Sei sulla pagina 1di 3

C-reactive protein in relation to fecundability and anovulation among eumenorrheic women

Supplemental Methods

Sampling scheme for urine specimens

This scheme was applied to all 1,228 women who enrolled in the EAGeR Trial. One specimen
was sampled from the early follicular phase, day 2 - 4. Also, one specimen specimen was
sampled from the peri-ovulatory window. In N=1,450 cycles with a peak fertility reading, this
was one day after the day a peak fertility reading appeared on the fertility monitor. In N=432
cycles in which the monitor detected ≥1 day with high fertility but not peak fertility, the day with
the maximum luteinizing hormone (LH) level from days 10-20 was the peak day. If only low
fertility was detected (N=89 cycles), the “peri-ovulatory” specimen was sampled from the 15th
day prior to the end of the cycle. The luteal phase specimens were sampled at two time points:
(1) nine days post-peak fertility, and (2) a randomly-selected day from four days post-peak
fertility until two days before the end of the cycle. Some cycles with peak fertility detected had
three specimens sampled from the luteal phase: 100 cycles with the highest maximum urinary
estrone-3 glucuronide and the 100 longest cycles. Only one luteal phase specimen was selected if
only low fertility was detected, from seven days before the end of the cycle.
Prediction model for percent body fat

The model for predicted percent body fat was developed using data from the BioCycle Study (1).
Participants (n=259) were healthy, eumenorrheic women, age 18-44 years. They were not
currently attempting pregnancy or using hormonal contraception, and they had not recently been
pregnant or breast feeding. At the baseline study clinic visit, study staff measured participant
weight and height; waist, hip, and upper-arm circumferences; and subscapular, suprailiac, and
triceps skinfolds. A dual energy X-ray absorptiometry scan performed two months later—at the
end of follow-up--measured overall percent body fat. An automated backwards selection
procedure was implemented using PROC GLMSELECT in SAS version 9.4 (SAS Institute,
Cary, NC). The selection procedure began with a model that included participant age (continuous
variable), non-white vs. white race, all variables for anthropometric measures (log-transformed),
and all two-way interaction variables. The selected model had adjusted r-square = 0.709. It is
presented here:

Parameter Linear Regression Coefficient Standard Error t-Statistic

Intercept 488.892726 118.528585 4.12


Log(subscapular skinfold) 4.842546 1.037292 4.67

Log(suprailiac skinfold) 60.288690 17.818182 3.38


Log(triceps skinfold) 5.642906 1.337324 4.22
Log(upper arm circumference) -158.061890 34.840228 -4.54
Log(Waist-hip ratio) -11.161908 3.344576 -3.34
Log(BMI) -206.770839 47.292281 -4.37

Age 0.072813 0.027095 2.69


Nonwhite (vs. white) -11.147251 4.775392 -2.33

interaction- log suprailiac and log upper arm -17.766003 5.324408 -3.34
Interaction- log triceps and nonwhite 3.696449 1.558473 2.37
Interaction- log upper arm and log BMI 64.714715 13.883123 4.66

Analysis of hsCRP and anovulation among all EAGeR participants

Because we previously found that LDA did not affect sporadic anovulation (2), this
secondary analysis included all 1,150 participants with hsCRP<10 mg/L, after excluding 14
women who withdrew before the first follow-up visit and 64 women with hsCRP≥10 mg/L. We
analyzed the data with the same methods described in the Materials and Methods section for the
analysis of the placebo group. The results are presented in Supplemental Table 2.
References

1. Wactawski-Wende J, Schisterman EF, Hovey KM, Howards PP, Browne RW, Hediger
M, et al. BioCycle study: design of the longitudinal study of the oxidative stress and
hormone variation during the menstrual cycle. Paed Perinat Epidemiol 2009; 23:171-
184.
2. Radin RG, Sjaarda LA, Perkins NJ, Silver RM, Chen Z, Lesher LL, Galai N, Wactawski-
Wende J, Mumford SL, Schisterman EF. Low-dose aspirin and sporadic anovulation in
the EAGeR randomized trial. J Clin Endocrinol Metab 2017;82:86-92.

Potrebbero piacerti anche