A pooled analysis of reproductive factors, exogenous hormone use and risk of multiple myeloma among women in the International Multiple Myeloma Consortium

Abstract

Background

Female sex hormones are known to have immunomodulatory effects. Therefore, reproductive factors and exogenous hormone use could influence the risk of multiple myeloma in women. However, the role of hormonal factors in multiple myeloma etiology remains unclear because previous investigations were underpowered to detect modest associations.

Methods

We conducted a pooled analysis of seven case–control studies included in the International Multiple Myeloma Consortium, with individual data on reproductive factors and exogenous hormone use from 1,072 female cases and 3,541 female controls. Study-specific odd ratios (ORs) and corresponding 95% confidence intervals (CIs) were estimated using logistic regression and pooled analyses were conducted using random effects meta-analyses.

Results

Multiple myeloma was not associated with reproductive factors, including ever parous (OR=0.92, 95%CI=0.68-1.25), or with hormonal contraception use (OR=1.04, 95%CI=0.80-1.36). Postmenopausal hormone therapy users had non-significantly reduced risks of multiple myeloma compared with never users, but this association differed across centers (OR=0.65, 95%CI=0.37-1.15, I²=76.0%, p-heterogeneity= 0.01).

Conclusions

These data do not support a role for reproductive factors or exogenous hormones in myelomagenesis.

Impact

Incidence rates of multiple myeloma are higher in men than in women, and sex hormones could influence this pattern. Associations with reproductive factors and exogenous hormone use were inconclusive despite our large sample size, suggesting that female sex hormones may not play a significant role in multiple myeloma etiology.

INTRODUCTION

Multiple myeloma (MM) is a malignancy characterized by the accumulation of clonal plasma cells in the bone marrow, abnormal secretion of monoclonal protein and end organ damage (1). Incidence rates are higher in men than in women (2). Because female sex hormones have immunomodulatory effects, reproductive factors and exogenous hormone use may affect risk for MM. However, the role of hormonal factors in MM etiology remains unclear. A few studies addressed possible associations between MM risk and reproductive factors, such as parity (3–6) or use of postmenopausal hormone therapy (HT) (6–8), but yielded inconsistent results as most studies were underpowered. We conducted a pooled analysis of case–control studies included in the International Multiple Myeloma Consortium (IMMC) to clarify the role of hormonal factors in the etiology of MM.

MATERIALS AND METHODS

We pooled individual-level questionnaire data from the seven IMMC case-control studies that collected information on reproductive factors among women (1,072 cases and 3,541 controls). These studies were: Los Angeles County Multiple Myeloma Case-Control Study (LAMMCC), Roswell Park Cancer Institute (RPCI), Utah, Epilymph, Fred Hutchinson Cancer Research Center (FHCRC) 1980s, National Cancer Institute (NCI)-Yale, and Molecular and Genetic Epidemiology Study (iMAGE). Enrollment period, age eligibility, study design, sample sizes, and participation rates within each study are summarized in Supplemental Table 1. Parity was defined as number of live-births in NCI-Yale, iMAGE, and RCPI, and as number of children in all other studies.

Within each study, we computed odds ratios (ORs) with corresponding 95% confidence intervals (CI) using unconditional logistic regression, adjusting for age group (four categories), race (except for Epilymph, which did not collect these data), and study center (for multicentric studies: Epilymph and FHCRC 1980s). Random-effects models were used to calculate pooled estimates using the DerSimonian & Laird method. Heterogeneity between studies was assessed using the I² statistic and p-heterogeneity using the Mantel-Haenszel method. Analyses on parity and gravidity were restricted to women aged 45 or older, as they are likely to have completed their reproductive history. Analyses on HT were restricted to postmenopausal women, defined as women who reported cessation of their menstrual periods. Wald tests were utilized to assess heterogeneity between strata.

RESULTS

In this pooled analysis, we did not observe any statistically significant association between MM and age at menarche or at menopause, ever pregnant, number of pregnancies, ever parous, number of children, age at first birth, or cause of menopause (Table 1). The association between MM and ever use of hormonal contraceptives was not significant (OR=1.04, 95%CI=0.80-1.36; Table 1). Similarly, we saw no significant associations or consistent patterns for age and year at first use, duration, or time since last hormonal contraceptive use.

Table 1.

Associations between reproductive factors and exogenous hormone use and multiple myeloma risk.

	Co	Ca	Pooled OR (95%CI)*	I2	P-het
REPRODUCTIVE FACTORS
Age at menarchea
Total	1335	482			No. centers=4
<11	256	86	Ref
12-13	717	271	1.20 (0.89, 1.63)	0.0%	0.58
>14	362	125	1.03 (0.73, 1.45)	0.0%	0.57
Ever pregnantc
Total	2321	691			No. centers=6
No	228	69	Ref
Yes	2093	622	0.90 (0.57, 1.44)	48.2%	0.09
No of pregnanciesb
Total	1494	593			No. centers=5
None	142	64	Ref
1	145	52	0.80 (0.39, 1.64)	55.2%	0.06
2	329	134	0.83 (0.57, 1.20)	0.0%	0.78
3	296	118	0.90 (0.50, 1.60)	51.8%	0.08
>4	582	225	0.91 (0.54, 1.54)	50.3%	0.09
Ever parousd
Total	3075	1007			No. centers=7
Never	408	150	Ref
Ever	2667	857	0.92 (0.68, 1.25)	41.8%	0.11
No. of childrend
Total	3075	1007			No. centers=7
None	408	150	Ref
1	410	138	0.95 (0.66, 1.38)	30.9%	0.19
2	802	274	0.96 (0.75, 1.24)	0.0%	0.46
3	635	190	0.91 (0.60, 1.37)	53.1%	0.05
>4	820	255	0.91 (0.64, 1.30)	41.3%	0.12
Age at first birthe
Total	1941	449			No. centers=4
Nulliparous	255	64	Ref
<20	237	66	0.97 (0.50, 1.89)	54.4%	0.09
20-<25	749	167	0.99 (0.58, 1.70)	56.1%	0.08
25+	700	152	0.96 (0.59, 1.57)	44.8%	0.14
Age at menopauseb
Total	1265	524			No. centers=5
<45	414	184	Ref
45-49	313	118	1.00 (0.74, 1.35)	0.0%	0.46
>50	538	222	1.12 (0.86, 1.45)	0.0%	0.58
Cause of menopausef
Total	800	397			No. centers=4
Natural	433	197	Ref
Surgical / Therapeutic	367	200	1.11 (0.80, 1.54)	36.7%	0.19
EXOGENOUS HORMONE USE
Ever hormonal contraceptiong
Total	2450	590			No. centers=5
Never used	1805	426	Ref
Ever used	645	164	1.04 (0.80, 1.36)	0.0%	0.66
Age at first hormonal contraceptiong
Total	2434	587			No. centers=5
Never used	1805	426	Ref
<25	421	101	1.07 (0.76, 1.49)	0.0%	0.80
>25	208	60	1.08 (0.76, 1.54)	0.0%	0.42
Year at first hormonal contraceptiong
Total	2434	587			No. centers=5
Never used	1805	426	Ref
<1975	353	119	1.19 (0.87, 1.62)	0.0%	0.87
≥1975	276	42	1.18 (0.65, 2.14)	19.6%	0.29
Time since last hormonal contraceptiong
Total	2421	588			No. centers=5
Never used	1891	428	Ref
<20	275	70	1.22 (0.84, 1.76)	0.0%	0.50
>20	255	90	1.09 (0.75, 1.58)	15.2%	0.32
Years of hormonal contraceptiong
Total	2415	582			No. centers=5
Never used	1805	426	Ref
<5	217	69	1.30 (0.92, 1.83)	0.0%	0.89
>5	393	87	0.96 (0.57, 1.63)	55.6%	0.06
Ever HTa
Total	1076	432			No. centers=4
Never used	703	307	Ref
Ever used	373	125	0.65 (0.37, 1.15)	76.0%	0.01
Age first used HTa
Total	1057	425			No. centers=4
Never used	703	307	Ref
<50	197	72	0.60 (0.31, 1.17)	71.5%	0.01
>50	157	46	0.61 (0.41, 0.90)	4.5%	0.37
Year first used HTa
Total	1057	425			No. centers=4
Never used	703	307	Ref
<1980	143	54	0.58 (0.33, 1.04)	38.9%	0.18
≥1980	211	64	0.81 (0.37, 1.77)	75.3%	0.01
Time since last HT consumptiona
Total	1055	424			No. centers=4
Never used	703	307	Ref
Current	150	41	0.84 (0.30, 2.38)	77.1%	<0.01
<10	116	46	0.90 (0.37, 2.16)	76.5%	0.01
>10	86	30	0.52 (0.21, 1.26)	57.9%	0.07
Years of HT usea
Total	1056	423			No. centers=4
Never used	703	307	Ref
<5	136	54	0.64 (0.30, 1.37)	71.4%	0.01
>5	217	62	0.56 (0.33, 0.97)	57.6%	0.07

^*Adjusted for center, age (four categories), and race (white, black and others).

Co: controls Ca: cases, OR: Odds Ratio, CI: Confidence interval, P-het: p-value for heterogeneity, HT: postmenopausal hormone therapy.

^aStudies with data on periods starting, ever HT use, number of years HT was used, years since last HT consumption and age at first post-menopausal HT use were LAMMCC, RPCI, NCI-Yale and iMAGE. Analyses on HT variables were performed among postmenopausal women.

^bAnalyses on periods stopping were performed among postmenopausal women. Analyses on number of pregnancies were performed among women aged 45 or older at reference date. Studies with data on periods stopping and number of pregnancies were LAMMCC, RPCI, NCI-Yale, iMAGE, and Utah.

^cAmong women aged 45 or older at reference date. Studies with data on ever being pregnant were LAMMCC, RPCI, NCI-Yale, iMAGE, Utah and Epilymph.

^dAmong women aged 45 or older at reference date. All studies collected data on parity and number of children.

^eAmong women aged 45 or older at reference date. Studies with data on age at first child were RPCI, Epilymph, NCI-Yale and iMAGE.

^fAmong postmenopausal women. Studies with data on cause of menopause were LAMMCC, RPCI, Utah and iMAGE.

^gStudies with data on hormonal contraception use, number of years hormonal contraception was used, years since last hormonal contraception consumption and age at first hormonal contraception use were LAMMCC, RPCI, Epilymph, NCI-Yale and iMAGE.

HT use showed non-significant decreased risks of MM (OR=0.65, 95%CI=0.37-1.15), but also showed significant heterogeneity between centers (I²= 76.0%, p=0.01, Figure 1). Further adjustment for BMI, education, tobacco, and alcohol yielded a similar risk estimate (OR=0.70, 95%CI=0.39-1.25). Inverse associations were observed among women taking HT at ages 50 or older, or for more than 5 years, compared with never use (OR=0.61, 95%CI=0.41-0.90; and OR=0.56, 95%CI=0.33-0.97, respectively), although heterogeneity between centers hampered interpretation (Supplemental Figure 1). Stratified analyses by cause of menopause, education, and BMI did not reveal statistically significant heterogeneity (data not shown).

Figure 1.

Study-specific risks of multiple myeloma for ever versus never parous, hormonal contraceptives and postmenopausal hormone therapy.

DISCUSSION

This large pooled analysis of 1,072 female cases and 3,541 controls yielded null associations between MM and reproductive factors. To our knowledge, 3 case-control studies (3–5,8) and 3 cohorts (4,6,7) have previously evaluated associations between reproductive factors, or exogenous hormone use and risk of MM. Inconsistent results were observed for parity and MM, with both significant inverse associations (5), increased risks (4), and null results (3,6). Previously reported associations for hormonal contraceptives and MM have been null (5,6). Significant inverse associations for HT use were observed in an Italian case-control study (8), but these associations were not corroborated in two cohort studies (6,7). However, conclusions in these studies have been limited by small sample sizes of women using HT.

Our study was based on a large dataset with individual-level information on reproductive factors and exogenous hormone use, yet we did not observe consistent patterns with these factors and MM risk. We had the ability to control for a variety of potential confounders, including education, BMI and alcohol use. Ignoring these variables may have biased previous studies of HT and cancer, due to the potential for selection bias and a healthy user effect. We did not observe clear evidence of confounding by those variables in the present analysis, although residual confounding cannot be discarded in explaining some of our results, in particular for HT. Also, use of controls that may not be representative of the population from which the cases arose was an inherent limitation of some of the participating studies’ design. In summary, our data do not support a significant role for reproductive factors or exogenous hormones in myelomagenesis.