Associations between plasma metals and hemoglobin in female college students with dysmenorrhea

Abstract

Background: Hemoglobin (HGB) was the most important factors which could cause dysmenorrhea in women. Metals exposure and hemoglobin level in dysmenorrhea female was unclear. We aimed to explore the associations of multi-metal exposure and HGB level in female college students with dysmenorrhea.

Methods: 253 female students who had dysmenorrhea was included in our study. The Last Absolute Shrinkage and Selection Operator (LASSO) regression, generalized linear model (GLM), and Bayesian Kernel Machine Regression (BKMR) models were used to explore the associations of multi-metal exposure and HGB levels in female college students with dysmenorrhea.

Results: GLM results showed that plasma Fe, Ni and Rb was positively associated with HGB and plasma Co was negatively associated with HGB. In menarche age ≤13 years old group, plasma Co and Rb only was negatively and positively associated with HGB level, respectively, and plasma Ni had positive association with HGB level in menarche age >13 years old group. BKMR results showed the reverse U-shaped relationship between the five metals mixture (Co, Fe, Ni, Cu and Rb) and HGB levels in overall and menarche age ≤13 years old group. However, there were positive association between the five metals mixture and HGB levels in menarche age >13 years old group.

Conclusion: Our present study revealed that metals (Fe, Ni, Co, Rb, Cu) mixture exposure could effect HGB levels in female college students with dysmenorrhea. And the relationships were different during different menarche age in female college students.

Keywords: Combined effect; Dysmenorrhea; Exposure; Hemoglobin; Multi-metal.

Fig. 1

Correlation of 22 plasma metals in female undergraduate with dysmenorrhea (−0.24 < r < 0.84).

Fig. 2

Restricted cubic spline (RCS) regression analysis. Analysis association between HGB concentrations and the concentration of metals which were selected by generalize linear regression. A: overall group; B: Menarche ≤13 years old; C: Menarche>13 years old. Adjusted factors: age, BMI, mother dysmenorrhea, age of menarche, passive smoking, night snack.

Fig. 3

The prediction error of the LASSO regression model as a function of the penalty parameter (log λ) (the figure at right) and the LASSO solution path with metal coefficient profiles based on the penalty parameter (log λ) (the figure at left). For each metal, the vertical red line represents the cross-validated optimum of λ (minimum mean-squared error), and forced variables and several metals were selected at the λ indicated by the vertical red line. As indicated by the vertical red line, Fe, Ni, Co and Rb were selected at the lambda (λ = 2.009) for analysis of the level of HGB.

Fig. 4

The overall effect of cumulative mixture with HGB. The lines were overall effect of three metals mixtures (estimates and 95 % credible intervals) on HGB level, when all of exposures were at a particular quantile as compared to when all of them were at their 50th value. A: overall group; B: Menarche ≤13 years old; C: Menarche>13 years old. Adjusted factors: age, BMI, mother dysmenorrhea, age of menarche, passive smoking, night snack.

Fig. 5

Univariate exposure-response functions for each plasma metal (95 % confidence interval) and serum HGB with the other two plasma metals fixed at the median value. Blue line acted as estimates and gray band acted as 95%CI. A: overall group; B: Menarche ≤13 years old; C: Menarche>13 years old. Adjusted factors: age, BMI, mother dysmenorrhea, age of menarche, passive smoking, night snack.

Fig. 6

The single-exposure effect (estimates and 95 % credible intervals) of each metal with serum HGB for increasing from 25th to 75th in metal distributions when the other metals were fixed at the P₂₅, P₅₀ or P₇₅. A: overall group; B: Menarche ≤13 years old; C: Menarche>13 years old. Adjusted factors: age, BMI, mother dysmenorrhea, age of menarche, passive smoking, night snack.