The National Health and Nutrition Examination Survey (NHANES, 2003–2004) is the most recently available nationally representative dataset that measures vaginal HPV infection with a medical exam. The survey included questions related to acculturation. A subset of the survey population participated in physical examinations. These were carried out in a set of specially-equipped Mobile Examination Centers (MEC). HPV DNA was detected by the HPV L1 consensus polymerase chain reaction with biotinylated PGMY09/11 primer sets and B-globin as internal control for sample amplification using self-collected vaginal swabs as instructed by a testing center physician (Catch-All Sample Collection Swabs Epicenter, Madison, WI) (Dunne et al., 2007
Women eligible for HPV laboratory assessment were ages 14–59. About 85% of eligible women participated. Of those, 95% of the tests were adequate for analysis. Positive samples were typed for 37 HPV types (Kahn et al., 2007
). Of these, 15 high-risk types, are known to lead to cancer (Munoz et al., 2004
NHANES measured acculturation using a series of questions that assessed primary language, country of birth, ethnic identity, country of father’s birth, and country of mother’s birth. A language proficiency score was developed using the Short Acculturation Scale (SAS) developed by Marin et al. (1987). This scale investigates level of language acculturation which has shown to also correlate highly with respondents’ generation, length of residence in the US among foreign-born individuals, and age at arrival to the US. The SAS included five items, which relate to language that the participant reads and speaks, speaks at home, spoke as a child, uses to currently think, and uses to speak with friends. Higher scores represent higher levels of acculturation. Past studies determined that individuals with an average score above 2.99 on the SAS are highly acculturated and those scoring an average of 2.99 or less demonstrate a low level of acculturation (Marin & Posner, 1995
In this study, acculturation was examined using the following three elements from the NHANES acculturation supplement:
- SAS language proficiency score (high or low language acculturation)
- Country of birth (US or elsewhere)
- Parents’ country of birth (Both born in the US, one born in the US, both born elsewhere)
These three elements have been found to account for a large portion of the variance in acculturation measures (Coronado et al., 2005
). We also investigated the relationship between HPV infection and a number of other sociodemographic variables, including family income (less than and equal or greater than $20,000), age, and marital status using chi-square tests. Characteristics that had a statistically significant relationship (p<0.05) with HPV infection and a possible association to level of acculturation were included in the multivariable models. Next, multivariable logistic regression analysis assessed the association of acculturation and HPV infection risk for US Latinas. Sensitivity analyses were conducted to explore the association between level of acculturation and high-risk oncogenic HPV infection (HPV types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, 68, and 73). Mediation by reported sexual behaviors was also explored using the Sobel-Goodman mediation test which quantifies the proportion of the total effect explained by the mediating variable.
Lastly, all multivariable analyses accounted for the stratification and clustering of data within the complex survey design of NHANES using multiple stage survey functions. STATA 10.0 was used to conduct all statistical analyses (College Station, Texas). The research was considered exempt by the University of Washington Institutional Review Board (IRB).