Independent attempts to confirm the results of Nicolson and colleagues have been unsuccessful. Gray et al. (1999) studied serum from symptomatic and asymptomatic Gulf War veterans who had given prewar and postwar blood samples. They used immunoblot banding for M. fermentans at the University of Alabama Diagnostic Mycoplasma Laboratory, and none of the banding profiles was associated with reported symptoms in veterans. The study revealed that 10.9% of Gulf War veterans and 9.3% of nondeployed veterans who served in the military during the same period as the Gulf War veterans had prewar antibodies to M. fermentans. Of those without pre-existing antibodies, 19.2% of Gulf War veterans and 13.7% of nondeployed veterans developed serologic evidence of new M. fermentans infections.
A matched case-control study was conducted to determine the prevalence of M. fermentans antibodies in military personnel before and after Gulf War deployment and seroconversion rates in veterans with and without complaints of GWI (Lo et al. 2000). The study found a predeployment prevalence of M. fermentans-specific antibodies of 4.8% in veterans with GWI and 5.2% in controls; no difference in rates of seroconversion (1.1% in GWI cases and 1.2% in controls) during deployment was found. Lo et al. noted that their serologic test has been shown to be highly sensitive and specific and that most patients, including immunocompromised patients with AIDS, produce detectable species-specific antibodies to M. fermentans. That specificity suggests that the results are not an artifact of intracellular infections that do not yield antibody responses. Lo et al. also noted that “it is difficult to assess the validity and specificity of the NGT [nuclear gene tracking] testing [of Nicolson and colleagues, as discussed above], as there is no precedent for identifying any viral, mycoplasmal, or bacterial infection in clinical specimens using this uncommon technique.”
A report prepared for the US Senate Committee on Veterans Affairs Special Investigation Unit on Persian Gulf War Illness stated in part that “M. fermentans has been at times suspected of causing various diseases in humans and, therefore, the center of some controversy” but that “this organism is considered to be a member of the normal human flora” (Dybvig, 1998). Dybvig noted that the NGT method used by Nicolson and colleagues was “an inappropriate diagnostic method for detection of M. fermentans” and that neither the specificity nor the sensitivity of the test had been established. He noted further that “M. fermentans DNA resides within the mycoplasmal cell and would not be present in the material assayed by this procedure, namely, host nucleoprotein.” Dybvig also wrote that genetic engineering of M. fermentans was not technically feasible at the time of his report and certainly did not occur before the Gulf War.
Because of the conflicting data related to M. fermentans infections and their possible association with GWI and the suggestion of possible benefits of treatment with doxycycline, the VA conducted a randomized placebo-controlled trial to determine whether doxycycline given at 200 mg/day for 12 months could improve functional status of persons with GWI (Donta et al. 2004). In the trial, 491 deployed Gulf War veterans with GWI and detectable mycoplasma DNA in the blood were randomized to receive either doxycycline or a placebo for 12 months. Of the participants, 324 (66%) had an M. fermentans infection, 197 (40.1%) had an M. genitalium infection, and 53 (10.8%) had an M. pneumoniae infection, either singly or in combination as detected with PCR testing. Although a higher fraction of doxycycline participants than controls showed improvement at 3 months (21.5% vs 9.9%), there was no statistically significant difference at 9, 12, and 18 months. Overall, there was no statistically significant difference between the doxycycline-treated and placebo groups on the primary outcome measures. There was a trend toward fewer unscheduled clinic visits and hospitalizations among doxycycline-treated veterans than placebo subjects, but this was not related to the presence of mycoplasma