Long-Term Outcome in Patients With Heart Failure Treated With Levothyroxine

An Observational Nationwide Cohort Study

Mette Nygaard Einfeldt; Anne-Marie Schjerning Olsen; Søren Lund Kristensen; Usman Khalid; Jens Faber; Christian Torp-Pedersen; Gunnar H Gislason; Christian Selmer

Disclosures

J Clin Endocrinol Metab. 2019;104(5):1725-1734. 

In This Article

Results

We identified a total of 224,670 patients from 1997 to 2012 discharged and alive 30 days after a first-time hospitalization for HF. The mean age was 70.7 years (SD 14.7), and 53% were men. Selection of the study cohort is illustrated in Figure 1, and the baseline characteristics of the study sample and distribution between the groups are presented in Table 2.

Figure 1.

Study population selection flowchart.

The patients treated with L-T4 at baseline had more comorbidities and were more likely to receive other medical therapy compared with the two other groups. The groups treated with L-T4 at baseline and during follow-up consisted primarily of women compared with the overall study population (79% and 68% vs 47%). The group treated with L-T4 at baseline was older than the overall study population (77.6 years vs 70.7 years).

A total of 9007 (4.0%) patients had at least one prescription for L-T4 after discharge, with an average treatment time of 64 (SD 450) days and an average prescribed daily dosage of 86.9 (SD 33.8) micrograms.

During a median follow-up time of 4.8 years (IQR 9.2) 35,725 MIs and 122,881 MACEs were registered and a total of 147,253 (66%) patients died, with 102,818 of these deaths due to cardiovascular causes. Causes of death are presented in Table 3.

A total of 1085 patients were censored due to emigration, and 14,697 patients were censored due to a prescription of amiodarone during the study period.

As a consequence of a significant interaction between L-T4 treatment and sex (P<0.0001) and L-T4 treatment and age (P<0.0001), we stratified the model on both sex and age in the main analyses.

All-cause Mortality

The unadjusted IRs for all-cause mortality were 86.1/1000 py for the group that was not treated, 194.9/1000 py for the group that was treated with L-T4 at baseline, and 112.3/1000 py for the group that initiated L-T4 during follow-up. The adjusted IRR showed that treatment with L-T4 at baseline was associated with a 25% (95% CI, 1.21 to 1.29) increased risk of death and that initiation of L-T4 treatment during follow-up was associated with a 13% (95% CI, 1.10 to 1.16) increased risk of death. The association between L-T4 treatment and risk of death was even more pronounced when we looked at the subdivided age groups. The adjusted IRR showed that treatment with L-T4 at baseline for patients < 65 years old was associated with a 58% (95% CI, 1.41 to 1.78) increased risk of death and a 24% (95% CI, 1.20 to 1.28) increase for patients ≥65 years old. As for the patients who initiated L-T4 treatment during follow-up, we observed an association between L-T4 treatment and an 8% (95% CI, 1.02 to 1.15) increased risk of death for patients <65 years old and a 15% (95% CI, 1.11 to 1.18) increase for patients ≥65 years old. Furthermore, in men, treatment with L-T4 was associated with a higher risk of all-cause mortality compared with in women. IRs are shown in Table 4, and IRRs are shown in Figure 2.

Figure 2.

IRRs for (a, e) all-cause mortality, (b, f) cardiovascular death, (c, g) MI, and (d, h) MACE, stratified on sex and age (<65 years and ≥65 years). Adjusted for age, sex, Charlson Comorbidity Index, and calendar year.

Cardiovascular Death

The unadjusted IRs for cardiovascular related death were 60.2/1000 py for the group that was not treated, 121.1/1000 py for the group treated with L-T4 at baseline, and 77.2/1000 py for the group that initiated L-T4 during follow-up. The adjusted IRR showed that treatment with L-T4 at baseline was associated with a 23% (95% CI, 1.18 to 1.27) increased risk of cardiovascular death and that initiation of L-T4 during follow-up was associated with an 11% (95% CI, 1.08 to 1.15) increased risk of cardiovascular death. The heightened risk of cardiovascular death was further increased among patients <65 years old for those treated at baseline (55%; 95%CI, 1.33 to 1.80) but not for those who initiated L-T4 during the study period (11%; 95% CI, 1.03 to 1.20). Treatment with L-T4 at baseline for patients ≥65 years old was associated with increased risk of cardiovascular death (21%; 95% CI, 1.17 to 1.26) and for patients who initiated L-T4 during follow-up (13%; 95% CI, 1.09 to 1.18). Furthermore, in men treatment with L-T4 was associated with a higher risk of cardiovascular death compared with L-T4 treatment in women. IRs are shown in Table 4, and IRRs are shown in Figure 2.

MI

The unadjusted IRs for MI were 22.2/1000 py for the group that was not treated, 45.7/1000 py for the group treated with L-T4 at baseline, and 18.7/1000 py for the group that initiated L-T4 during follow-up. The adjusted IRR showed that treatment with L-T4 at baseline was associated with a 32% (95% CI, 1.23 to 1.41) increased risk of MI but with a significantly reduced risk of MI in the group that initiated L-T4 during follow-up. The lowered risk of MI remained significant for those who initiated L-T4 during follow-up for patients ≥65 years old and for women in general. However, the association of L-T4 treatment during follow-up with lowered risk of MI was no longer significant among patients <65 years old and for men in general. The significantly increased risk of MI observed for patients treated at baseline remained significant for patients ≥65 years old (30%; 95% CI, 1.22 to 1.40) and was further increased for patients <65 years old (43%; 95% CI, 1.17 to 1.76). L-T4 treatment at baseline was associated with an increased risk of MI in both women (28%; 95% CI, 1.18 to 1.38) and men (35%; 95% CI, 1.18 to 1.54). IRs are shown in Table 4, and IRRs are shown in Figure 2.

MACE

The unadjusted IRs for MACE were 78.6/1000 py for the group that was not treated, 174.5/1000 py for the group treated with L-T4 at baseline, and 91.3/1000 py for the group that initiated L-T4 during follow-up. The adjusted IRR showed that L-T4 treatment at baseline was associated with a 26% (95% CI, 1.22 to 1.31) increased risk of MACE and a 5%(95% CI, 1.02 to 1.09) increased risk of MACE in the group that initiated L-T4 during follow-up. The association between L-T4 treatment at baseline and increased risk of MACE was further increased for patients <65 years old (38%; 95% CI, 1.22 to 1.57) and for patients <65 years old who initiated L-T4 during follow-up (8%; 95% CI, 1.01 to 1.16). A similar association was observed between L-T4 treatment and increased risk of MACE in patients ≥65 years old. L-T4 treatment was associated with a higher increased risk of MACE in men when compared with the risk observed in women. IRs are shown in Table 4, and IRRs are shown in Figure 2.

Sensitivity Analysis

Adjusting the main model for ischemic heart disease and loop diuretic status at baseline did not alter the result obtained with the main model, nor did the sensitivity analysis including only patients with a Charlson Comorbidity Index of 0 (no known comorbidities). Adjusting the main model for socioeconomic status showed no difference from our main results except when we looked at MI, where the results were no longer significant. When we included patients who did not survive beyond the first 30 days after their first HF discharge, the results did not differ significantly from the main results. Dividing the follow-up time into smaller intervals yielded no differences in the main results. The results from the sensitivity analyses are presented in Table 5.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....