Lactobacilli Dominance and Vaginal pH: Why Is the Human Vaginal Microbiome Unique?

Abstract

The human vaginal microbiome is dominated by bacteria from the genus Lactobacillus, which create an acidic environment thought to protect women against sexually transmitted pathogens and opportunistic infections. Strikingly, lactobacilli dominance appears to be unique to humans; while the relative abundance of lactobacilli in the human vagina is typically >70%, in other mammals lactobacilli rarely comprise more than 1% of vaginal microbiota. Several hypotheses have been proposed to explain humans' unique vaginal microbiota, including humans' distinct reproductive physiology, high risk of STDs, and high risk of microbial complications linked to pregnancy and birth. Here, we test these hypotheses using comparative data on vaginal pH and the relative abundance of lactobacilli in 26 mammalian species and 50 studies (N = 21 mammals for pH and 14 mammals for lactobacilli relative abundance). We found that non-human mammals, like humans, exhibit the lowest vaginal pH during the period of highest estrogen. However, the vaginal pH of non-human mammals is never as low as is typical for humans (median vaginal pH in humans = 4.5; range of pH across all 21 non-human mammals = 5.4-7.8). Contrary to disease and obstetric risk hypotheses, we found no significant relationship between vaginal pH or lactobacilli relative abundance and multiple metrics of STD or birth injury risk (P-values ranged from 0.13 to 0.99). Given the lack of evidence for these hypotheses, we discuss two alternative explanations: the common function hypothesis and a novel hypothesis related to the diet of agricultural humans. Specifically, with regard to diet we propose that high levels of starch in human diets have led to increased levels of glycogen in the vaginal tract, which, in turn, promotes the proliferation of lactobacilli. If true, human diet may have paved the way for a novel, protective microbiome in human vaginal tracts. Overall, our results highlight the need for continuing research on non-human vaginal microbial communities and the importance of investigating both the physiological mechanisms and the broad evolutionary processes underlying human lactobacilli dominance.

Keywords: estrogen; evolution; lactobacilli; mammals; pH; vaginal microbiome.

Figure 1

Vaginal pH across 22 species of mammals including humans. Open circles represent mean pH from individual studies and diamonds represent the overall mean for that species. Diamonds are color-coded based on taxonomic order. Error bars represent the standard deviation from the mean. Humans are divided into two groups, one with bacterial vaginosis (BV) and one without BV. Dendrogram indicates the evolutionary distance between species in millions of years (Myr).

Figure 2

The mean relative abundance of Lactobacillus spp. vs. other bacteria in (A) humans and (B) non-human mammals. For non-human mammals, lactobacilli relative abundance was calculated as the mean across all species (N = 14). The standard error of the mean for lactobacilli was ±0.046% in humans and ±0.39% in other mammals. See Table 2 for the list of studies used to generate this figure.

Figure 3

Mean vaginal pH between periods of high and low estrogen of 11 mammalian species during the ovarian cycle. Paired black diamonds represent the overall mean for each species at both estrogen levels. The open diamonds and dashed line show the mean human vaginal pHs during the high estrogen phases (follicular phase and ovulation) and low estrogen phases (luteal phase and menstruation) of the ovarian cycle. Letters next to diamonds identify the mammalian species. See Table S1 for the list of studies used to generate this figure. Abbreviations: C, Cow; D, Dog; H, Horse; OB, Olive baboon; P, Common brushtail possum; PM, Pig-tailed macaque; R, Brown rat; S, Sheep; SM, Black-handed spider monkey; YB, Yellow baboon.

Figure 4

Mean vaginal pH as a function of STD or obstetric risk across mammals. Each point represents one species. Asterisks show where humans fall within each comparison. STD risk proxies are (A) relative testes mass, (B) baseline white blood cell count, (C) annual sexual receptivity, (D) maximum lifetime reproductive events, and (E) intromission pattern (SBI, single brief intromission; MBI, multiple brief intromissions; SPI, single prolonged intromission). Obstetric risk proxies are (F) gestation length, (G) relative neonatal mass, and (H) relative maternal pelvic area. Risk level increases moving left to right on all plots (note the reversed x-axis in plot H). The solid lines represent the best-fit linear models without humans.