Pelvic Pain Symptoms and Inflammation Among Adolescents and Adults with and Without Endometriosis

Abstract

We evaluated inflammatory markers among 389 surgically confirmed endometriosis cases and 505 controls from the Women's Health Study: From Adolescence to Adulthood (A2A) cohort. Participants reported dysmenorrhea, acyclic pelvic pain, dyspareunia, and pain with bowel movements. Using multiplex assays, we measured their levels of plasma interleukin (IL)-1β, -6, -8, -10, and -16, tumor necrosis factor (TNF)-α, monocyte chemotactic protein (MCP)-1 and -4, thymus and activation-regulated chemokine (TARC), and interferon gamma-induced protein (IP)-10. For each symptom, we computed biomarker-level geometric means (GMs) with 95% confidence intervals (95% CI) using multivariate linear regression among the endometriosis cases and controls, with interactions with case/control status tested using Wald statistics. Among the controls, those with dyspareunia had lower levels of IL-8 (GM_present = 4.64 [95% CI = 4.41-4.89] pg/mL vs. GM_absent = 4.99 [95% CI = 4.82-5.17] pg/mL; p = 0.02), and the IL-8 levels were lower for controls reporting pain with bowel movements (GM_present = 4.66 [95% CI = 4.43-4.89] vs. GM_absent = 4.96 [95% CI = 4.82-5.11] pg/mL, p = 0.03). No significant associations between pelvic pain symptoms and inflammatory markers were observed among the endometriosis cases; however, the relationship between inflammatory marker levels and pain experience varied by analgesic use at blood draw. Dyspareunia and pain with bowel movements were associated with inflammatory markers among the controls, while the associations between pelvic pain symptoms and inflammatory markers among the endometriosis cases differed by analgesic use.

Keywords: bowel pain; chemokines; cytokines; dysmenorrhea; dyspareunia; endometriosis; inflammation; pelvic pain.

Figure 1

Overall direction and magnitude of effect for associations between pelvic pain symptoms and plasma cytokine and chemokine levels among A2A participants with and without endometriosis are presented in Panels (A,B). Arrows show direction of association for changes in geometric mean biomarker levels of 10% or greater or where p-value for association is <0.10. Panel (A) displays direction of association (arrows) and strength of association (shaded boxes) for controls. Panel (B) shows direction of association (arrows) and strength of association (shaded boxes) for endometriosis cases. Double dagger indicates significant heterogeneity between endometriosis cases and controls for association between pelvic pain symptoms and inflammation markers.

Figure 2

Geometric mean levels of CRP, IL-6, IL-8, and IL-10 in relation to dysmenorrhea severity and frequency and presence, severity, and frequency of dyspareunia among endometriosis cases (blue diamonds) and controls (red squares) in A2A cohort. All estimates were adjusted for age at blood draw (continuous), body mass index at blood draw (under-/normal weight, overweight, obese), pain medication use within 48 h of blood draw (yes, no, missing), and hormone use within 30 days of blood draw (yes, no). Asterisks (*) denote borderline and significant p-values for trend (frequency and severity) and p-values (presence) (no asterisk: p-value > 0.10; one asterisk: p-value = 0.05–<0.10; two asterisks: p-value < 0.05). Double dagger symbols (‡) denote significant heterogeneity (p-het < 0.05) between endometriosis cases and controls. Full results, including sample size numbers, can be found in Table S3.

Figure 3

Geometric mean levels of MCP-1, MCP-4, TARC, IL-16, and IP-10 in relation to dysmenorrhea severity and frequency and presence, severity, and frequency of dyspareunia among endometriosis cases (blue diamonds) and controls (red squares) in A2A cohort. All estimates were adjusted for age at blood draw (continuous), body mass index at blood draw (normal weight, overweight, obese), pain medication use within 48 h of blood draw (yes, no, missing), and hormone use within 30 days of blood draw (yes, no). Asterisks (*) denote borderline and significant p-values for trend (frequency and severity) and p-values (presence) (no asterisk: p-value > 0.10; one asterisk: p-value = 0.05–<0.10). Full results, including sample size numbers, can be found in Table S3.

Figure 4

Geometric mean levels of CRP, IL-6, IL-8, and IL-10 in relation to presence, severity, and frequency of acyclic pelvic pain and presence of pain with bowel movements among endometriosis cases (blue diamonds) and controls (red squares) in A2A cohort. All estimates were adjusted for age at blood draw (continuous), body mass index at blood draw (normal weight, overweight, obese), pain medication use within 48 h of blood draw (yes, no, missing), and hormone use within 30 days of blood draw (yes, no). Asterisks (*) denote borderline and significant p-values for trend (frequency and severity) and p-values (presence) (no asterisk: p-value < 0.10; one asterisk: p-value = 0.05–<0.10; two asterisks: p-value < 0.05). Full results, including sample size numbers, can be found in Table S3.

Figure 5

Geometric mean levels of MCP-1, MCP-4, TARC, IL-16, and IP-10 in relation to presence, severity, and frequency of acyclic pelvic pain and presence of pain with bowel movements among endometriosis cases (blue diamonds) and controls (red squares) in A2A cohort. All estimates were adjusted for age at blood draw (continuous), body mass index at blood draw (normal weight, overweight, obese), pain medication use within 48 h of blood draw (yes, no, missing), and hormone use within 30 days of blood draw (yes, no). Asterisks (*) denote borderline and significant p-values for trend (frequency and severity) and p-values (presence) (no asterisk: p-value < 0.10; one asterisk: p-value = 0.05–<0.10). Full results, including sample size numbers, can be found in Table S3.

Figure 6

Timeline of A2A cohort enrollment and blood sample collection in relation to assaying of inflammatory markers using Luminex and Ella assays.