Serum samples from US Marines 18–45 years of age who served >180 days in Afghanistan during 2001–2010 were obtained from the US Department of Defense Serum Repository (DoDSR). Documentation of prior exposure to Q fever or rickettsioses and sample volume <0.5 mL were exclusion criteria. We selected the most recent 1,000 postdeployment specimens that fit the inclusion criteria for our study.

Military service members have blood drawn every 2 years for HIV testing and during postdeployment screenings, and a portion of each sample is stored at the DoDSR. All predeployment samples were collected <1 year before the start of Afghanistan service, and postdeployment samples were collected within ≈1 year after the end of Afghanistan service. Paired predeployment and postdeployment samples were irreversibly stripped of personal identifiers by DoDSR and labeled with an internal DoDSR code. Only DoDSR has access to the key linking the code to personal identifiers. The Naval Medical Research Center Institutional Review Board approved the study.

We screened specimens for antibodies against TGR, SFGR, and STGO by ELISA, as described (11) and determined titers for positive specimens. Seroconversion (i.e., titer <100 in the predeployment sample and titer >400 in the paired postdeployment sample) or a 4-fold rise in titer between the predeployment and paired postdeployment sample was used to determine acute infection.

We tested specimens for antibodies against C. burnetii by using the Q fever Immunodot assay (GenBio, San Diego, CA, USA), according to the manufacturer’s instructions. Specimens were considered positive for acute infection if antibodies against phase I and phase II antigen were present. Seroconversion was defined by the presence of antibodies against phase I antigen in the postdeployment sample but not in the predeployment sample.

Of the 1,000 postdeployment serum samples screened for SFGR, TGR, and STGO, only 879 were screened for Q fever because the volume of 121 samples was depleted from earlier testing. The screening assays showed that 87, 18, and 1 samples were positive for antibodies against C. burnetii, SFGR, and STGO, respectively, before deployment. No antibodies against TGR were detectable in any samples. Seroconversions to C. burnetii and SFGR occurred in 3.4% and 0.5% of the paired serum samples, respectively (Table). C. burnetii infection was most prevalent among the agents tested both before deployment (n = 87) and during deployment (n = 30). Of 879 specimens, 117 (13.3%) were positive.

Figure

Figure. Serum samples assessed for evidence of seroconversion for Coxiella burnetii among US Marines deployed to Afghanistan, by year deployment began, 2001–2010. A) Number of Coxiella burnetii seroconversions (n = 30). B)...

Most of the 30 C. burnetii seroconversions occurred in Marines who began deployment in 2008 (n = 12) and 2009 (n = 8) (Figure, panel A). However, most (634/879 [72.1%]) of the sample population were deployed during this period (Figure, panel B), so higher rates for these years likely do not indicate higher risk. With the exception of 1 fixed-wing pilot, all C. burnetii and SFGR seroconversions occurred in general infantrymen, who represented 93.1% of C. burnetii–positive samples and 92.6% of total samples.

Range of deployment time was 180–450 days, and average length of deployment was 271 (SD 104.49) days. Average length of deployment for those who seroconverted for C. burnetii or SFGR was 286.2 (SD 112.01) and 342.4 (SD 131.65) days, respectively. Analysis by t-test showed no correlation between length of deployment and seroconversion for either C. burnetii (p = 0.98) or SFGR (p = 1.0).

Predeployment serum samples were collected 3–364 (mean 183, SD 111.07) days before deployment, and postdeployment samples were collected <443 (mean 61.3, SD 111.02) days after return. Samples showing seroconversions for C. burnetii and SFGR were collected an average of 179.3 (SD 93.94) and 151.6 (SD 126.76) days, respectively, predeployment and 76.9 (SD 87.64) and 129.8 (SD 127.85) days, respectively, postdeployment.

Acute febrile illness can be difficult to diagnose because many infections have similar symptoms and signs and are difficult to differentiate without appropriate diagnostic tools. The rate of C. burnetii seroconversions (3.4%) in Marines in our study is similar to that reported in United Kingdom military personnel deployed to Afghanistan during 2008–2011 (7); the study from the United Kingdom also confirms the presence of rickettsiae in the region. The rate of C. burnetii seropositivity in the United States is ≈3% but ranges dramatically; nearby Nova Scotia, Canada, has rates >14% (8,12). With proper treatment of infections, the case-fatality rate for Q fever and rickettsial infections is <2%. However, rates can be >30%; among hospitalized Mediterranean spotted fever patients in Portugal in 1997, the case-fatality rate was 32.5% (13).

In our study, gaps between dates of blood draws and start and end dates of deployment provided an opportunity for exposure to agents of Q fever and SFGR outside the deployment period. The nature of retrospective, blinded studies prevents follow-up with participants to ascertain additional information, including whether symptoms developed or treatment was sought. This limitation is especially true for studies of military populations, for whom routine blood draws are typically used, rather than samples being specifically collected for research studies. However, results from these studies provide valuable information, and in our study, this limitation does not discount the documented prevalence of Q fever in Iraq and Afghanistan (4–10,14) or the inherent risk to deployed US military personnel. Our results highlight the risk of contracting Q fever and the need for Q fever diagnostics in military engagements in Central Asia.