Herpes Simplex Virus Type 2 Seroprevalence and Ultrasound-diagnosed Uterine Fibroids in a Large Population of Young African-American Women

Kristen R. Moore; Jennifer S. Smith; Stephen R. Cole; Victor J. Schoenbach; Katherine Schlusser; Charlotte A. Gaydos; Donna D. Baird*


Am J Epidemiol. 2016;183(11):961-968. 

In This Article


Study Participants and Data Collection

We used transvaginal ultrasound results, self-reported questionnaire data, and stored frozen serum specimens from participants in SELF, an ongoing study based in the Detroit, Michigan, area. SELF is a prospective cohort study of fibroid development. Enrollment data and specimen collection protocols have been described previously.[18] In brief, from November 2010 to December 2012, the study recruited 1,696 African-American volunteers; all of the women were aged 23–34 years. Women were not eligible for SELF if they had previously been diagnosed with uterine fibroids; had had a hysterectomy; had ever taken medication to treat lupus, Grave's disease, Sjögren syndrome, scleroderma, or multiple sclerosis; or had ever had any type of cancer treated with radiation or chemotherapy. The study was approved by the institutional review boards of the National Institute of Environmental Health Sciences (Research Triangle Park, North Carolina) and the Henry Ford Health System (Detroit, Michigan).

Fibroid Assessment

Transvaginal ultrasound, the standard procedure for the detection and diagnosis of fibroids, was used in this study.[37] Fibroids were assessed by study sonographers as described previously[18] at one of 3 clinics. Focal fibroids of 0.5 cm diameter or greater were measured in triplicate. For each measurement, the 3 perpendicular diameters (longitudinal, anterior-posterior, and transverse) were recorded.

Outcome Definitions

The primary outcome of this study was the presence of fibroids (yes or no) at the transvaginal ultrasound examination completed at enrollment. Fibroids were classified as "questionable" when at least 1 diameter could not be measured. Participants with at least 1 fibroid or questionable fibroid greater than or equal to 0.5 cm in diameter at the enrollment ultrasound were considered to have fibroids, and all women without a fibroid or questionable fibroid at least 0.5 cm in diameter were considered not to have fibroids. The secondary outcomes for this analysis were size of the largest fibroid, number of fibroids, and total fibroid volume. The size of the largest fibroid was estimated by averaging the maximum diameter (longitudinal, anterior-posterior, or transverse) of each of the triplicate fibroid measurements. Fibroid volume (cm3) was measured by computing the volumes of each of the triplicate fibroid measurements using the ellipsoid formula (longitudinal × anterior-posterior × transverse × 0.5233) and averaging across the 3 volumes. Total fibroid volume (cm3) was calculated by adding the average volumes from each of a woman's fibroids. Total fibroid volume was not computed for women with only questionable fibroids (n = 6) in whom at least 1 diameter was not measured.

HSV-2 Assessment

HSV-2 antibody serostatus was assessed in the International Sexually Transmitted Diseases Research Laboratory at Johns Hopkins University (Baltimore, Maryland) using the Focus Diagnostics HerpeSelect 2 ELISA IgG test (Focus Diagnostics, Cypress, California) according to the package instructions.[38] This type-specific assay (glycoprotein gG2) allows for the qualitative detection of HSV-2 immunoglobulin G antibodies with or without the presence of herpes simplex virus type 1 (HSV-1).[38] Antibody levels greater than 1.10 optical density units were categorized as seropositive and levels less than 0.90 were seronegative. Levels of 0.9 to 1.10 were considered indeterminate, and those samples were retested. The level of antibody response cannot be used to determine active infection or recency of initial infection.

For quality control, we included blinded duplicate samples. Aliquots were created from unused serum samples collected from SELF participants during a special blood drawing performed 6 months after enrollment.[18] Two aliquots from each of 42 specimens served as 84 blinded control samples. Forty of the 42 duplicate pairs had identical results. For each of the other 2 discordant pairs, 1 of the aliquots was considered indeterminate. The result for the other aliquot of each of these 2 discordant pairs matched the result found for the enrollment specimen from the same participant.

Statistical Analyses

Due to the use of telephone and Web-based interview methods that did not allow participants to skip questions, there was a minimal amount of missing data on covariates. However, participants were able to respond "prefer not to answer." This response (<1%) was coded as missing, and complete case analysis was performed. Variables were categorized on the basis of the distribution of the data and comparability with the literature. All analyses were completed using data from women who had blood samples taken at enrollment. Standard descriptive statistics were performed for all variables of interest. We also evaluated the relationship between HSV-2 serostatus and number of sexual partners and age at first intercourse for comparison with the literature. All analyses were conducted using SAS, version 9.3 (SAS Institute, Inc., Cary, North Carolina).

Primary Analyses. Logistic regression was used to compute odds ratios and 95% confidence intervals to evaluate the relationship between HSV-2 serostatus and the presence of fibroids. Although the relative odds overestimate the relative risk for associations with an outcome as common as fibroids, overestimation is minimal for weak associations and does not affect statistical tests. Potential confounders and variables of interest were determined based on a review of the literature, and a directed acyclic graph[39] was used to provide a conceptual framework. Potential confounders included age in years (continuous), age at menarche (7–10 years or 11–19 years), alcohol consumption (low, moderate, or heavy), and use of DMPA (ever or never). The alcohol-consumption variable reflected the drinking level each woman reported for the age(s) at which she was drinking the most. Low alcohol consumption was defined as having had less than 10 alcoholic drinks in a year. Heavy drinkers were those who usually consumed 6 or more drinks on days when they imbibed alcohol or who consumed 4 or more drinks per sitting at least 2–3 times a month. All others were considered moderate drinkers. Variables of interest based on the literature included education (high school graduation, General Educational Development certificate, or less; some college, an associate degree, or technical education; or bachelor's, master's, or doctoral degree), and body mass index (15–24, 25–29, 30–34, or ≥35), which was calculated as weight (kg)/height (m)2. Parity (nulliparous or parous) was included to increase precision because of the consistent association between parity and fibroids in the literature. We included age, age at menarche, parity, and DMPA use in a full model a priori. The alcohol variable was not included a priori because the amount of literature supporting an association between alcohol consumption and fibroids is small. After adding alcohol and the additional variables of interest, we used backwards elimination and the 10%-change-in-estimate approach to determine the minimally adjusted final model. Only the covariates included a priori remained in the final model.

Secondary analyses. In a secondary analysis, we examined the association between HSV-2 and fibroid characteristics: size of the largest fibroid, number of fibroids, and total fibroid volume. Median values rounded to the nearest whole number were used to determine category cutpoints for size of the largest fibroid and total fibroid volume (2 cm and 2 cm3, respectively). We used multinomial logistic regression models to estimate the odds ratios and 95% confidence intervals for the associations of HSV-2 serostatus with fibroid size, number, and total volume. Moreover, because the majority of HSV-2 infections are asymptomatic, self-reported infections are likely to be those that are symptomatic. Thus, in a secondary analysis, we evaluated the multivariable-adjusted odds ratios and 95% confidence intervals for the relationship between HSV-2 and fibroid presence using a 3-level exposure variable: 0 = seronegative, 1 = not symptomatic (seropositive with no self-reported diagnosis), and 2 = symptomatic (seropositive with self-reported diagnosis). For both of these secondary analyses, we adjusted for the same confounders as those included in the final model in the primary analysis (age, age at menarche, parity, and DMPA use).

Sensitivity Analyses. To evaluate the robustness of our findings, we examined multivariable-adjusted odds ratios and 95% confidence intervals for the association of HSV-2 with fibroids in a series of sensitivity analyses by restricting or stratifying the sample. Cervical treatment is an indicator of cervical lesions secondary to persistent human papillomavirus infection and was found to be inversely associated with fibroids in a previous study.[40] Thus, any HSV-2 associations we observed might have been attenuated. Therefore, we repeated the primary analyses after excluding those who reported cervical treatment (231 women excluded). To evaluate age as an effect modifier (due to possible decline in antibody levels over time), we evaluated the association between HSV-2 and fibroid presence in 2 age strata (23–29 years or 30–35 years).

To investigate temporality of HSV-2 exposure in relation to fibroid development, we assessed the association between HSV-2 and fibroids among strata of size of the largest fibroid (<2 cm vs. ≥2 cm). Those with smaller fibroids are more likely to have developed fibroids more recently and thus may be more likely to have had exposure to HSV-2 before fibroid development compared with women with larger fibroids. In addition, we looked at the association of HSV-2 and fibroid presence within 2 strata of number of sexual partners before age 20 years (0 or 1 vs. 2 or more). Those with more sexual partners before the age of 20 years would have been more likely to be exposed to HSV-2 prior to fibroid development.

A total of 1,696 women were enrolled in the SELF study. HSV-2 serology was conducted for the 98% of participants with available enrollment blood samples (n = 1,662). Four of the samples had indeterminate results that were excluded. Therefore, our analyses were performed on the 1,658 participants with HSV-2 serology results.