Are Vaginal Symptoms Ever Normal? A Review of the Literature

Abstract

Purpose

Vaginal symptoms such as discharge, odor, and itch are among the most common presenting complaints in primary care. We undertook to determine if the symptoms associated with vaginitis (discharge, odor, irritation) occur in normal women.

Methods

To answer this question, we performed a literature review. We conducted a Medline search using the following terms: "vagina," "vaginal discharge," "secretion," "odors," discharge," "pruritus," "normal," "irritation," "itch," "physical examination," "healthy," "asymptomatic," "quantity," and "physiology." To find additional references we reviewed textbooks in gynecology, primary care, and physical diagnosis and contacted authors.

Results

There are few primary studies, and most are not of high quality. Existing data show that the quantity and quality of vaginal discharge in healthy women vary considerably both across individuals and in the same individual during the menstrual cycle. Most studies indicate that discharge is greatest at midcycle. Vaginal fluid contains malodorants, and one study of intact vaginal fluid found it to be malodorous. Two studies found that normal women reported irritative symptoms in the course of their menstrual cycle.

Conclusions

The primary literature indicates that there is a wide variation in the normal vagina and that some of the symptoms associated with vaginal abnormality are found in well women. Both clinicians and their patients would benefit from a better understanding of the range of normal as well as what constitutes a meaningful departure from that range.

Introduction

Vaginal symptoms are among the most common reasons for gynecologic consultation in primary care.[1] In current practice, the evaluation of vaginal complaints in premenopausal women is based primarily on the diagnosis of 1 of 3 conditions: bacterial vaginosis (BV), trichomoniasis, and vaginal candidiasis.[2-4] Despite the general acceptance of this approach, studies in a variety of settings have demonstrated that neither BV nor any pathogenic microbe can be found in approximately a third of symptomatic women.[5,6] Furthermore, asymptomatic women can be infected with candida[7] and trichomonas,[8] whereas Gardnerella vaginalis is considered part of the normal vaginal flora.[9] Thus, the presence or absence of a microbe corresponds poorly with the presence or absence of symptoms. Clinicians are often faced with symptoms for which there is no obvious cause.

One possible explanation for women who have vaginal symptoms and no identifiable microbial cause is that they are interpreting normal physiologic processes as evidence of disease. This explanation, in turn, raises the question of how patients and physicians decide what is normal and what is diseased.

In previous work, we found that patients with the diagnosis of vaginitis agreed very little on the characteristics of a normal vaginal discharge.[10] Although most women thought that some vaginal discharge was normal, a substantial minority felt that a healthy vagina should be dry. Among those women who accepted vaginal wetness as normal, there was little agreement about the characteristics of this discharge, such as timing or color.

In informal conversations with our physician colleagues, we encountered similar disagreement. Rules for determining whether discharge looked "abnormal" varied from person to person and seemed to be based more on personal experience than on textbook learning. Physicians commented that some women with a "normal" discharge complained that it was abnormal, while other women with an abnormal discharge seemed unaware of the fact that something was amiss.

To clarify what constituted "normal," we turned to textbooks. Again we found disagreement. Textbooks offered a variety of conflicting terms to describe the color of normal discharge (white, gray, yellow, clear) and its consistency (clumpy, floccular, semisolid, cream, high viscosity, nonhomogeneous, thin).[11-14] Although most textbook authors stated that vaginal fluids were either odorless, minimally odorless, or did not have an unpleasant odor,[12,13,15-17] 1 book mentioned the possibility that the normal vaginal odor might be unpleasant.[18] Textbooks, however, generally agreed that irritative symptoms -- itching, irritation, redness, and swelling -- should not normally be present.[12,15,17]

Given the conflicts in patient accounts, physician attitudes, and textbook teachings, we decided to conduct a review of the primary literature. Initially we had considered undertaking a systematic review. The literature, however, did not lend itself well to systematization given the diversity of methodologies and end points used by researchers in this field.

Methods

Literature Review/Primary Sources

We searched Medline using combinations of the terms "vagina," "vaginal discharge," "secretion," "odors," discharge," "pruritus," "normal," "irritation," "itch," "physical examination," "healthy," "asymptomatic," "quantity," and "physiology." Abstracts from all articles were reviewed, and those that appeared relevant were retrieved. Additional references were obtained from both these articles and textbooks in primary care, physical diagnosis, family medicine, and ambulatory care. We attempted to contact, via email or regular mail, all authors of all primary articles to obtain additional references; we were able to contact 3 (Dr. Richard Doty, Dr. George Preti, and Dr. Hanna Klaus), and Dr. Klaus provided us with an additional reference.

Results

Our search led us to 8 primary articles and 1 review. We have grouped the information from these sources into 3 categories: (1) normal vaginal discharge, (2) normal vaginal odor, and (3) normalcy of irritative symptoms.

Characteristics of Normal Vaginal Discharge

In 1959, Stone[19] examined a group of 113 New York City women, an unknown number of whom showed evidence of cervical or vaginal infection; the nature of that evidence was not specified in the article. Stone used weight as a method to quantitate discharge. He swabbed the entire vagina during the course of an exam and noted the weight change in the swab. He found a mean vaginal discharge of 0.76 g for all patients, 1.0 g for women with vaginal infection and 0.50 g in women who had douched on the day before the exam. The mean weight of vaginal discharge increased among women who were peri-ovulational.

Godley[20] evaluated a group of 27 women who had no vaginal symptoms and considered their discharge normal. Women were provided with tampons that were worn for 8 hours and then mailed to the investigator in a plastic container. Discharge was quantitated by the weight change in the tampon. In this study, mean discharge was 1.55 g/8 hours, with a standard deviation of 0.6. Godley also reported a midcycle increase in discharge. Lowest values for discharge weight were obtained on Day 7 of the cycle (1.38 g/8h) and Day 26 (1.37 g/8h), and the highest value was on Day 14 (1.96 g/8h).

In a study looking at multiple aspects of the menstrual cycle, Moghissi[21] evaluated 10 women who had undergone a thorough medical examination to exclude both medical and gynecologic disease. They were studied throughout the menstrual cycle, and cervical mucus production was evaluated during a speculum examination. Mucus was evaluated in terms of quantity (measured in milliliters), viscosity, and spinnbarkeit. Cervical mucus increased from 2.68 mL during the follicular cycle to 3.97 mL during midcycle and declined to 2.13 mL during the luteal phase. Viscosity showed an inverse pattern, decreasing at midcycle and increasing to a maximum in the luteal phase. Spinnbarkeit was maximal at midcycle, specifically on the day before the luteinizing hormone surge, dropping dramatically in the luteal phase.

The observation that variation in cervical secretions was associated with changes in vaginal discharge led the Australian physician, John Billings,[22,23] to study vaginal secretions as a way of predicting ovulation. Billings studied several hundred women using self-report, often correlated with biochemical markers of ovulation. He described a typical pattern of vaginal discharge on the basis of changes in cervical mucus. The pattern began with a postmenses "dry period" followed by a period of discharge attributable to increasing production of cervical mucus. The discharge resulting from this mucus was initially opaque and sticky. At the time of ovulation, the discharge was stretchy, wet, and slippery (reflecting spinnbarkheit), becoming opaque and tacky later on in the cycle. The typical pattern might be altered by infection or semen. The "ovulation method" or "Billings ovulation method" is promoted either for purposes of family planning or infertility treatment.[24]

The finding of a peri-ovulational increase in vaginal fluid was not confirmed in a 2000 study by Eschenbach and colleagues.[25] The authors evaluated 74 women (24 of whom had evidence of "asymptomatic BV") at 3 different times of the menstrual cycle (Days 1-5, 7-12, and 19-24), collecting information on symptoms and physical examination findings. The amount of vaginal discharge was estimated by instilling the vagina with 3 mL of phosphate buffered saline (PBS) using a pipette, removing all fluid, and estimating the increase in the aspirate. Discharge was graded as scant (< 1 mL), normal (1-3 mL), or copious (> 3 mL). Discharge volume increased over the 3 phases of the study; cervical mucus was greatest during Days 1 to 5. "Most" subjects had white or clear discharge; women with yellow discharge were excluded from the study, but yellow discharge later occurred in about 10% of the women. The discharge from 65 women had a "normal" consistency, the discharge from 8 women was judged homogenous, and 1 woman had a curdy discharge. Forty-eight subjects had a discharge that was pooled; 24 had a diffuse discharge, and 2 had a patchy discharge.

Characteristics of Normal Vaginal Odor

There is an extensive literature discussing the biochemical composition of vaginal fluid and the odor of its constituents. This literature demonstrates clearly that there are malodorous components to vaginal fluids.[26] However, we could locate only 1 study examining the odor of intact vaginal fluid and 1 study that provided information on patient self-report.

Doty[27] examined 4 women who collected samples of their vaginal secretions on a tampon every other day (excluding weekends) for 4 cycles. Bottles containing tampons from 15 ovulatory cycles were presented to 73 blinded volunteers (37 men) who rated the odor of the bottle on scales of intensity and pleasantness. The majority of raters found the odors unpleasant. There was a wide variation in pleasantness and intensity of odor, both among individual women throughout the cycle and between different women at same stages of the cycle. Several of the blinded observers remarked that the substances they smelled were "either deodorizing products, cheeses, or preservatives for food such as turkey or fish."

In the Eschenbach study cited above,[25] only 1 of the 50 women without evidence of BV reported a vaginal odor; no information was provided on the women with evidence of BV.

Normalcy of Irritative Symptoms

Two studies addressed the presence of irritative symptoms in normal women. In 1997, Priestley and colleagues[28] reported on 26 volunteer healthcare workers who took vaginal swabs periodically throughout an 8-week period and maintained symptom diaries. The authors divided the women into 4 groups on the basis of microscopy and cultures: a normal group (8 subjects), a candida group (8 subjects), a BV group (10 subjects), and an ureaplasma group (10 subjects). An individual woman might change her allocation during the study depending upon the results of her smears. Of the 8 normal patients, 6 had some symptoms of irritation, discharge, or odor during the course of the study. Intriguingly, there was no significant difference noted between the groups with regard to the number of women who had symptoms more than 2 days per week.

Eschenbach[25] reported that 6 of 74 subjects (8%) experienced pruritus during the premenstrual period (Days 19-24 of the cycle); 4 of these had positive candida cultures.

Discussion

A review of the primary literature on vaginal symptoms suggests a number of clinically useful conclusions. First, there is a wide variation in the normal, and many of the symptoms associated with vaginal abnormality are found in well women. Healthy women experience vaginal discharge. The quantity of discharge appears to vary from woman to woman and across an individual woman's menstrual cycle. Normal vaginal fluid contains malodorants and may have an odor that can be unpleasant. There is a suggestion that healthy women can experience irritative symptoms.

These conclusions challenge some of what is written in the textbooks we consulted. This is a troubling finding. Physicians who are misinformed as to the nature of vaginal wetness, odor, and irritation may impose the diagnosis of vaginitis on normal women who would then be subject to needless worry and medication. If true, this would be consistent with medicine's historical tendency to interpret the normal functions of the female reproductive system as diseased.[29] Although discharge, odor, and irritation may be normal, they may also be legitimate indicators of disease. Given the prevalence of vaginal symptoms, it is important for primary care researchers to develop clinically useful ways of making the distinction between normal and pathologic.

Unfortunately, there were important limitations to some of the articles we reviewed. For example, not all of the papers provided clear and detailed inclusion and exclusion criteria,[25] and some studies included women with infections. What would better-quality studies look like?

Such studies must have some criteria for "normality." In general, the studies reviewed in this article used 1 of 2 definitions of normal. Women were normal if either they had been examined thoroughly and no disease found ("clinically normal") or if they declared themselves normal ("self-report"). There are other possible definitions of normality, including those based on a statistically derived norm or on a threshold value beyond which diagnosis and treatment become beneficial.[30]

Better studies will require more clinically useful measures. Measuring grams of discharge or milliliters of cervical mucus is not practical during a pelvic examination. If there is a wide variation in normal, it may be impossible to provide precise measures of normality for measures such as quantity of discharge. Alternative definitions of normal, such as change from usual pattern[31] or functional difficulties, may be more useful.

Indeed, for patients, social factors may play the key role in determining what is or is not normal. Some patients with vaginal symptoms seek medical consultation when the discharge interferes with functioning -- either sexual relations or the ability to go to work or attend school.[10] "Symptoms" are a nearly ubiquitous human experience, yet the decision to seek medical consultation is often triggered by some social difficulty.[32]

Although vaginal symptoms are only rarely portents of serious disease, they are distressing to patients. Unfortunately, a good evidence base upon which to decide when a discharge "looks like something" is lacking. Both clinicians and their patients would benefit from a better understanding of the range of normal as well as what constitutes a meaningful departure from that range.