Lim et al. (2008) studied exposure to BFRs using the U.S. National Health and Nutrition Examimation Survey (NHANES) dataset, a U.S.-wide survey of chemical exposures, nutrition, and health conditions. They found that two of six (PBB-153 and PBDE-153) were associated with both diabetes (the type of diabetes was not specified) and metabolic syndrome (a cluster of conditions that are common in people with type 2 diabetes, as well as type 1). The two chemicals had different dose response curves. PBB-153 was positively associated with diabetes, in that the risk increased as the levels of exposure increased. PBDE-153 was also associated with diabetes, but in the highest exposure group, the trend decreased slightly. Similar trends were seen in the associations with metabolic syndrome: increasing for PBB-153, and increasing then decreasing for PBDE-153. These unusual dose-response relationships have been seen in animal studies of endocrine disrupting compounds.

In a study from Finland, where PBDE levels are lower, however, researchers found that PBDE-47 and PBDE-153 were not associated with type 2 diabetes in elderly adults (Airaksinen et al. 2011).

A long-term study from Michigan found that levels of PBBs were not associated with diabetes risk (although PCB levels were). The people in this study were exposed to very high levels of PBBs for about eight months during the 1970s (but not unusually high levels of PCBs) (Vasiliu et al. 2006). It may be that the effects of PBBs show up at lower doses, and at higher exposure levels, the risk would level off or even decrease, thus explaining why an association was found between diabetes and PBBs in people with lower levels of exposure, but not higher levels (Lim et al. 2008). 

There have been no studies on type 1 diabetes and PBDEs, although I have met at least one person who found his child had high levels of PBDEs in blood after diagnosis with type 1. In Sweden, PBDEs were banned in the 1990s, and by the end of that decade, levels in Swedish women's breastmilk began to decline. (Overall, PBDE levels in the US are much higher than in Sweden). Curiously, Swedish children born beginning in the year 2000 show slightly lower risks of type 1 diabetes than those born earlier. (The incidence is still increasing in Swedish children, but not as rapidly). I wonder if these two trends could be related? (Howard 2011)

Chronic exposure to the PBDE known as penta-BDE can disturb glucose and insulin metabolism in fatty tissue of rats, characteristics associated with type 2 diabetes, insulin resistance, and obesity. These effects suggest that PBDE exposure should be examined as possible contributors to the obesity epidemic (Hoppe and Carey 2007).

PBDEs are endocrine disrupting compounds and accumulate in fatty tissue (Hoppe and Carey 2007). PBDEs can also affect the immune system, although to my knowledge have not been studied in relation to autoimmunity. One study found that mice exposed to PBDEs had lower levels of the immune cells (cytokines) important in the defense of coxsackie virus (Lundgren et al. 2009). (Coxsackie virus is one of the viruses linked to type 1 diabetes; see the viruses page).

Exposure to two types of brominated flame retardants are associated with diabetes and metabolic syndrome in people with lower, background levels of exposure.