A comprehensive review of nearly 130 studies shows the strongest evidence to date that “forever chemicals” are associated with gestational diabetes, while underscoring uncertainty around other diabetes outcomes and the need for more prospective research.
Study: Associations of perfluoroalkyl and polyfluoroalkyl substances with markers of glycaemic control, insulin secretion and sensitivity, and diabetes risk: a systematic review and meta-analyses. Image credit: Reshetnikov_art/Shutterstock.com
In a recent study published in eClinicalMedicine, researchers assessed associations between exposure to per- and polyfluoroalkyl substances (PFAS) and diabetes outcomes.
PFAS exposure emerges as a potential diabetes risk factor
Diabetes mellitus (DM) is a chronic disease in which the body cannot efficiently use insulin or the pancreas does not produce sufficient insulin. Its prevalence has increased since the 1990s, and is currently estimated to affect more than 828 million individuals worldwide. Beyond lifestyle and genetic factors, exposure to environmental chemicals can contribute to the etiology of DM.
PFAS are persistent and ubiquitous chemicals that can interfere with the endocrine system and increase the risk of DM. They are referred to as forever chemicals due to their pervasiveness and long half-lives. PFAS have been linked to metabolic syndrome and its components, and may promote insulin resistance, induce cellular stress and inflammation, or alter pancreatic function and homeostasis.
Examining the PFAS-diabetes link
In the present study, researchers characterized associations between PFAS exposure and markers of glycemic control, pancreatic β-cell function, insulin resistance, and diabetes risk. First, a systematic literature search was conducted to identify human studies from the Medline and Embase databases. Eligible studies were population-based observational studies that analyzed associations between diabetes and PFAS.
Title, abstract, and full-text screening of eligible studies was performed, and relevant information was extracted. For each outcome of interest, the number of studies reporting at least one significantly positive or negative association with individual PFAS was summarized. A qualitative synthesis was conducted as an exploratory analysis for studies examining multipollutant or mixture PFAS associations.
Further, random-effects meta-analyses were performed between PFAS and gestational DM (GDM), type 2 diabetes (T2D), homeostatic model assessments of β-cell function (HOMA-β) and insulin resistance (HOMA-IR), fasting insulin, glycated hemoglobin (HbA1c), and fasting glucose. The risk of bias of studies and the quality and strength of the evidence were assessed using the Navigation Guide.
Strongest evidence linked PFAS exposure to gestational diabetes
Of the 738 records identified through database searches, 129 records were included after screening. Most studies were conducted in the US (54) and China (30) and were cross-sectional in design (70). Across studies, 45 different PFAS were quantitatively measured, with the most common being perfluorooctanoic acid (PFOA), perfluorooctanesulfonic acid (PFOS), perfluorohexanesulfonic acid (PFHxS), perfluorodecanoic acid (PFDA), and perfluorononanoic acid (PFNA).
Most studies examined PFAS exposure in adulthood (75), the gestational or prenatal period (42), adolescence (22), and childhood (16), highlighting potential differences by exposure timing. The most common diabetes outcomes were T2D, GDM, fasting glucose, and HOMA-IR. Study populations were predominantly the general public. Nearly one-third of studies examined outcomes in pregnant individuals.
Only three studies examined type 1 diabetes, and their findings were inconsistent. The meta-analyses, which included 79 studies assessing 18 distinct PFAS, showed that several PFAS were associated with higher odds of gestational diabetes. Specifically, each doubling of PFOS and perfluorobutanesulfonic acid (PFBS) levels was linked to increased GDM risk in prospective studies.
Nested case, control studies reported positive associations for several long-chain and emerging PFAS, including PFOA, PFNA, PFDA, PFBS, perfluoroheptanesulfonic acid (PFHpS), and 6:2 chlorinated polyfluoroalkyl ether sulfonate (6:2 Cl-PFESA), while cross-sectional or case, control studies found positive associations for perfluorododecanoic acid (PFDoDA) and PFOA, as well as a negative association for perfluoroheptanoic acid (PFHpA).
For T2D, the associations were insignificant, although PFNA, PFOA, and PFOS showed a positive overall direction in prospective studies. In homeostatic model assessments, several significant associations were observed: higher β-cell function and insulin resistance with each doubling of PFAS levels. Specifically, HOMA-IR was positively associated with PFOS and PFNA in prospective studies.
HOMA-β was associated with PFOA, PFOS, and PFNA in cross-sectional studies and PFNA in prospective studies. For fasting insulin measures, positive associations were observed with PFOS and PFNA in prospective studies. There were generally no significant associations with HbA1c in the main analyses. However, sensitivity analyses restricted to studies with a lower risk of bias identified a positive association between PFOS and HbA1c in cross-sectional studies.
Associations with fasting glucose were limited and inconsistent across PFAS and study designs, with mixed or largely null findings, particularly in pregnancy-specific analyses. Most studies were classified as having a low or probably low risk of bias regarding outcome and exposure assessment, study population, baseline differences, and confounding.
Sixty studies examined multiple PFAS or jointly with other chemicals. Of these, 35 included PFAS as part of an exposure mixture, and six also considered other chemicals beyond PFAS. Most mixture studies assessing GDM or T2D reported positive associations, whereas studies using summed PFAS measures often reported inconsistent or null findings.
Finally, the evidence was of moderate quality for some long-chain or legacy PFAS and selected outcomes, but low quality for short-chain or emerging PFAS and for outcomes such as T2D, fasting glucose, and HbA1c, and was limited or inadequate to determine with certainty the associations between PFAS exposure and diabetes outcomes in humans.
Modest metabolic effects, but uncertainty remains
The most consistent finding was the association between increased PFAS exposure and greater odds of GDM. Evidence on PFAS, T2D associations is yet to be explained, while that on type 1 diabetes remains scarce. Meta-analyses revealed positive associations between PFAS and HOMA-IR, and limited associations with fasting glucose, HOMA-β, and fasting insulin.
Effect sizes for continuous metabolic markers were generally small, and many findings were derived from cross-sectional studies, limiting causal inference and raising the possibility of reverse causation. Overall, the evidence suggests that exposure to some PFAS is associated with an increased GDM risk and modest changes in insulin sensitivity and secretion in the general population.
Download your PDF copy now!
Natural daylight during office hours improves glucose stability in type 2 diabetes