appropriate poverty threshold.

Figure 1.

Figure 1.

Open in a new tab

Predictive margins (multivariate adjusted cytomegalovirus [CMV] seroprevalences) in the United States, National Health and Nutrition Examination Survey (NHANES) 1999–2004, stratified by age, sex, and race/ethnicity. The circles representing the female and male prevalences are distinguished by the female (♀) and male (♂) icons. To better distinguish between females and males, the circles representing the prevalences in the 6–11-year-old age groups are plotted slightly above and below their true values; true prevalences are shown in the text next to the circles.

Figure 2.

Figure 2.

Open in a new tab

Factors associated with cytomegalovirus (CMV) seroprevalence in the United States: adjusted odds ratios and 95% confidence intervals (error bars) by sex and race/ethnicity, Health and Nutrition Examination Survey (NHANES) 1999–2004. Upper confidence limit for non-Hispanic (NH) black women aged 30–39 years was 49.2 but was truncated because of space constraints.

Household education level was significantly associated with CMV seropositivity for females of all 3 race/ethnicities and non-Hispanic black males (Figure 2). Those with an education level of less than high school had higher CMV prevalences than those with an education level of more than high school (57.8% vs 42.3% for non-Hispanic white females, 81.5% vs 72.7% for non-Hispanic black females, 83.4% vs 73.9% for Mexican American females, and 66.3% vs 58.6% for non-Hispanic black males). Household crowding was significantly associated with CMV seropositivity only for Mexican American females (69.4% for the low crowding group, 78.6% for the average crowding group, and 83.6% for the high crowding group) and for 1 of the group comparisons for non-Hispanic white females (48.3% for the average crowding group and 42.1% for the low crowding group). Nevertheless, for all sex and racial or ethnic groups, the ORs were greater than 1 and increased as the level of crowding increased (Figure 2). Low household income level was significantly associated with CMV seropositivity for all race and sex groups except non-Hispanic black males (Figure 2). Conversely, non-Hispanic black males were the only race-sex group for which not having insurance was a significant risk factor for CMV seropositivity (Figure 2). Mexican Americans born in Mexico had higher CMV prevalences compared with those born in the United States (OR for females, 3.8; 95% CI, 2.3–6.2; OR for males, 3.7; 95% CI, 2.7–5.0). For non-Hispanic whites and non-Hispanic blacks, sample sizes for the “other” birthplace category were too small to make any inferences about an association with CMV seropositivity.

Comparison of 1988–1994 and 1999–2004 data.

The overall age-adjusted seroprevalence of CMV did not change significantly from 1988–1994 to 1999–2004 (50.8% and 50.4%, respectively) (Table 1). After stratifying by sex and race/ethnicity, there were a handful of statistically significant changes in CMV seroprevalence among certain levels of age and household income, but no consistent patterns were observed. For non-Hispanic whites, no significant differences were observed for females; for males, the multivariate-adjusted CMV seroprevalences among 40–49-year-olds decreased between 1988–1994 and 1999–2004 (from 49.9% to 40.3%). For non-Hispanic black females, statistically significant increases occurred among 6–11-year-olds (from 38.6% to 48.7%) and 30–39-year-olds (from 81.8% to 90.3%), and a decrease occurred among 40–49-year-olds (from 94.4% to 88.5%) (Figure 3). For non-Hispanic black males, an increase in multivariate-adjusted CMV seroprevalence was observed among 12–19-year-olds (from 43.4% to 52.7%); also, as with non-Hispanic black females, a decrease occurred at the 40–49-year age level (from 86.3% to 69.8%) (Figure 3). For Mexican Americans, the only significant difference between 1988–1994 and 1999–2004 was an increase in age-adjusted CMV seroprevalence among males and females within the middle household income level (from 70.9% to 75.2%).

Figure 3.

Figure 3.

Open in a new tab

Differences in predictive margins (multivariate adjusted cytomegalovirus [CMV] seroprevalences) in the United States between Health and Nutrition Examination Survey (NHANES) 1988–1994 (dashed lines) and NHANES 1999–2004 (solid lines), stratified by age, sex, and race/ethnicity. GED, Graduate Educational Development diploma; NH, non-Hispanic; Ref, reference.

DISCUSSION

CMV seroprevalence across most age, sex, and racial/ethnic groups in the United States showed few changes between 1988–1994 and 1999–2004. Given this relative lack of temporal trends, differences in CMV seroprevalence by age approximate seroconversion rates among persons moving from one age group to the next. Thus, we can conclude that among the substantial proportion of US women who are CMV seronegative as they enter their reproductive years, many experience seroconversion. For example, of the ~45% of non-Hispanic black women who are CMV seronegative during their teen years, nearly all (~8 of 9) seroconvert by the time they are in their 30s. Approximately one-half of CMV-seronegative Mexican American women and one-fourth of CMV-seronegative non-Hispanic white women also seroconvert during the same period. This means that many women are at risk of experiencing a primary CMV infection during pregnancy, which is associated with a higher risk of congenital infection [8] and permanent damage to the child [21].

Such risk highlights the urgent need for interventions, such as a vaccine, which can reduce the likelihood of adverse outcomes, such as maternal infection, congenital infection, or childhood disability. Although no licensed vaccines are currently available, a number are in various stages of development, including a gB subunit vaccine that showed efficacy in a recent phase 2 trial [22]. Several target populations have been proposed for future vaccines, including women of reproductive age [22], preadolescents [23], and children [24]. Our study shows that all of these groups contain large proportions of CMV-seronegative individuals who could be protected from CMV infection by vaccination.

Sex, race or ethnicity, country of origin, and factors associated with low socioeconomic status, such as crowding and low household income, were all independently associated with and substantially affected CMV seropositivity. These multiple associations are not surprising, given the multiple ways that CMV can be transmitted. The main transmission routes for CMV infection are breast-feeding [5, 25], close contact with young children [26–31], and intimate contact with adults (ie, kissing or sexual intercourse) [27, 32, 33]. The chance of becoming infected depends primarily on 2 factors: the frequency of these contacts and the likelihood that any given contact will be with a person who is shedding CMV in their bodily fluids. These factors differ for each of the main transmission routes and change during a lifetime, making it difficult to precisely explain what drives CMV seroprevalence results and what accounts for racial or ethnic differences. For instance, possible explanations include breast-feeding rates, household demographic factors and child care arrangements, and sexual behaviors and networks, all of which differ substantially by race or ethnicity [34–39]. However, there is no clear correlation between these racial/ethnic variations in exposure to prominent CMV transmission modes and the likelihood of being CMV seropositive. Thus, although this study is useful for identifying at-risk populations, it is less able to assess the relative importance of different modes of CMV transmission.

The major strengths of this study are that it was nationally representative, reported a current estimate of CMV seroprevalence in the United States, and assessed time trends in the US population as a whole. The major limitation was its cross-sectional design (both NHANES III and NHANES 1999–2004 were cross-sectional samples), which does not allow for the definitive detection of a birth cohort effect. However, comparison of data between 1988–1994 and 1999–2004 indicated that with the exception of some minor changes in CMV seroprevalences in the oldest age group, CMV seroprevalences were relatively constant between individuals of the same age who were born ~10 years apart, suggesting that any birth cohort effect was minimal and that incidence has not changed substantially in the recent past. Also, a cross-sectional design does not reveal when seropositive individuals became infected with CMV. Furthermore, the time-dependent variables, such as crowding or household income level, were measured at only one point in time (ie, the time of the NHANES survey). Thus, it was impossible to determine whether the exposures of interest occurred at a time that was relevant to the CMV infection.

In summary, many women of reproductive age in the United States are still at risk of primary CMV infection during pregnancy. As a result, there is an urgent need for vaccine development and other clinical and public health interventions that can benefit children and their families. The substantial disparities in CMV risk among seronegative women suggest that prevention strategies should include an emphasis on reaching racial or ethnic minorities and women of low socioeconomic status.

Acknowledgments

We thank Kay Radford, Brandon Radmall, and Minal Amin for excellent technical assistance with serologic testing. We also thank Dr Donna Brogan, Deanna Kruszon-Moran, Dr. Geraldine McQuillan, Dr Stephanie Staras, and Erica Din for their assistance.

Financial support.

This work was supported through a grant from GlaxoSmithKline and the CDC Foundation and in part by an appointment (S.L.B.) to the research participation program at the CDC, National Center for Immunization and Respiratory Diseases, Division of Viral Diseases administered by the Oak Ridge Institute for Science and Education through an interagency agreement between the US Department of Energy and CDC.

Footnotes

Potential conflicts of interest. All authors: no conflicts.

The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention

Presented in part: 2008 Congenital Cytomegalovirus Conference, Atlanta, Georgia, 5 November 2008.

References

  • 1.Mocarski ES Jr, Shenk T, Pass RF. Cytomegaloviruses. In: Knipe DM, Howley PM, eds. Fields’ virology. 5th ed. Philadelphia: Lippincott Williams & Wilkins, 2007:2702–2772. [Google Scholar]
  • 2.Cannon MJ, Davis KF. Washing our hands of the congenital cytomegalovirus disease epidemic. BMC Public Health 2005;5:70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Dollard SC, Grosse SD, Ross DS. New estimates of the prevalence of neurological and sensory sequelae and mortality associated with congenital cytomegalovirus infection. Rev Med Virol 2007;17:355–363. [DOI] [PubMed] [Google Scholar]
  • 4.Grosse SD, Ross DS, Dollard SC. Congenital cytomegalovirus (CMV) infection as a cause of permanent bilateral hearing loss: a quantitative assessment. J Clin Virol 2008;41:57–62. [DOI] [PubMed] [Google Scholar]
  • 5.Hamprecht K, Maschmann J, Vochem M, Dietz K, Speer CP, Jahn G. Epidemiology of transmission of cytomegalovirus from mother to preterm infant by breastfeeding. Lancet 2001;357:513–518. [DOI] [PubMed] [Google Scholar]
  • 6.Stagno S, Remington JS, Klein JO. Cytomegalovirus. In: Remington JS, Klein JO, eds. Infectious diseases of the fetus and newborn infant. Philadelphia: WB Saunders, 2001:389–424. [Google Scholar]
  • 7.Fowler KB, Stagno S, Pass RF. Interval between births and risk of congenital cytomegalovirus infection. Clin Infect Dis 2004;38:1035–1037. [DOI] [PubMed] [Google Scholar]
  • 8.Kenneson A, Cannon MJ. Review and meta-analysis of the epidemiology of congenital cytomegalovirus (CMV) infection. Rev Med Virol 2007;17:253–276. [DOI] [PubMed] [Google Scholar]
  • 9.Revello MG, Gerna G. Diagnosis and management of human cytomegalovirus infection in the mother, fetus, and newborn infant. Clin Microbiol Rev 2002;15:680–715. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Fowler KB, Stagno S, Pass RF. Maternal immunity and prevention of congenital cytomegalovirus infection. JAMA 2003;289:1008–1111. [DOI] [PubMed] [Google Scholar]
  • 11.Staras SAS, Dollard SC, Radford KW, Flanders WD, Pass RF, Cannon MJ. Seroprevalence of cytomegalovirus infection in the United States, 1988–1994. Clin Infect Dis 2006;43:1143–1151. [DOI] [PubMed] [Google Scholar]
  • 12.Arvin AM, Fast P, Myers M, Plotkin S, Rabinovich R. Vaccine development to prevent cytomegalovirus disease: report from the National Vaccine Advisory Committee. Clin Infect Dis 2004;39:233–239. [DOI] [PubMed] [Google Scholar]
  • 13.Bate SL, Cannon MJ. A social marketing approach to building a behavioral intervention for congenital cytomegalovirus. Health Promot Pract 2009. [EPub ahead of print]. [DOI] [PubMed] [Google Scholar]
  • 14.Centers for Disease Control and Prevention. Analytic and Reporting Guidelines: The Third National Health and Nutrition Examination Survey, NHANES III (1988–1994). Hyattsville, MD: National Center for Health Statistics, 1996. [Google Scholar]
  • 15.National Center for Health Statistics. Plan and operation of the Third National Health and Nutrition Examination Survey, 1988–94. Series 1: programs and collection procedures. Vital Health Stat 1 1994;32:1–407. [PubMed] [Google Scholar]
  • 16.Centers for Disease Control and Prevention (CDC), National Center for Health Statistics (NCHS). Analytic and reporting guidelines: the National Health and Nutrition Examination Survey (NHANES). Hyattsville, MD: US Dept of Health and Human Services, Centers for Disease Control and Prevention, 2006. http://www.cdc.gov/nchs/data/nhanes/nhanes_03_04/nhanes_analytic_guidelines_dec_2005.pdf. Accessed 12 April 2010. [Google Scholar]
  • 17.Centers for Disease Control and Prevention (CDC), National Center for Health Statistics (NCHS). National Health and Nutrition Examination Survey: Documention, Codebook, and Frequencies. NHANES 2001–2002 Data Documentation 2009. http://www.cdc.gov/nchs/nhanes/vardemo_b.htm. Accessed 12 April 2010.
  • 18.Klein RJ, Schoenborn CA. Age adjustment using the 2000 projected U.S. population. Healthy People 2010 Stat Notes 2001;20:1–10. [PubMed] [Google Scholar]
  • 19.Graubard BI, Korn EL. Predictive margins with survey data. Biometrics 1999;55:652–659. [DOI] [PubMed] [Google Scholar]
  • 20.Korn EL, Graubard BI. Analysis of health surveys. New York: John Wiley & Sons, 1999. [Google Scholar]
  • 21.Fowler KB, Stagno S, Pass RF, Britt WJ, Boll TJ, Alford CA. The outcome of congenital cytomegalovirus infection in relation to maternal antibody status. N Engl J Med 1992;326:663–667. [DOI] [PubMed] [Google Scholar]
  • 22.Pass RF, Zhang C, Evans A, et al. Vaccine prevention of maternal cytomegalovirus infection. N Engl J Med 2009;360:1191–1199. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Institute of Medicine Committee to Study Priorities for Vaccine D. Vaccines for the 21st century: a tool for decision making. Washington, DC: National Academy Press, 2000. [Google Scholar]
  • 24.Griffiths PD, McLean A, Emery VC. Encouraging prospects for immunisation against primary cytomegalovirus infection. Vaccine 2001;19:1356–1362. [DOI] [PubMed] [Google Scholar]
  • 25.Schleiss MR. Role of breast milk in acquisition of cytomegalovirus infection: recent advances. Curr Opin Pediatr 2006;18:48–52. [DOI] [PubMed] [Google Scholar]
  • 26.Adler SP. Molecular epidemiology of cytomegalovirus: evidence for viral transmission to parents from children infected at a day care center. Pediatr Infect Dis 1986;5:315–318. [PubMed] [Google Scholar]
  • 27.Fowler KB, Pass RF. Risk factors for congenital cytomegalovirus infection in the offspring of young women: exposure to young children and recent onset of sexual activity. Pediatrics 2006;118:e286–e292. [DOI] [PubMed] [Google Scholar]
  • 28.Murph JR, Baron JC, Brown CK, Ebelback CL, Bale JF. The occupational risk of cytomegalovirus infection among day-care providers. JAMA 1991;265:603–608. [PubMed] [Google Scholar]
  • 29.Pass RF, Hutto C, Ricks R, Cloud GA. Increased rate of cytomegalovirus infection among parents of children attending day-care centers. N Engl J Med 1986;314:1414–1418. [DOI] [PubMed] [Google Scholar]
  • 30.Staras SA, Flanders WD, Dollard SC, Pass RF, McGowan JE Jr, Cannon MJ. Cytomegalovirus seroprevalence and childhood sources of infection: a population-based study among pre-adolescents in the United States. J Clin Virol 2008;43:266–271. [DOI] [PubMed] [Google Scholar]
  • 31.Taber LH, Frank AL, Yow MD, Bagley A. Acquisition of cytomegaloviral infections in families with young children: a serological study. J Infect Dis 1985;151:948–952. [DOI] [PubMed] [Google Scholar]
  • 32.Staras SA, Flanders WD, Dollard SC, Pass RF, McGowan JE Jr, Cannon MJ. Influence of sexual activity on cytomegalovirus seroprevalence in the United States, 1988–1994. Sex Transm Dis 2008;35:472–479. [DOI] [PubMed] [Google Scholar]
  • 33.Stover CT, Smith DK, Schmid DS, et al. Prevalence of and risk factors for viral infections among human immunodeficiency virus (HIV)-infected and high-risk HIV-uninfected women. J Infect Dis 2003;187:1388–1396. [DOI] [PubMed] [Google Scholar]
  • 34.Auerswald CL, Muth SQ, Brown B, Padian N, Ellen J. Does partner selection contribute to sex differences in sexually transmitted infection rates among African American adolescents in San Francisco? Sex Transm Dis 2006;33:480–484. [DOI] [PubMed] [Google Scholar]
  • 35.Federal Interagency Forum on Child and Family Statistics. America’s children: key national indicators of well-being, 2009. Washington, DC: US Government Printing Office, 2009. [Google Scholar]
  • 36.Gibson-Davis CM, Brooks-Gunn J. Couples’ immigration status and ethnicity as determinants of breastfeeding. Am J Public Health 2006;96:641–646. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Laumann EO, Youm Y. Racial/ethnic group differences in the prevalence of sexually transmitted diseases in the United States: a network explanation. Sex Transm Dis 1999;26:250–261. [DOI] [PubMed] [Google Scholar]
  • 38.Li RW, Darling N, Maurice E, Barker L, Grummer-Strawn LM. Breast-feeding rates in the United States by characteristics of the child, mother, or family: the 2002 National Immunization Survey. Pediatrics 2005;115:E31–E37. [DOI] [PubMed] [Google Scholar]
  • 39.US Census Bureau. Census 2000 Summary File 2, Table PCT27, United States. Washington, DC: US Census Bureau, 2008. [Google Scholar]

RESOURCES