OBJECTIVE: To examine risk factors for vulvovaginal candidiasis among women with or at risk for human immunodeficiency virus (HIV) infection. METHODS: Data were from 856 HIV-infected women and 421 at-risk uninfected women observed semiannually at four study sites from April 1993 through February 1999. At enrollment women were 15-55 years old and had no acquired immunodeficiency syndrome-defining conditions. Three definitions for vulvovaginal candidiasis of differing severity were constructed using data from vaginal Candida culture and Gram stains scored for yeast and three signs on pelvic examination (vulvovaginal edema, erythema, or discharge): 1) culture or Gram stain positivity plus at least one clinical sign, 2) culture or Gram stain positivity plus at least two clinical signs, and 3) visible yeast on Gram stain plus at least one clinical sign. RESULTS: The prevalence and cumulative incidence of each definition of vulvovaginal candidiasis were greater among HIV-infected women than among women not infected with HIV (P <.01 for all comparisons). Stratified by status at the preceding visit, vulvovaginal candidiasis was most likely among women with prior vulvovaginal candidiasis, least likely among women without earlier Candida colonization, and intermediately likely among women with preceding subclinical Candida colonization. Among HIV-infected women, lower CD4 count and higher HIV viral load were associated with vulvovaginal candidiasis. Several other factors were independently associated with vulvovaginal candidiasis, with strong associations for diabetes mellitus and pregnancy in particular. Vulvovaginal candidiasis was not more severe among HIV-infected women. CONCLUSION: Vulvovaginal candidiasis occurred with higher incidence and greater persistence, but not greater severity, among HIV-infected women.
OBJECTIVE: To examine risk factors for vulvovaginal candidiasis among women with or at risk for human immunodeficiency virus (HIV) infection. METHODS: Data were from 856 HIV-infectedwomen and 421 at-risk uninfected women observed semiannually at four study sites from April 1993 through February 1999. At enrollment women were 15-55 years old and had no acquired immunodeficiency syndrome-defining conditions. Three definitions for vulvovaginal candidiasis of differing severity were constructed using data from vaginal Candida culture and Gram stains scored for yeast and three signs on pelvic examination (vulvovaginal edema, erythema, or discharge): 1) culture or Gram stain positivity plus at least one clinical sign, 2) culture or Gram stain positivity plus at least two clinical signs, and 3) visible yeast on Gram stain plus at least one clinical sign. RESULTS: The prevalence and cumulative incidence of each definition of vulvovaginal candidiasis were greater among HIV-infectedwomen than among women not infected with HIV (P <.01 for all comparisons). Stratified by status at the preceding visit, vulvovaginal candidiasis was most likely among women with prior vulvovaginal candidiasis, least likely among women without earlier Candida colonization, and intermediately likely among women with preceding subclinical Candida colonization. Among HIV-infectedwomen, lower CD4 count and higher HIV viral load were associated with vulvovaginal candidiasis. Several other factors were independently associated with vulvovaginal candidiasis, with strong associations for diabetes mellitus and pregnancy in particular. Vulvovaginal candidiasis was not more severe among HIV-infectedwomen. CONCLUSION:Vulvovaginal candidiasis occurred with higher incidence and greater persistence, but not greater severity, among HIV-infectedwomen.