Age at Menarche and the Risk of Cardiac Events in Women With Congenital Long QT Syndrome

Abstract

Background

Women with congenital long‐QT syndrome (LQTS) experience increased risk of cardiac events (CE) after the onset of adolescence, possibly due to the effect of sex hormones on the cardiac ion channels. We hypothesized that late menarche may affect the risk of CE in women with LQTS.

Methods

Beginning in 2010, information on age at menses onset was obtained from all women enrolled in the Rochester LQTS Registry. Multivariable modeling was employed to evaluate the association of age at menarche with the burden of CE (total number of syncope, aborted cardiac arrest, and LQTS‐related sudden cardiac death) during the subsequent 20 years of follow‐up. Menarche groups were defined as early (<12 years), normal (12–16 years), and late (>16 years).

Results

We report data on 435 genetically confirmed LQTS women, of whom 68, 346, and 21 were in the early, normal, and late‐onset groups, respectively. The mean cumulative rate of CE at 20 years of follow‐up in women with late menarche was 28% and 64% higher than in women with normal or early menarche, respectively; P<0.001 for the overall difference during follow‐up. Consistently, multivariable analysis showed that late menarche was associated with a pronounced 52% increased risk of CE compared with normal menarche (P=0.02).

Conclusions

Late menarche is associated with increased risk of CE in women with congenital LQTS. Further research is warranted to validate these findings and ascertain the explanatory pathophysiological pathways.

Nonstandard Abbreviations and Acronyms

BB

beta blockers

CE

cardiac events

LCSD

left cardiac sympathetic denervation

LQTS

congenital long QT syndrome

Clinical Perspective.

What Is New?

• Late (>16 years old) menarche is associated with a significant increase in cardiac events among women with congenital long QT syndrome, including long QT syndrome‐related syncope, aborted cardiac arrest or long QT syndrome‐related sudden cardiac death.

• Beta‐blocker therapy, an important treatment option, was associated with a significant, nearly 70% reduction in the risk of recurrent cardiac events, with a protective effect consistent across menarche ages.

What Are the Clinical Implications?

• Clinicians should consider hormonal factors, including age at menarche, for risk stratification in women with congenital long QT syndrome.

In patients with congenital long QT syndrome (LQTS), the risk of cardiac events (CE) differs by sex and age. In comparison with male patients, female patients with LQTS have an increased risk for cardiac events after the onset of adolescence, possibly due to a modulating effect of female sex hormones on arrhythmic risk, especially in women with mutations in the rapidly activating delayed rectifier potassium current (I_kr). ¹ , ²

Furthermore, female sex was shown to be associated with a baseline longer QTc and is also an independent risk factor for the development of torsades de pointes in patients with acquired (drug‐induced) LQTS, suggesting that the effect of sex hormones on arrhythmic risk may be related to cardiac ion channel mechanisms. ³ , ⁴ , ⁵

Progesterone has been shown to exert protective effects on cardiac repolarization potassium channel activity, whereas estrogen has opposite effects. ⁵ , ⁶ Hence, changes in hormonal balance (ie, estrogen/progesterone ratio) due to altered menstrual physiology may affect the QTc interval. ⁷ Nonetheless, an assessment of the association of the age at menarche with the burden of CE after the onset of adolescence in women with LQTS has not been previously conducted.

This study, based on the Rochester LQTS Registry, aimed to (1) evaluate the association between age at menarche and the risk of recurrent CE; and (2) assess the protective effect of beta‐blocker (BB) therapy on recurrent CE risk by menarche age.

METHODS

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Study Population

This study comprised women from the Rochester LQTS Registry who responded to a questionnaire about their menstrual and biological hormone‐driven periods. The questionnaire was sent via email or as a package to female patients who met the following criteria: (1) ≥18 years of age; (2) genotype positive for LQT1 to LQT3, or clinically diagnosed with QTc interval ≥450 ms and are identified clinically as LQTS in the registry (patients from families with LQT who did not undergo genetic testing); and (3) who are alive and consent to the study.

The University of Rochester Medical Center Research Subjects Review Board approved this study. All patients provided written informed consent.

Questionnaire and Menarche Definitions

The questionnaire included 3 main questions about menarche, oral contraceptive use, and menopause (Table S1). Due to the retrospective nature of the data, patients reported their best estimate of the date of menarche. To account for potential recall inaccuracy, age at menarche was analyzed using rounded chronological years rather than decimal ages. Information regarding pregnancy, BB, or abortion is captured prospectively as part of the regular follow‐up beginning in 2010.

Menarche periods were categorized in accordance with existing literature as follows ⁸ , ⁹ :

• Early menarche: if the first menstrual period occurred before the age of 12 years.

• Normal menarche: if the first menstrual period occurred between the ages of 12 and 16 years.

• Late menarche: menstrual period occurred after the age of 16 years.

Study Outcomes

To facilitate the evaluation of the association of age at menarche with the risk of recurrent CE, we employed statistical methodology that allowed for the inclusion of repeated CE in the multivariable models. Therefore, the primary end point of the present study was the occurrence of total CE of any type (comprising LQTS‐related syncope, aborted cardiac arrest, and LQTS‐related sudden cardiac death) during follow‐up. LQTS‐related syncope was defined as a transient loss of consciousness with abrupt onset and offset, in a setting strongly suggestive of having stemmed from LQTS‐associated torsades de pointes, as determined by review of patient data and the clinical manifestation of the event, as previously reported by the study authors. ¹⁰ To minimize the bias, syncopal events were clinically adjudicated by the expert electrophysiologist cardiologists involved in the Rochester LQTS Registry (W.Z., I.G.). Aborted cardiac arrest was defined as an event requiring defibrillation as part of the resuscitation or appropriate shock among patients with implantable cardioverter‐defibrillators; and LQTS‐related sudden cardiac death was defined as abrupt in onset without evident cause, if witnessed, or death that was not explained by any other reason if it occurred in a nonwitnessed setting. The circumstances of the events (including data on onset, prodromal symptoms, and seizures) were corroborated by the study coordinators through the patient's medical files and interviews with individuals about themselves or about family members and categorized by the study specialists using prespecified codes.

Statistical Analysis

Patients were categorized into 3 groups based on their first menstruation date in relation to the onset of menstruation (early menarche, normal menarche, and late menarche). Previous studies have demonstrated a varied association between age at menarche and cardiac events, suggesting that clinically, these are distinct patient populations. ¹¹ Accordingly, the study hypothesis was that the prespecified groups may differ in outcomes due to clinical differences such as hormonal exposure, which has been previously demonstrated to be associated with LQTS outcomes. Thus, comparative analyses were performed according to the prespecified categorization. Patients were compared using the Wilcoxon test for continuous variables and the chi‐square test for categorical variables. The clinical characteristics of the study population were summarized by the mean±SD for continuous variables and by frequencies and proportions for categorical variables.

Analysis of the primary end point was based on the unconditional Andersen and Gill non‐gap‐time model, an extension of the stratified Cox model. The model was used to estimate the hazard ratio (HR) for recurrent CE (both first and subsequent occurrences). The hazard ratio was adjusted for the history of CE before menarche, baseline QT interval corrected for heart rate (QTc), birth decade, LQT genotype, a time‐dependent variable for oral contraceptive use, and a time‐dependent variable for BB use. Moreover, to account for the inflated SE inherent in recurrent events, we employed the robust sandwich estimate proposed by Lin and Wei. ¹² Interaction‐term analysis was used to assess the effect of BB use on the risk of CE within the 3 menarche groups. The primary aim was to evaluate the burden of cardiac events overall by menarche age, without differentiating by event type or order of occurrence. Therefore, the order of events was not the focus of this analysis.

The mean cumulative event rate in the 3 groups is presented using the Ghosh–Lin curves, as appropriate, with the time of follow‐up beginning at menarche in each group. Patients were censored at age 40 years or age at menopause, whichever came first, to avoid the potential bias associated with the effects of early menopause and concomitant risk of non‐LQTS‐related cardiovascular events.

Analyses were conducted using SAS software (version 9, SAS Institute, Cary, NC).

RESULTS

Baseline Characteristics

Among the 435 study subjects, 346 (80%) were in the normal menarche group, 68 patients (15%) had menarche before 12 years (early menarche‐onset group), and 21 patients (5%) experienced menarche at the age of 16 or above (late menarche group) (Figure 1). Clinical characteristics, including baseline QTc, oral contraceptive use, and pregnancy rates, did not differ significantly among the 3 menarche age groups (Table 1). The rate of LQTS‐related therapies, including BBs, left cardiac sympathetic denervation, and implantable cardioverter‐defibrillators, was also similar among the 3 groups (Table 1).

Figure 1. Distribution of the cohort.

Table 1.

Baseline Characteristics

Clinical characteristics	Early menarche (n=68)	Normal menarche (n=346)	Late menarche (n=21)	P value 1–2	P value 2–3
Mean age at menarche, y	10.5±0.6	12.9±1	16.7±0.7
Patients with pregnancy	49 (72)	256 (73)	14 (67)	0.89	0.45
Average of pregnancies per pregnant woman	2.2	2.9	3	0.47	0.76
Menopause before the age of 40	2 (3)	10 (3)	3 (14)	0.86	0.032
ECG at enrollment
QRS	79±13	80±13	83±21	0.57	0.47
QTc, ms	487±45	485±49	480±38	0.47	0.99
Genotype
LQT1	16 (24)	108 (31)	5 (24)	0.25	0.63
LQT2	26 (38)	113 (33)	8 (38)	0.40	0.64
LQT3	6 (9)	29 (8)	2 (10)	0.82	0.69
Other	20 (29)	96 (28)	6 (28)	0.73	0.87
Treatments before menarche
Beta blocker	11 (16)	57 (16)	6 (29)	1.0	0.23
ICD	2 (3)	3 (1)	1 (5)	0.19	0.21
Treatments during follow‐up
Oral contraceptive	22 (32)	85 (25)	2 (10)	0.23	0.18
Beta blockers	42 (62)	235 (68)	14 (67)	0.32	0.98
Left cardiac sympathetic denervation	None	8 (2)	None	0.43	0.76
ICD	16 (24)	77 (22)	4 (19)	0.87	0.97

ICD indicates implantable cardioverter‐defibrillator; LCSD, left cardiac sympathetic denervation; and LQT, long QT.

Menstruation Onset and Risk of Cardiac Events

In the 3 groups, early, normal, and late, the number and proportion of subjects experiencing at least 1 cardiac event were 29 (42.7%), 181 (52.3%), and 13 (61.9%), respectively. Most patients experiencing these events had an LQTS‐related syncope event (N=27, 176, and 13 in each group, respectively), whereas fewer patients suffered more severe events, aborted cardiac arrest or LQTS‐related sudden cardiac death (N=8, 34, and 3 in the groups, respectively). Figure 2 illustrates the mean cumulative total of CE since menarche onset. At a 20‐year follow‐up from menarche, the mean cumulative event rate was significantly higher in the late menarche group compared with the normal menarche and early menarche groups; P<0.001 for the overall comparison during follow‐up.

Figure 2. Mean cumulative number of cardiac events after the onset of adolescence by menarche group.

Consistent with the univariate findings, multivariable Andersen and Gill modeling (Table 2) demonstrated that, compared with normal menarche, late menarche was associated with a significant 52% increased risk of recurrent CE (HR, 1.52 [95% CI, 1.04–2.21]; P=0.029). In contrast, early menarche was not associated with a statistically significant difference in the risk of recurrent CE compared with normal menarche (HR, 0.88 [95% CI, 0.72–1.19]; P=0.15). These findings were consistent when age at menarche was treated as a continuous rather than a categorical variable, with an HR of 1.09 (95% CI, 1.04–1.15, P=0.001) for recurrent CE per 1‐year increment in age at menarche.

Table 2.

Multivariable Analysis: Association of Age at Menarche With Risk of Recurrent Cardiac Events in Women With LQTS^*

	HR (95% CI)	P value
Early menarche	0.88 (0.72–1.19)	0.15
Normal menarche	1.0 (Reference)
Late menarche	1.52 (1.04–2.21)	0.029
Beta blockers	0.31 (0.22–0.44)	<0.001
QTc ≥550 ms	1.57 (1.24–2.00)	<0.001
LQT2 vs LQT1	1.35 (1.08–1.68)	0.009
LQT3 vs LQT1	0.51 (0.33–0.79)	0.003

HR indicates hazard ratio; LQT, long QT; and LQTS, congenital long QT syndrome.

^*Multivariable analysis was carried out with Anderson–Gill modeling, with the normal menarche‐onset cohort assessed as the reference covariates in the multivariable model. Findings are adjusted for time‐dependent β‐blocker therapy, time‐dependent oral contraceptive, prolonged QTc (QTc ≥550 ms) at baseline, genotype (LQT1, LQT2, LQT3, and other LQT genotypes [LQT4–LQT9]), and birth decade.

Additional covariates associated with increased risk for recurrent CE in the multivariable model included a prolonged QTc duration (≥550 ms), which was associated with a significant 57% risk increase (P<0.001) and LQT2 versus LQT1 genotype, which was associated with 35% risk increase (P=0.009), whereas LQT3 versus LQT1 was associated with a 49% lower risk (P=0.003) (Table 2).

Beta Blockers Use and CE Burden

Multivariable analysis demonstrated that BB (assessed as a time‐dependent variable) was associated with a significant, nearly 70% risk reduction for recurrent CE (HR, 0.31 [95% CI, 0.22–0.44], P<0.001) (Table 2). Interaction‐term analysis revealed that the protective effect of BB therapy was consistent across menarcheal ages (all P values for interaction >0.10).

DISCUSSION

To our knowledge, the present study is the first to assess the association between the age at menarche and the burden of CE after the onset of adolescence in women with congenital LQTS. Our findings suggest several important clinical implications regarding risk assessment of women with this inherited cardiac disorder. (1) Women with LQTS and a late menarche (age >16 years) may experience an increased risk for recurrent CE compared with women with normal or early menarche. These findings are independent of QTc duration, genotype, BB, and oral contraceptive use. (2) Nonetheless, no difference in CE was observed between women with LQTS and early menarche (age <12 years) and those with normal menarche. Finally (3) BB therapy is associated with a significant and pronounced reduction in the risk of recurrent CE in adolescent women with LQTS regardless of age at menarche.

Importantly, although the number of patients in the late menarche group in this study was small (N=21), the number of subjects that experienced a cardiac event (N=13), as well as the total number of events they had experienced (N=160), allowed for reasonable parameter estimation in the regression models employed. The study was underpowered to assess associations between age at menarche and ever experiencing a cardiac event; however, the trend of this association was in the same direction as for recurrent events, thereby further supporting the consistency of our findings.

Sex Hormones and Cardiac Ion Channel Activity

Progesterone has been shown to shorten the action potential duration and protect against arrhythmias. ¹³ In contrast, estrogen has been shown to exhibit both acute and genomic effects on the I_Kr channel, including reduction in channel expression and prolongation of ventricular repolarization, resulting in prominent QTc prolongation and an increased risk of arrhythmia. ⁶ , ¹⁴ , ¹⁵ Several previous reports have shown that women with LQTS experience a more pronounced increase in the risk of CE after the onset of adolescence, suggesting a modulating effect of female sex hormones on arrhythmic risk, especially in women with I_kr mutations. ¹ , ²

At birth, levels of sex hormones are high, but they decrease within a few months and remain low until puberty. Early in puberty, levels of luteinizing hormone and follicle‐stimulating hormone increase, stimulating the production of these sex hormones. The increased levels of sex hormones persist during adulthood, with notable variability during menstrual cycles, pregnancy, and postpartum periods. As women age, sex hormone production declines, leading to premenopause followed by menopause. ¹⁶ We have recently demonstrated that the risk of CE is highest during puberty and the postpartum period. However, the absolute estrogen levels at the beginning of puberty are usually still lower than during maternity and are remarkably lower during the postpartum period than they are during pregnancy. ¹⁷ These findings suggest the risk for CE in women with LQTS may be influenced by the relationship between estrogen and progesterone and their relative levels, rather than by their absolute hormonal levels.

Age at Menarche and Cardiac Events

In women without LQTS, early menarche and late menarche were both associated with increased risk for cardiovascular disease, and the relationship between age at menarche and CE was U shaped. ¹¹ , ¹⁸ Yet, the exact mechanism to explain this correlation remains unknown. In the current study, we report for the first time that the association between age at menarche and the risk of CE in patients with LQTS may differ from that observed in the general population. Early menarche was not associated with a statistically significant increased risk of CE, whereas late menarche was associated with a higher risk of CE. Several possible mechanisms may explain these findings. (1) Most causes of late menarche are associated with anovulation, thus limiting the protective effects progesterone has been shown to have on the QTc interval, while often increasing duration and quantity of estrogen exposure. (2) Genetic studies have indicated that the genetic contribution may be more important than any other environmental effects in determining the age of menarche, because 53% to 74% of the variation in age of menarche can be attributed to genetic factors, which may play a significant role especially in patients with LQTS, where genetic variation is the most important determinant of the course of the disease. ¹⁹ Finally, (3) excessive exercise is a cause of delayed menarche (secondary amenorrhea) and is a known trigger for CE among patients with LQTS, especially LQTS1. ²⁰ , ²¹

LIMITATIONS

Although the onset of menstruation is known to be associated with varying levels of estrogen, progesterone, and androgens, these levels may fluctuate significantly between patients, and hormonal blood levels were not collected for this study. The majority of the CE in the present study were recurrent syncope, which may also include events that are nonarrhythmic in nature and yet presented in a clinical manifestation suggestive of arrhythmic syncope. To minimize the bias associated with this limitation, syncopal events were categorized by the study specialists. Thus, nonabrupt events associated with prodromal symptoms of dizziness or lightheadedness, as well as those associated with seizures, were not included in the present analysis. Similar to other studies based on questionnaires, the respondents to the questionnaire may represent a lower‐risk group of patients with LQTS, as high‐risk patients may be less likely to respond due to disease‐related morbidity. In addition, sample size limitations, especially in the late menarche group, warrant further confirmatory studies in this field.

CONCLUSIONS AND CLINICAL IMPLICATIONS

In women with congenital LQTS, late menarche is associated with a higher burden of cardiac events. We hypothesize this may be due to differences in the exposure of these women to estrogen and progesterone compared with women with early or normal onset of menarche. Genetics might also be a crucial factor. Moreover, this study found that BB treatment was associated with a lower burden of CE regardless of the age of menarche. The findings of this study warrant further validation, as well as assessment of the hypothesized factors that may influence the risk of developing CE events after the onset of adolescence among women with LQTS.

Sources of Funding

This study was supported through the American Heart Association Arrhythmias Sudden Cardiac Death Strategically Focused Research Network Grant 19SFRN34930007.

Disclosures

None.

Supporting information

Table S1

STROBE Checklist