Density-dependent mortality of the human host in onchocerciasis: relationships between microfilarial load and excess mortality

Abstract

Background: The parasite Onchocerca volvulus has, until recently, been regarded as the cause of a chronic yet non-fatal condition. Recent analyses, however, have indicated that in addition to blindness, the parasite can also be directly associated with human mortality. Such analyses also suggested that the relationship between microfilarial load and excess mortality might be non-linear. Determining the functional form of such relationship would contribute to quantify the population impact of mass microfilaricidal treatment.

Methodology/principal findings: Data from the Onchocerciasis Control Programme in West Africa (OCP) collected from 1974 through 2001 were used to determine functional relationships between microfilarial load and excess mortality of the human host. The goodness-of-fit of three candidate functional forms (a (log-) linear model and two saturating functions) were explored and a saturating (log-) sigmoid function was deemed to be statistically the best fit. The excess mortality associated with microfilarial load was also found to be greater in younger hosts. The attributable mortality risk due to onchocerciasis was estimated to be 5.9%.

Conclusions/significance: Incorporation of this non-linear functional relationship between microfilarial load and excess mortality into mathematical models for the transmission and control of onchocerciasis will have important implications for our understanding of the population biology of O. volvulus, its impact on human populations, the global burden of disease due to onchocerciasis, and the projected benefits of control programmes in both human and economic terms.

Figure 1. Relative mortality risk with increasing Onchocerca volvulus microfilarial load in the Onchocerciasis Control Programme cohort.

Observed (open squares) and fitted (solid line) relative risk of mortality, with fitted (log-)sigmoid dose-response model adjusted to the average age of individuals within the cohort. Shaded (grey) area represents the 95% Bayesian credible interval around the fitted line; vertical error bars are 95% confidence intervals around observations. Inset permits visual inspection of the mortality relative risk at parasite loads ≤40 microfilariae per skin snip.

Figure 2. Observed and fitted relative mortality risk with Onchocerca volvulus microfilarial skin load according to age-group.

Individuals <20 years old (open squares and solid line respectively); individuals ≥20 years old (open circles and dashed line respectively). The fitted sigmoid dose-response model is adjusted to the average age of the respective age groups. Shaded (grey) areas around the fitted lines represent 95% Bayesian credible intervals; error bars represent 95% confidence intervals around observations. Inset permits visual inspection of the mortality relative risk at parasite loads ≤40 microfilariae per skin snip.

Figure 3. Measurement error-adjusted microfilarial loads plotted against their observed values.

Each panel corresponds to an adjustment for a different assumed magnitude of measurement error as defined by parameter k of the negative binomial measurement error model (see Measurement Error section in the main text). In panel A, k→∞ which corresponds to Poisson measurement error. In panel B, k = 15 as estimated from published data (see Protocol S3). In panel C, k = 1, which corresponds to an arbitrarily large degree of measurement error. In each panel the solid red line is the diagonal representing perfect agreement between observed and adjusted microfilarial loads.

Figure 4. Measurement error-adjusted and fitted relative mortality risk in the Onchocerciasis Control Programme cohort.

Panels from top to bottom correspond to adjustments for increasing assumed magnitudes of measurement error as defined by parameter k of the negative binomial measurement error model (see Measurement Error section in the main text). In panel A, k→∞, corresponding to Poisson measurement error. In panel B, k = 15, as estimated from published data (Protocol S3). In panel C, k = 1. In all panels the solid and dashed lines represent, respectively, the fitted (log-)sigmoid and (log-)linear dose-response models. Note the absence of error bars around the model-derived point estimates of relative risk (Protocol S1) and around the fitted dose-response. This is because regression calibration cannot account fully for the uncertainty introduced by adjusting the observed data for measurement error (see Discussion).