Serology reflects a decline in the prevalence of trachoma in two regions of The Gambia
======================================================================================

* Stephanie J Migchelsen
* Nuno Sepúlveda
* Diana L Martin
* Gretchen Cooley
* Sarah Gwyn
* Harry Pickering
* Hassan Joof
* Pateh Makalo
* Robin Bailey
* Sarah E. Burr
* David CW Mabey
* Anthony W Solomon
* Chrissy h Roberts

## ABSTRACT

Trachoma is caused by *Chlamydia trachomatis* (Ct). It is targeted for global elimination as a public health problem. In 2014, a population-based cross-sectional study was performed in two previously trachoma-endemic areas of The Gambia. Participants of all ages from Lower River Region (LRR) (N = 1028) and Upper River Region (URR) (N = 840) underwent examination for trachoma and had blood collected for detection of antibodies against the Ct antigen Pgp3, by ELISA. Overall, 30 (1.6%) individuals had active trachoma; the prevalence in children aged 1–9 years was 3.4% (25/742) with no statistically significant difference in prevalence between the regions. There was a significant difference in overall seroprevalence by region: 26.2% in LRR and 17.1% in URR (p<0.000l). In children 1–9 years old, seroprevalence was 4.4% in LRR and 3.9% in URR. Reversible catalytic models using information on age-specific seroprevalence demonstrated a decrease in the transmission of Ct infection in both regions, possibly reflecting the impact of improved access to water, health and sanitation as well as mass drug administration campaigns. Serological testing for antibodies to Ct antigens is potentially useful for trachoma programmes, but consideration should be given to the co-endemicity of sexually transmitted Ct infections.

## INTRODUCTION

Trachoma is caused by ocular infection with the obligate intracellular bacterium *Chlamydia trachomatis* (Ct) and is the leading infectious cause of blindness worldwide1. Infection is associated with clinical signs of inflammation in the conjunctiva known as active trachoma; these include trachomatous inflammation—follicular (TF) and trachomatous inflammation—intense (Tl). Many repeated episodes of active trachoma over years to decades can lead to trachomatous trichiasis (TT), which may lead to impaired vision. The World Health Organization (WHO) estimates that over 200 million people in 42 countries are at risk of blindness from trachoma 2. In 2010, approximately 1.9 million people suffered from visual impairment or blindness due to trachoma1. The WHO Alliance for the Global Elimination of Trachoma by 2020 (GET2020) aims to eliminate trachoma as a public health problem by 20202 through the SAFE Strategy (Surgery, Antibiotics, Facial cleanliness and Environmental improvement). The target is to reduce the prevalence of TF to <5% in children aged 1–9 years, and the prevalence of unmanaged TT to <0.2% in adults aged 15 years and above3. By 2014, seven countries had reported having met these targets nationally4.

Evaluating the effectiveness of trachoma elimination programmes can be difficult, and current guidelines for post-intervention surveillance of Ct transmission intensity, like the elimination threshold, are based on the prevalence of TF in children aged 1–9 years. This is biologically problematic because in areas of low endemicity5 and in populations where mass drug administration (MDA) of azithromycin has been undertaken6, the correlation between the clinical signs of trachoma and the presence of ocular Ct infection is poor. There is therefore a need for better means by which to assess transmission.

Antibodies against *Chlamydia trachomatis* reflect cumulative exposure to Ct7 and it has been suggested that programmes could use some measure of seroprevalence as an alternative indicator of changes in transmission8,9. Previous work has begun to investigate for the use of age-specific seroprevalence for surveillance in the post-MDA setting9–11.

Serological techniques for the detection of antibodies against Ct have been used to study the epidemiology of both urogenital and ocular infections12,13. Recently, an enzyme-linked immunoassay (ELISA) which detects antibodies against Pgp3 (pCT03)—a highly immunogenic Ct-specific protein encoded by the plasmid, which is highly conserved among Ct isolates14—has been used for analysis of samples collected from trachoma-endemic regions15,16.

Extensive trachoma research has been undertaken in The Gambia for over 50 years, and it is therefore possible to track the declining prevalence of trachoma over time in many areas. Two National Surveys of Blindness and Low Vision demonstrated nation-wide decreases of comparable magnitude in the prevalence of active trachoma and TT between 1986 and 200017,18. Specific elimination efforts, run by the National Eye Health Programme (NEHP), included mass drug administration (MDA) of azithromycin in 23 districts across the country between 2007 and 2010, including three rounds in the Lower River Region (LRR). The Partnership for the Rapid Elimination of Trachoma (PRET)5 was embedded within the national programme and measured the prevalence of Ct infection in children residing in four districts in which MDA had been administered. MDA was not administered in URR as the TF prevalence in URR was already below the WHO threshold for elimination as a public health problem.

The aim of the current study was to use serological data from a Pgp3-specific ELISA to gain insight into the dynamics of Ct transmission in two well-characterized regions of The Gambia. The seroconversion rate (SCR)—the yearly average rate by which seronegative individuals become seropositive upon disease exposure and a surrogate for the underlying force of infection (Fol)—was estimated under different epidemiological settings in order to explore the utility of serology as a tool for surveillance in trachoma elimination programmes.

## RESULTS

We recruited participants of all ages from LRR (n=1028, 41.9% male) and URR (n=840, 42.5% male). Ten participants were excluded from the study because they either declined to provide a blood sample (n=l) or had incomplete examination data (n=9). The median participant age was 13 years in LRR (range: 1-88, IQR 6–34) and 11 years in URR (range: 0–90, IQR 5–40). The proportion of participants by age group is shown in Table 1. There were significantly more females than males overall (X2=45.332, p<0.0001), which held true in both LRR (X2=26.483, p<0.0001) and URR (X2=18.601, p<0.0001) and is representative of Gambian demographics19. Age distributions were approximately equal between the two regions, as seen in Table 1 (X2=27.703, p=0.14). 

View this table:
[Table 1.](http://biorxiv.org/content/early/2017/06/14/149237/T1)

Table 1. 
Age distribution of study participants, Lower River Region and Upper River Region, The Gambia, 2014.

### Examination findings

Thirty cases of TF were found in total (1.6%, 95% CI=1.1–2.3%), of which 25 were in children aged 1–9 years (3.4%, 95% CI=2.2–4.9, Table 2). There were 122 cases of trachomatous conjunctival scarring (TS), of which 102 were in participants aged 15 years and above (n=875,11.6%, 95% CI=9.6–14.0); eight cases of TT, all of which were in those aged 40 years and older (0.9%, 95% CI=0.4–1.8); and one case of corneal opacity (CO), in a participant over 40 years of age (0.1%, 95% CI=0.0–0.6) (Table 2). The prevalence of TS was significantly different (X2=4.9852, p=0.03) between LRR (78/1028, 7.6%, 95% CI=6.1–9.4%) and URR (44/840, 5.2%, 95% CI=3.9–7.0%). The TF prevalence in children aged 1-9 years was not significantly different (X2=0.1343, p=0.71) between LRR (18, 1.8%) and URR (12, 1.4%). The prevalence of Tl, TT and CO in this population was too low for further statistical analysis. 

View this table:
[Table 2.](http://biorxiv.org/content/early/2017/06/14/149237/T2)

Table 2. 
Frequency of signs of trachoma in study participants, Lower River Region and Upper River Region, The Gambia, 2014.

Data from this study show very low prevalence of active trachoma in these regions. In both regions, the prevalence of TF in 1–9-year-olds was below the 5% elimination threshold specified by WHO.

### Antibody responses in the populations

The threshold for seropositivity was set at the mean of the Gaussian distribution of the seronegative population plus four standard deviations, 0.810 OD450nm to ensure high specificity. Previous studies using the Pgp3 ELISA have commonly used a threshold set as three standard deviations above the mean of the negative population (the 97.5% confidence interval) 15,20’21. Using that lower threshold resulted in the same qualitative conclusions being drawn (See Supplementary Information).

The seroprevalence of antibodies against Pgp3 for each region, by age and gender, is summarised in Table 3. The overall seroprevalence in LRR and URR was 26.2% (95% CI=23.5–29.0) and 17.1% (95% CI=14.7–19.9%), respectively. There was a significant difference in overall seroprevalence between the two regions (X2=20.72, p<0.000l). Figure 1 shows the seroprevalence by age group and region. In children 1–9 years old, seroprevalence was 4.4% in LRR and 3.9% in URR. As expected, the prevalence of anti-Pgp3 antibodies increased with age using the non-parametric test for trend (z-score=23.35, p<0.000l). The seroprevalence doubled between 10–19-year-olds and the next oldest age group, 20–29-year-olds, both in the two regions combined, and in each region. Across both regions, in study participants who had no signs of trachoma, seroprevalence was 21% (95% CI = 19–23). Of those who had active trachoma (TF and/or Tl) (n=30), 3% were seropositive (95% CI = 0.1–17) and of those who had scarring trachoma (TS and/or TT and/or CO in either eye; n=122) 56% were seropositive (95% CI = 44–67). 

View this table:
[Table 3.](http://biorxiv.org/content/early/2017/06/14/149237/T3)

Table 3. 
Seroprevalence of anti-Pgp3 antibodies by region, gender and age, Lower River Region and Upper River Region, The Gambi a, 2014.

![Figure 1.](http://biorxiv.org/https://www.biorxiv.org/content/biorxiv/early/2017/06/14/149237/F1.medium.gif)

[Figure 1.](http://biorxiv.org/content/early/2017/06/14/149237/F1)

Figure 1. 
Proportion of participants who were seropositive for anti-Pgp3 antibodies by age group and region, Lower River Region (LRR) and Upper River Region (URR), The Gambia, 2014. Vertical bars indicate 95% CIs.

The prevalence of antibodies was significantly higher in females than in males (z-score=6.384, p<0.000l). The same was true in each region (LRR: z-score=4.881, p<0.000l; URR: z-score=4.114, p<0.000l). When data were considered for each age group, the seropositivity difference between males and females was only significant for 10–19-year-olds (z-score=2.667, p=0.0077) and 30–39- year-olds (z-score=0.2551, p=0.0107). Seroprevalence by gender and age group is shown in Figure 2. 

![Figure 2.](http://biorxiv.org/https://www.biorxiv.org/content/biorxiv/early/2017/06/14/149237/F2.medium.gif)

[Figure 2.](http://biorxiv.org/content/early/2017/06/14/149237/F2)

Figure 2. 
Proportion of participants who were seropositive for anti-Pgp3 antibodies, by age group and gender, Lower River Region and Upper River Region, The Gambia, 2014. Vertical bars indicate 95% confidence intervals.

### Reduction in seroconversion rates over time

We used a reversible catalytic model together with the profile likelihood method to identify reductions in seroconversion rates (SCR) in both LRR and URR. The reversible catalytic model is based on the premise that individuals transit between seropositive and seronegative states with specific average rates22. Abrupt reductions in SCRs were identified in both LRR and URR (Table 5 and Figure 3). These changes in SCR are estimated to have occurred 23 and 19 years before data collection in LRR and URR, respectively (Figure 3). For LRR, the profile likelihood plot (Figure 3) could be cautiously interpreted to indicate that an additional, smaller change in SCR occurred approximately 10 years ago, which coincides with the azithromycin MDA undertaken between 2007 and 20105. For LRR, the SCR would appear to have dropped from an incidence of 0.062 yearly events per person, to 0.010 yearly events per person. These estimates implied a putative 6.2-fold decrease in transmission intensity. In URR, the estimate of the past SCR dropped from 0.050 yearly events per person to 0.008 yearly events per person, a 6.3-fold decrease (Table 5). Figure 4 shows the expected seroprevalence curves as function of age assuming a change in transmission intensity. 

View this table:
[Table 5.](http://biorxiv.org/content/early/2017/06/14/149237/T4)

Table 5. 
Maximum likelihood estimates for the past and current seroconversion and seroreversion rates (SCR and SRR, respectively) associated with data collected from participants in Lower River Region and Upper River Region, The Gambia, 2014 where the respective 95% confidence intervals are shown in brackets. P-values <0.05 are indicative of a change in transmission intensity when comparing two reversible catalytic models, one assuming constant and stable transmission over time and another assuming an abrupt reduction in transmission intensity somewhere in the past.

![Figure 3.](http://biorxiv.org/https://www.biorxiv.org/content/biorxiv/early/2017/06/14/149237/F3.medium.gif)

[Figure 3.](http://biorxiv.org/content/early/2017/06/14/149237/F3)

Figure 3. 
Profile likelihood plots for antibodies against Pgp3 in Lower River Region (LRR) and Upper River Region (URR), The Gambia, 2014. Profile likelihood plots show the log likelihood of a reversible catalytic model allowing for an abrupt change in SCR occurring at iterative years, with the maximum being the time point at which that change is most likely to have occurred. The plot of LRR (left) indicates one major peak 23 years prior to sample collection. The plot of URR (right) suggests a change 19 years prior to sample collection.

### Evidence for elimination

Data from this study show very low prevalence of active trachoma in these regions. In children aged 1-9 years, only 4.2% had detectable antibodies against Pgp3. In both regions, the prevalence of TF in 1-9-year-olds was below the 5% elimination threshold specified by WHO.

## DISCUSSION

This study examines the age-specific prevalence of anti-Pgp3 antibodies in two of The Gambia’s eight regions. One of the regions (LRR) had received azithromycin MDA for trachoma from 2007–2010, whilst the other (URR) had not because the TF prevalence in URR was already below the WHO threshold for elimination as a public health problem. We have demonstrated significant changes in SCR that would be consistent with either (a) a significant drop in the transmission intensity of ocular Ct in the mid to late 1990s or (b) a confounding signal, manifesting as a continuous process of seroconversion, which likely results from exposure to urogenital infection with Ct following sexual debut. It is possible that the data reflect both things, but without current data describing the population prevalence of urogenital and ocular infections, the proportional contributions of STIs and trachoma to the SCR cannot be fully assessed.

The WHO’s guidelines for validation of elimination of trachoma as a public health problem focus programmatic attention on the prevalence of TF in children aged 1–9 years and TT in people aged 15 years and older. The Gambian government is currently compiling evidence for validation of trachoma elimination, following completion of its three-year TF/TT surveillance plan, which began in 201123,24. Previous studies have demonstrated declines in the prevalence of active trachoma in The Gambia both prior to6,25 and in response to the implementation of specific trachoma-control interventions26–28. Increased access to water, education and healthcare in The Gambia during recent decades are thought to have had an impact29, manifesting in a secular decline. Regardless of cause, the prevalence of active trachoma in 0–14-year-olds fell from 10.4% in 198618 to less than 5% in 199617, and has subsequently remained low. Prevalence data from historical studies is presented in Table 6. More recently, six years before the survey reported here, communities in LRR received azithromycin to treat ocular Ct infection, further reducing the prevalence of trachoma in this region5,24, while communities in URR did not. Although not measured in our study, previous work has shown that there has been a reduction in the prevalence of ocular Ct infection in two villages in LRR30,31 with the most recent measurements showing 0.5% infection prevalence in PRET villages, a portion of which are in LRR5. In line with the findings of those previous studies, we provide a further data point showing TF prevalences <5% for each region. The seroprevalence estimates and Fol modelling performed here are measures of trachoma transmission intensity that strongly support The Gambia’s claim to have eliminated trachoma from these areas. 

View this table:
[Table 6.](http://biorxiv.org/content/early/2017/06/14/149237/T5)

Table 6. 
Previously published data on the prevalence active trachoma and trachomatous trichiasis, Lower River Region and Upper River Region, The Gambia, 1986-2013. Data from the 1986 survey was not available for 0-9 year olds, thus we have used the data for 0-14 year olds.

Our analysis of age-specific seroprevalence suggests that the Fol is currently very low, with a substantial decrease in SCR in children having occurred approximately 20 years ago (compared to the children that grew up before then), or a substantial increase in SCR occurring now in those aged around 20 years (compared to their contemporary juniors). The change is too acute to reflect secular decline in the Fol of trachoma. If it is not related to transmission of urogenital Ct, it could be interpreted to reflect the impact of the door-to-door distribution of topical 1% tetracycline ointment across the country, which took place in the late 1990s32.

The antigenic overlap between ocular Ct infection and sexually transmitted infections with the same organism is a potential barrier to making correct inferences about ocular Ct transmission intensity from anti-Pgp3 seroprevalence data. The urogenital Ct consideration is extremely important. No currently available serological test distinguishes between exposure to ocular and urogenital CT infection. The inflections in the SCR curves (Figure 4) could be signals related to seroconversion in response to acquisition of urogenital infection following sexual debut. Interestingly, the seroprevalence among 10–19-year-old Gambian females in our study was almost double that of their male counterparts of the same age. In the Gambia, the median age at which females first have sex is 18.6 years, with 52% of women aged 20–24 years surveyed as part of the Demographic and Health Survey (2013) having had sexual intercourse by age 20; in males the median age is 23.1 years, with 48% of men aged 25–29 years surveyed having had sexual intercourse by age 2233. This could in part explain why we observed a gender difference in seroprevalence in the 10–19-year-old age range, as those with earlier sexual debut would be expected to be more likely to acquire STIs34–36. 

![Figure 4.](http://biorxiv.org/https://www.biorxiv.org/content/biorxiv/early/2017/06/14/149237/F4.medium.gif)

[Figure 4.](http://biorxiv.org/content/early/2017/06/14/149237/F4)

Figure 4. 
Predicted age-seroprevalence curves for antibodies against Pgp3, Lower River Region and Upper River Region, The Gambia, 2014, where the dots represent the observed seroprevalences when the age distribution was broken down in ten-year age groups. Models allowing for a change in transmission intensity at a point in time were fitted.

Conversely, in a population in which the urogenital Ct infection prevalence has been consistently low, it might be expected that anti-Pgp3 serological data more accurately reflect longitudinal trends in ocular Ct transmission. Whilst data suggest that the prevalence of urogenital Ct infections has historically been very low in rural areas of The Gambia37,38, there are no recent data, nor data based on modern molecular testing methods. A 2003 study from Malicounda in Thiès Region of neighbouring Senegal showed the prevalence of urogenital Ct infection there to be 0.3% (n=73)39. Additionally, a systematic review of global estimates of incidence and prevalence of sexually transmitted infections (STI), including urogenital chlamydia, estimated the prevalence of urogenital chlamydia to be 2.9% in low-income countries40 such as The Gambia, although it is noted that this study did not include data from The Gambia.

For the Lower River Region, the profile likelihood plot (Figure 3) could be cautiously interpreted to indicate that a second, smaller change in SCR occurred approximately 10 years ago, a time frame that coincides with the azithromycin MDA undertaken between 2007 and 20105. A similar secondary maximum was not observed in the plot for URR, where MDA was not given. Given the contrast between LRR and URR, it seems plausible that this change in Fol reflects the impact of the NEHP’s azithromycin MDA, which would have reduced the risk of acquiring either ocular41 or urogenital Ct42. Any impact of MDA that could be inferred from the increase in profile likelihood at 7–10 years in LRR is small by comparison to the process of change in FOI that peaked some two decades ago and which probably relates to longer-term health and social improvements in the Gambia.

The confidence intervals associated with SCRs seen in Figure 4 are very broad (which in part reflects the uncertainty of modelling approaches) and comparison between the charts for LRR and URR is indicative but not conclusive of a difference in SCR between the two regions. Although a larger sample size would reduce the uncertainty, interpretation of the model depends to a large extent on the magnitude of the change as well as the timing between the change in SCR and sample collection, as seen in malaria modelling work21. The very large sample sizes required of studies that could delineate SCR changes with high precision could be prohibitive.

A recent serological study using samples from Tanzania9 examined the age-specific seroprevalence of anti-Pgp3 antibodies in a trachoma-endemic community that had received two rounds of high coverage azithromycin MDA. As a result ofthat intervention, the all-ages prevalence of ocular Ct infection had fallen from 9.5% to 0.1% two years after MDA43, and to 0% five years after MDA44, with a corresponding 11-fold decrease in SCR9. This change in infection prevalence occurred in a more defined (and probably narrower) timeframe than the one we have studied in The Gambia. This resulted in a more acute change in SCR, as has been demonstrated in malaria modelling exercises45,46. Previous trachoma modelling studies suggest that an individual may require upwards of 100 lifetime ocular Ct infections in order to develop TT47, so even a modest reduction in transmission may have significant public health implications and reduce the future incidence of TT.

Research is ongoing to address remaining challenges in interpreting trachoma seroprevalence. It is unclear how many infections are required for seroconversion to occur. Studies involving urogenital Ct infection suggest 68% of infected women produce antibodies against Ct48. The intensity of the inflammatory response, and the surface area of inflamed mucosa, however, are both likely to differ between the infected conjunctiva and infected female urogenital tract. Additionally, further work is needed to determine the half-life of Pgp3 antibodies and seroreversion rates. A previous study that examined a high-prevalence community before and after 1 round of azithromycin suggests that individual anti-Pgp3 antibody levels decreased slightly six months after drug treatment, but not enough to be considered seroreversion49. This is similar to results seen for urogenital Ct infection, where anti-Pgp3 antibody titres decreased over 4–7 years, but patients were still in the seropositive range50. Although we have estimated SRR in this study (Table 5), a more accurate estimate could be obtained from a longitudinal study collecting serum samples over a period of years. In low transmission settings, such as post-MDA communities, the SRR may be under-estimated if the assumption of a balance between total number of seropositive and seronegative individuals does not hold true. Further studies to determine SRR are currently underway.

Although dichotomising antibody levels to a simple seropositive/seronegative classification provides a straightforward estimation of seroprevalence, SCR estimates could potentially be improved by using a model based on antibody levels and multiple sampling time points, as suggested by Yman *et al51.* Such models might assume that antibody levels increase with age, as exposure is age-dependent and that transmission intensity can be calculated by measuring the boost in antibody levels. The use of age group-specific geometric mean antibody levels could be explored51, in addition to SCR.

Approximately 4% of 1–9-year-olds were positive for antibodies to Pgp3, which may be due to previously acquired ocular Ct infections and/or to ocular or respiratory Ct infections acquired at birth from mothers with urogenital Ct infections52. This seroprevalence is within the range of prevalence values previously estimated in post-MDA surveys in Tanzania and Nepal 8’9. There was also no observed increase in anti-Pgp3 antibody positivity with age in 1–9-year-olds (Supplementary Table 2), in contrast to what is observed in trachoma-endemic settings, whether treatment-naive 11 or after 3 rounds of MDA 10,15. Focusing on age-specific changes in seropositivity as a measure of cumulative exposure to ocular Ct infection might offset antibody responses from peri-natal infection, as the latter would be expected to be consistent across all ages, or even to decline with increasing age. The data from The Gambia presented here, combined with those from a variety of pre- and post-MDA settings, contribute to an understanding of the potential use of antibody-based surveillance of children to ensure a lack of infection recrudescence.

Data from paired pre- and post-MDA surveys would substantially improve modeling efforts to describe changes over time in. The inclusion of infection data for both ocular and urogenital Ct infection are needed to further clarify how urogenital Ct contributes to observed seropositivity rates.

## METHODS

### Ethical Review

This study was conducted in accordance with the Declaration of Helsinki. It received approval from the London School of Hygiene & Tropical Medicine Ethics Committee (LSHTM; reference 7059) and The Gambia Government/Medical Research Council Joint Ethics Committee (SCC1408v2). CDC investigators did not engage with study participants.

### Survey methodology

We conducted a population-based, cluster-random-sampled survey was conducted in February-March 2014. The Gambia is divided into geographically-defined census Enumeration Areas (EAs) of approximately 600-800 people each. Sampling by EA is equivalent to sampling settlements with probability proportional to their size53. Twenty EAs in each of URR and LRR were randomly selected for participation. Trained field workers sensitised villagers and obtained verbal community-consent from each village chief *(alkalo).* Field workers and *alkalos* compiled a list of households for each EA, from which households were randomly selected for census and recruitment. The study and consent form were explained to the head of each selected household and prospective participants. All members of selected households were invited to participate, regardless of age. Written (thumbprint or signature) consent was obtained from each participant aged ≥18 years, while a parent or guardian provided written consent for each participating child aged under 18 years. Children aged 12-17 years provided assent before participating.

The trachoma graders were experienced in field grading for active trachoma and had received regular training according to PRET5 and Global Trachoma Mapping Project (GTMP) protocols54,55. After informed consent was received, the grader examined both eyes of the subject using a binocular loupe (2.5x) and a torch. The grader changed gloves between each participant to minimise the risk of carry-over infection. In accordance with the Gambian NEHP policy, antibiotics were provided to individuals with evidence of active trachoma and to residents of their household.

Each participant had a finger-prick blood sample collected onto filter paper (Trop-Bio, Townsville, Australia) using a sterile single-use lancet (BD Microtrainer, Dublin, Ireland). Each filter paper had six extensions, calibrated to absorb 10 μL of blood each. Samples were air-dried for approximately five hours and then placed in individual Whirl-Pak plastic bags (Nasco, Modesto, California) which were stored with desiccant sachets (Whatman, Little Chalfont, UK) at −20°C. All samples were shipped to LSHTM for testing.

### ELISA Assay

Dried blood spots (DBS) were tested for antibodies against Pgp3 according to the method previously described15. Briefly, serum was eluted from dried blots spots then applied to a plate coated with Pgp3 protein10; known standards were included and assayed in triplicate on each plate. Following incubation, bound antibody was detected with HRP-labelled mouse anti-human lgG(Fc)-HRP (Southern Biotech, Birmingham, USA). Plates were incubated and washed, and then TMB (KPL, Gaithersburg, USA) was added to develop the plates. The reaction was stopped with 1N H2S04 and optical density was read at 450 nm (OD450) on a Spectramax M3 plate reader (Molecular Devices, Wokingham UK). Readings were corrected for background by subtracting the average absorbance of three blank wells containing no serum, using Softmax Pro5 software (Molecular Devices).

### Statistical Methods

Blanked OD450 values for samples were normalised against the 200 U standard included on each plate15. Laboratory work was undertaken masked to demographic and clinical information. Statistical analyses were carried out using R56. Using the “survey” package and assuming a design effect of 2.6554, the 95% confidence intervals (CI) were calculated using the Clopper-Pearson interval57. Wilcoxon-Mann-Whitney z-scores58 were calculated to compare the proportion seropositive between different regions, ages, and genders. The non-parametric test for trend was used to measure the increase in prevalence of anti-Pgp3 antibodies with age.

A finite mixture model21 was used to classify the samples as seropositive or seronegative based on normalised OD450 values. The data were fitted using maximum likelihood methods, estimating the distribution parameters for each classification group (seropositive or seronegative) as well as the proportion of samples in each category to fit the overall distribution of results59. To ensure that the assay had high specificity, the threshold for seropositivity was set using the mean of the Gaussian distribution of the seronegative population plus four standard deviations (the quantile inclusive of 99.994%) of the seronegative population21,59.

Population age groups were categorized according to known time points of changes in disease prevalence. The youngest age group was 1–9-year-olds, who in LRR are likely to have been born during or after MDA with azithromycin. The oldest group included people aged 40 years and above, who experienced secular declines in trachoma prevalence prior to 1986, the year of the first National Survey of Blindness and Low Vision18. Participants aged between 10 and 39 years were grouped into 10-year categories.

Maximum likelihood methods were used to fit a simple reversible catalytic model (assuming a constant transmission intensity over time, and that people enter into the model as seronegative) to the seropositivity data22. The model is described as function of SCR (the mean rate by which seronegative individuals become seropositive upon disease exposure) and a SRR (the mean rate by which seropositive individuals revert to a seronegative status in the absence of re-infection). SCR is usually interpreted as a proxy of the Fol for each region. As there are no long-term data regarding the SRR for Ct, we allowed it to vary as determined by the model. Age of each individual was rounded to the nearest integer in order to generate age-specific seroprevalence curves for each region. A second set of curves was generated using a model allowing for a sudden change in SCR, as would be seen if there were a change in the Fol, due to an intervention, for example. Both models assume the age distribution of the populations match the typical age distribution of Africa. Wilks’ likelihood ratio test was used to compare the single-Fol model to the model allowing a change in SCR at some point in time. P-values<0.05 indicated that the latter model described the data better than the single-Fol model. Profile likelihood plots were generated to determine the most likely timing of the change in Fol as described elsewhere21.

## AUTHOR CONTRIBUTIONS

S.J.M., R.B., D.C.W.M., A.W.S. and C.h.R designed the study and wrote the paper, D.L.M., G.C., S.G. C.h.R. and S.J.M. developed the protocol for serological analysis, S.J.M. performed serological analysis, S.J.M., C.h.R and N.S. performed the statistical analysis, R.B. trained the trachoma graders, S.J.M., H.J., P.M., R.B., S.E.B., D.C.W.M. and C.h.R. organized field data collection. H.J. and P.M. performed the field grading of trachoma. H.P. provided critical feedback on the paper. All authors reviewed and approved the manuscript.

## ADDITIONAL INFORMATION

Sarah Gwyn is employed by the commercial company IHRC, Inc. and is a contractor at the Centres for Disease Control and Prevention.

## COMPETING FINANCIAL INTERESTS

The authors declare no competing financial interests.

## ACKNOWLEDGMENTS

We are very grateful to the communities that generously gave their time to participate in this study. We thank the field team and public health staff in The Gambia, Ms PJ Hooper from the International Trachoma Initiative, and Prof Hannah Faal from the University of Calabar and Teaching Hospital. NS is partially supported by the Portuguese Fundação para a Ciência e Tecnologia (Grant # UID/MAT/00006/2013).

The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.

*   Received June 12, 2017.
*   Revision received June 12, 2017.
*   Accepted June 14, 2017.


*   © 2017, Posted by Cold Spring Harbor Laboratory

This pre-print is available under a Creative Commons License (Attribution 4.0 International), CC BY 4.0, as described at [http://creativecommons.org/licenses/by/4.0/](http://creativecommons.org/licenses/by/4.0/)

## REFERENCES

1.  1.Bourne, R. R. A. et al. Causes of vision loss worldwide, 1990-2010: A systematic analysis. Lancet Glob. Heal. 1, 339–349 (2013).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1016/S2214-109X(13)70113-X&link_type=DOI) 

2.  2.World Health Organization Alliance for the Global Elimination of Trachoma by 2020. Eliminating trachoma: accelerating towards 2020. (2016). at <[http://www.trachomacoalition.org/GET2020/](http://www.trachomacoalition.org/GET2020/)>
    
    

3.  3.World Health Organization. Validation of elimination of trachoma as a public health problem (WHO/HTM/NTD/2016.8). (2016). at <[http://apps.who.int/iris/bitstream/10665/20890l/l/WHO-HTM-NTD-2016.8-eng.pdf?ua=1](http://apps.who.int/iris/bitstream/10665/20890l/l/WHO-HTM-NTD-2016.8-eng.pdf?ua=1)>
    
    

4.  4.WHO. Weekly epidemiological record Relevé épidémiologique hebdomadaire. World Heal. Organ. Wkly. Epidemiol. Rec. 96, 421–428 (2014).
    
    

5.  5.Harding-Esch, E. M. et al. Mass Treatment with Azithromycin for Trachoma: When Is One Round Enough? Results from the PRET Trial in The Gambia. PLoS Negl. Trop. Dis. 7, e2115 (2013).
    
    

6.  6.Harding-Esch, E. M. et al. Active trachoma and ocular Chlamydia trachomatis infection in two Gambian regions: on course for elimination by 2020? PLoS Negl. Trop. Dis. 3, e573 (2009).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1371/journal.pntd.0000573&link_type=DOI) 
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=20027217&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 

7.  7.Wang, S. & Grayston, J. T. Human seroloy in Chlamydia trachomatis: Infection with Microimmunofluorescence. J. Infect. Dis. 130, 388–397 (1974).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1093/infdis/130.4.388&link_type=DOI) 
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=4613759&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 
    
    [Web of Science](http://biorxiv.org/lookup/external-ref?access_num=A1974U610700010&link_type=ISI) 

8.  8.West, S. K. et al. Can We Use Antibodies to Chlamydia trachomatis as a Surveillance Tool for National Trachoma Control Programs * ? Results from a District Survey. PLoS Negl. Trop. Dis. 10, 1–11 (2016).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1371/journal.pntd.0004412&link_type=DOI) 

9.  9.Martin, D. L. et al. Serology for trachoma surveillance after cessation of mass drug administration. PLoS Negl. Trop. Dis. 9, e0003555 (2015).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1371/journal.pntd.0003555&link_type=DOI) 
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=25714363&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 

10. 10.Goodhew, E. B. et al. CT694 and pgp3 as Serological Tools for Monitoring Trachoma Programs. PLoS Negl. Trop. Dis. 6, e1873 (2012).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1371/journal.pntd.0001873&link_type=DOI) 
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=23133684&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 

11. 11.Martin, D. L. et al. Serological Measures of Trachoma Transmission Intensity. Sei. Rep. 5, 18532 (2015).
    
    

12. 12.Woodhall, S. C. et al. Can we use postal surveys with anonymous testing to monitor chlamydia prevalence in young women in England? Pilot study incorporating randomised controlled trial of recruitment methods. Sex. Transm. Infect. 91, 412–4 (2015).
    
    [Abstract/FREE Full Text](http://biorxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6ODoic2V4dHJhbnMiO3M6NToicmVzaWQiO3M6ODoiOTEvNi80MTIiO3M6NDoiYXRvbSI7czozNzoiL2Jpb3J4aXYvZWFybHkvMjAxNy8wNi8xNC8xNDkyMzcuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 

13. 13.Wills, G. S. et al. Pgp3 antibody enzyme-linked immunosorbent assay, a sensitive and specific assay for seroepidemiological analysis of Chlamydia trachomatis infection. Clin. Vaccine Immunol. 16, 835–43 (2009).
    
    [Abstract/FREE Full Text](http://biorxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoiY2RsaSI7czo1OiJyZXNpZCI7czo4OiIxNi82LzgzNSI7czo0OiJhdG9tIjtzOjM3OiIvYmlvcnhpdi9lYXJseS8yMDE3LzA2LzE0LzE0OTIzNy5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 

14. 14.Wang, J. et al. A genome-wide profiling of the humoral immune response to Chlamydia trachomatis infection reveals vaccine candidate antigens expressed in humans. J. Immunol. 185, 1670–80 (2010).
    
    [Abstract/FREE Full Text](http://biorxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6ODoiamltbXVub2wiO3M6NToicmVzaWQiO3M6MTA6IjE4NS8zLzE2NzAiO3M6NDoiYXRvbSI7czozNzoiL2Jpb3J4aXYvZWFybHkvMjAxNy8wNi8xNC8xNDkyMzcuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 

15. 15.Migchelsen, S. J. et al. Defining Seropositivity Thresholds for Use in Trachoma Elimination Studies. PLoS Negl. Trop. Dis. 11, e0005230 (2017).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1371/journal.pntd.0005230&link_type=DOI) 

16. 16.Cocks, N. et al. Community seroprevalence survey for yaws and trachoma in the Western Division of Fiji. Trans R Soc Trop Med Hyg 1–6 (2016). doi:10.1093/trstmh/trw069
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1093/trstmh/trw069&link_type=DOI) 
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=27852877&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 

17. 17.Faal, H. et al. Evaluation of a national eye care programme: re-survey after 10 years. Br. J. Ophthalmol. 84, 948–51 (2000).
    
    [Abstract/FREE Full Text](http://biorxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6MTI6ImJqb3BodGhhbG1vbCI7czo1OiJyZXNpZCI7czo4OiI4NC85Lzk0OCI7czo0OiJhdG9tIjtzOjM3OiIvYmlvcnhpdi9lYXJseS8yMDE3LzA2LzE0LzE0OTIzNy5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 

18. 18.Dolin, P. J. et al. Trachoma in The Gambia. Br. J. Ophthalmol. 82, 930–3 (1998).
    
    [Abstract/FREE Full Text](http://biorxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6MTI6ImJqb3BodGhhbG1vbCI7czo1OiJyZXNpZCI7czo4OiI4Mi84LzkzMCI7czo0OiJhdG9tIjtzOjM3OiIvYmlvcnhpdi9lYXJseS8yMDE3LzA2LzE0LzE0OTIzNy5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 

19. 19.United Nations Statistics Division. UN: a world of information. (2016). at <[http://data.un.org/](http://data.un.org/)>
    
    

20. 20.Corran, P. H. et al. Dried blood spots as a source of anti-malarial antibodies for epidemiological studies. Malar. J. 7, 195 (2008).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1186/1475-2875-7-195&link_type=DOI) 
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=18826573&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 

21. 21.Sepúlveda, N., Stresman, G., White, M. T. & Drakeley, C. J. Current mathematical models for analyzing anti-malarial antibody data with an eye to malaria elimination and eradication. J. Immunol. Res. 2015, (2015).
    
    

22. 22.Drakeley, C. J. et al. Estimating medium- and long-term trends in malaria transmission by using serological markers of malaria exposure. Proc. Natl. Acad. Sci. U. S. A. 102, 5 108–13 (2005).
    
    

23. 23.World Health Organization. Report of the 17th meeting of the WHO alliance for the global elimination of blinding trachoma. (2013). at <[http://www.who.int/blindness/publications/GET17Report_final.pdf?ua=1](http://www.who.int/blindness/publications/GET17Report_final.pdf?ua=1)>
    
    

24. 24.Burr, S. E. et al. Cross-Sectional Surveys of the Prevalence of Follicular Trachoma and Trichiasis in The Gambia: Has Elimination Been Reached? PLoS Negl Trop Dis 10, (2016).
    
    

25. 25.Faal, H., Minassian, D., Sowa, S. & Foster, A. National survey of blindness and low vision in The Gambia: results. Br. J. Ophthalmol. 73, 82–7 (1989).
    
    [Abstract/FREE Full Text](http://biorxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6MTI6ImJqb3BodGhhbG1vbCI7czo1OiJyZXNpZCI7czo3OiI3My8yLzgyIjtzOjQ6ImF0b20iO3M6Mzc6Ii9iaW9yeGl2L2Vhcmx5LzIwMTcvMDYvMTQvMTQ5MjM3LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 

26. 26.Harding-Esch, E. M. et al. Trachoma prevalence and associated risk factors in the Gambia and Tanzania: baseline results of a cluster randomised controlled trial. PLoS Negl. Trop. Dis. 4, e861 (2010).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1371/journal.pntd.0000861&link_type=DOI) 
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=21072224&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 

27. 27.Chen, C. et al. Incremental cost of conducting population-based prevalence surveys for a neglected tropical disease: the example of trachoma in 8 national programs. PLoS Negl. Trop. Dis. 5, e979 (2011).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1371/journal.pntd.0000979&link_type=DOI) 
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=21408130&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 

28. 28.Burton, M. J. et al. Profound and sustained reduction in Chlamydia trachomatis in The Gambia: a five-year longitudinal study of trachoma endemic communities. PLoS Negl. Trop. Dis. 4, (2010).
    
    

29. 29.Dolin, P. J. et al. Reduction of trachoma in a sub-Saharan village in absence of a disease control programme. Lancet 349, 1511–2 (1997).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1016/S0140-6736(97)01355-X&link_type=DOI) 
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=9167460&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 
    
    [Web of Science](http://biorxiv.org/lookup/external-ref?access_num=A1997XA90100010&link_type=ISI) 

30. 30.Bailey, R. L. et al. Polymerase Chain Reaction For The Detection Of Ocular Chlamydial Infection In Trachoma-Endemic Communities. J. Infect. Dis. 170, 709–712 (1994).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1093/infdis/170.3.709&link_type=DOI) 
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=8077735&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 
    
    [Web of Science](http://biorxiv.org/lookup/external-ref?access_num=A1994PE26500035&link_type=ISI) 

31. 31.Mabey, D. C., Bailey, R. L, Ward, M. E. & Whittle, H. C. A longitudinal study of trachoma in a Gambian village: implications concerning the pathogenesis of chlamydial infection. Epidemiol. Infect. 108, 343–51 (1992).
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=1582475&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 

32. 32.Sillah, A. Gambia Country Plan for SAFE with Zithromax. (2006).
    
    

33. 33.The Gambia Bureau of Statistics (GBOS) & ICF International. The Gambia Demographic and Health Survey. (2014). at <[http://dhsprogram.com/pubs/pdf/FR289/FR289.pdf](http://dhsprogram.com/pubs/pdf/FR289/FR289.pdf)>
    
    

34. 34.Speizer, I. S., Christophe Fotso, J., Davis, J. T., Saad, A. & Otai, J. Timing and Circumstances of First Sex among Female and Male Youth from Select Urban Areas of Nigeria, Kenya, and Senegal. J Adolesc Heal. 53, 609–616 (2013).
    
    

35. 35.Ghebremichael, M., Larsen, U. & Paintsil, E. Association of age at first sex with HIV-1, HSV-2, and other sexual transmitted infections among women in northern Tanzania. *Sex.* Transm. Dis. 36, 570–6 (2009).
    
    

36. 36.Sauvain-Dugerdil, C. et al. The start of the sexual transition in Mali: risks and opportunities. Stud. Fam. Plann. 39, 263–80 (2008).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1111/j.1728-4465.2008.00174.x&link_type=DOI) 
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=19248714&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 

37. 37.Mabey, D. C. W. & Whittle, H. C. Genital and neonatal chlamydial infection in a trachoma endemic area. Lancet 320, 300-301 (1982).
    
    

38. 38.Walraven, G. et al. The burden of reproductive-organ disease in rural women in The Gambia, West Africa. Lancet (London, England) 357, 1161–1167 (2001).
    
    

39. 39.Gueye Ndiaye, A. et al. [Screening for HIV, syphilis, Chlamydia trachomatis and Neisseria gonorrhoreae during a combined survey conducted in Malicouna, a Senegalese rural area]. Bull Soc Pathol Exot 102, 150–154 (2009).
    
    

40. 40.Newman, L., Rowley, J., Hoorn, S. Vander & Wijesooriya, N. S. Global Estimates of the Prevalence and Incidence of Four Curable Sexually Transmitted Infections in 2012 Based on Systematic Review and Global Reporting. PloS ONEone 10, 1–17 (2015).
    
    

41. 41.Evans, J. R. & Solomon, A. W. Antibiotics for trachoma. Cochrane database Syst. Rev. CD001860 (2011). doi:10.1002/14651858.CD001860.pub3
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1002/14651858.CD001860.pub3&link_type=DOI) 

42. 42.Marks, M. et al. Mass drug administration of azithromycin for trachoma reduces the prevalence of genital Chlamydia trachomatis infection in the Solomon Islands. Sex. Transm. Infect. 92, 261–5 (2016).
    
    [Abstract/FREE Full Text](http://biorxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6ODoic2V4dHJhbnMiO3M6NToicmVzaWQiO3M6ODoiOTIvNC8yNjEiO3M6NDoiYXRvbSI7czozNzoiL2Jpb3J4aXYvZWFybHkvMjAxNy8wNi8xNC8xNDkyMzcuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 

43. 43.Solomon, A. W. et al. Mass treatment with single-dose azithromycin for trachoma. N. Engl. J. Med. 351, 1962–71 (2004).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1056/NEJMoa040979&link_type=DOI) 
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=15525721&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 
    
    [Web of Science](http://biorxiv.org/lookup/external-ref?access_num=000224851300008&link_type=ISI) 

44. 44.Solomon, A. W. et al. Two doses of azithromycin to eliminate trachoma in a Tanzanian community. N. Engl. J. Med. 358, 1870–1 (2008).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1056/NEJMc0706263&link_type=DOI) 
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=18434662&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 
    
    [Web of Science](http://biorxiv.org/lookup/external-ref?access_num=000255209200037&link_type=ISI) 

45. 45.Sepúlveda, N., Paulino, C. D. & Drakeley, C. Sample size and power calculations for detecting changes in malaria transmission using antibody seroconversion rate. Malar. J. 14, 1–14 (2015).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1186/1475-2875-14-1&link_type=DOI) 
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=25557741&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 

46. 46.van den Hoogen, L. L. et al. Serology describes a profile of declining malaria transmission in Farafenni, The Gambia. Malar. J. 14, 416 (2015).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1186/s12936-015-0939-1&link_type=DOI) 

47. 47.Gambhir, M. et al. The development of an age-structured model for trachoma transmission dynamics, pathogenesis and control. PLoS Negl. Trop. Dis. 3, e462 (2009).
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=19529762&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 

48. 48.Horner, P. J. et al. Effect of time since exposure to Chlamydia trachomatis on chlamydia antibody detection in women: a cross-sectional study. Sex. Transm. Infect. 89, 398–403 (2013).
    
    [Abstract/FREE Full Text](http://biorxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6ODoic2V4dHJhbnMiO3M6NToicmVzaWQiO3M6ODoiODkvNS8zOTgiO3M6NDoiYXRvbSI7czozNzoiL2Jpb3J4aXYvZWFybHkvMjAxNy8wNi8xNC8xNDkyMzcuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 

49. 49.Goodhew, E. B. et al. Longitudinal analysis of antibody responses to trachoma antigens before and after mass drug administration. BMC Infect. Dis. 14, 216 (2014).
    
    

50. 50.Gijsen, A. P., Land, J. a, Goossens, V. J., Slobbe, M. E. P. & Bruggeman, C. a. Chlamydia antibody testing in screening for tubal factor subfertility: the significance of IgG antibody decline over time. Hum. Reprod. 17, 699–703 (2002).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1093/humrep/17.3.699&link_type=DOI) 
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=11870123&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 
    
    [Web of Science](http://biorxiv.org/lookup/external-ref?access_num=000174424600028&link_type=ISI) 

51. 51.Yman, V. et al. Antibody acquisition models: A new tool for serological surveillance of malaria transmission intensity. Sci. Rep. 6, 19472 (2016).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1038/srep19472&link_type=DOI) 

52. 52.Hobson, D. & Rees, E. Maternal genital chlamydial infection as a cause of neonatal conjunctivitis. Postgrad.Med. J. 53, 595–597 (1977).
    
    

53. 53.Harding-Esch, E. et al. Costs of Testing for Ocular Chlamydia trachomatis Infection Compared to Mass Drug Administration for Trachoma in The Gambia: Application of Results from the PRET Study. PLoS Negl. Trop. Dis. 9, e0003670 (2015).
    
    

54. 54.Solomon, A. W. et al. The Global Trachoma Mapping Project: Methodology of a 34-Country Population-Based Study. Ophthalmic Epidemiol. 22, 214–25 (2015).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.3109/09286586.2015.1037401&link_type=DOI) 
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=26158580&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 

55. 55.Courtright, P. et al. Global Trachoma Mapping Project Training for mapping of trachoma, 3rd edition. (2015). at <[http://www.trachomacoalition.org/resources/global-trachoma-mapping-project-training-mapping-trachoma](http://www.trachomacoalition.org/resources/global-trachoma-mapping-project-training-mapping-trachoma)>
    
    

56. 56.R Development Core Team, R Core Team & R Development Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing (R Foundation for Statistical Computing, 2008). at <[http://www.r-project.org](http://www.r-project.org).>
    
    

57. 57.Clopper, C. J. & Pearson, E. S. The Use of Confidence or Fiducial Limits Illustrated in the Case of the Binomial. Biometrika 26, 404–413 (1934).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1093/biomet/26.4.404&link_type=DOI) 

58. 58.Fay, M. P. & Proschan, M. A. Wilcoxon-Mann-Whitney or t-test? On assumptions for hypothesis tests and multiple interpretations of decision rules. Stat. Surv. 4, 1–39 (2010).
    
    [CrossRef](http://biorxiv.org/lookup/external-ref?access_num=10.1214/09-SS051&link_type=DOI) 
    
    [PubMed](http://biorxiv.org/lookup/external-ref?access_num=20414472&link_type=MED&atom=%2Fbiorxiv%2Fearly%2F2017%2F06%2F14%2F149237.atom) 

59. 59.Kaneko, A. et al. Characteristic age distribution of Plasmodium vivax infections after malaria elimination on Aneityum Island, Vanuatu. Infect. Immun. 82, 243–52 (2014).
    
    [Abstract/FREE Full Text](http://biorxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6MzoiaWFpIjtzOjU6InJlc2lkIjtzOjg6IjgyLzEvMjQzIjtzOjQ6ImF0b20iO3M6Mzc6Ii9iaW9yeGl2L2Vhcmx5LzIwMTcvMDYvMTQvMTQ5MjM3LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==)