Abstract
This matched-cohort study uses data from the French National Health Insurance database to assess whether a 19.5-mg levonorgestrel intrauterine system, vs a 52-mg system, is associated with increased use of antidepressant, hypnotic, and anxiolytic medications.
Studies have shown associations between levonorgestrel intrauterine systems and depression,1 antidepressant use,2 or anxiolytic use.3 These studies all focused on the use of the intrauterine system containing 52 mg of levonorgestrel, recommended in France for contraception and some gynecologic conditions such as endometriosis. A 19.5-mg levonorgestrel intrauterine system, indicated for contraceptive use, has been marketed in France since 2018 and was used by 30% of levonorgestrel intrauterine system users in 2019.4 We hypothesized that the dose of levonorgestrel may be associated with the incidence of psychotropic drug use, because the low-dose devices may result in lower serum progestin levels.5
Methods
We conducted a nationwide matched-cohort study using data from the French National Health Insurance database (SNDS), which provides information on health insurance claims for 99% of the population in France. We identified females aged 13 to 40 years who did not use a levonorgestrel intrauterine system in the 6 years before 2019 or psychotropic drugs in the previous year. Females dispensed a 52-mg levonorgestrel intrauterine system in 2019 and a comparator group who received a 19.5-mg levonorgestrel intrauterine system were matched (1:1) for age (±3 years), gravidity in the 10 years prior to the study, month of device dispensation (±3 months), and physician specialty. Standardized differences were used to assess matching, with a difference of ±10% indicating a good match. Conditional logistic regression was used to calculate odds ratios, adjusted for sociodemographic characteristics, childbirth in the year before intrauterine system insertion, medical history (gynecological/vascular/oncological history, obesity, use of smoking cessation drugs), and the matching variables to assess the associations between the type of levonorgestrel intrauterine system and at least 2 prescriptions for anxiolytics, hypnotics, and/or antidepressants within 2 years after each dispensing of each device. The final date of follow-up was December 30, 2021. We conducted sensitivity analyses among subgroups of different ages (testing for interactions) and among females without medical history. A 95% CI that did not cross 1 indicated statistical significance; P < .05 (2-sided) was used to assess significant interactions. Analyses were performed with SAS version 9.4 (SAS Institute Inc). The research group has permanent access to the anonymized data from the SNDS according to a specific French regulation and thus is exempted from institutional review board approval.
Results
In total, 45 736 females who received a 52-mg levonorgestrel intrauterine system and 45 736 who received a 19.5-mg levonorgestrel intrauterine system (mean age, 32.3 [SD, 5.4] and 31.8 [SD, 5.4] years, respectively) were well matched on most characteristics (Table 1). After adjustment, dispensation of a 52-mg levonorgestrel intrauterine system vs a 19.5-mg levonorgestrel intrauterine system was associated with a statistically significantly increased use of an antidepressant within 2 years (4.0% for the 52-mg group, 3.6% for the 19.5-mg group; adjusted odds ratio [aOR], 1.13 [95% CI, 1.05-1.21]) but not with use of anxiolytics (5.2% vs 4.9%; aOR, 1.05 [95% CI, 0.99-1.12]) or hypnotics (2.9% vs 2.6%; aOR, 1.09 [95% CI, 0.99-1.19]) (Table 2). The association with antidepressant use persisted in females without a medical history (3.8% vs 3.3%; aOR, 1.19 [95% CI, 1.09-1.31]). Females younger than 25 years using a 52-mg vs 19.5-mg levonorgestrel intrauterine system had a significantly increased risk of using hypnotics (3.2% vs 2.2%; aOR, 1.61 [95% CI, 1.04-2.48]; P = .04 for interaction). For other age categories, the associations between levonorgestrel dose and outcomes did not reach statistical significance.
Table 1. Characteristics of Females Receiving a 19.5-mg or 52-mg Levonorgestrel IUS in 2019 After Matching.
| Characteristic | Levonorgestrel IUS, No. (%) | Standardized differencea | |
|---|---|---|---|
| 19.5 mg (n = 45 736) | 52 mg (n = 45 736) | ||
| Age, yb | −0.099 | ||
| Mean (SD) | 31.8 (5.4) | 32.3 (5.4) | |
| Median (IQR) | 32.0 (28 to 36) | 33.0 (29 to 37) | |
| Month of dispensationb | 0.075 | ||
| 1 | 3954 (8.6) | 4557 (10.0) | |
| 2 | 3432 (7.5) | 3850 (8.4) | |
| 3 | 3951 (8.6) | 4227 (9.2) | |
| 4 | 3733 (8.2) | 3883 (8.5) | |
| 5 | 3880 (8.5) | 3857 (8.4) | |
| 6 | 3626 (7.9) | 3518 (7.7) | |
| 7 | 3685 (8.1) | 3604 (7.9) | |
| 8 | 2884 (6.3) | 2937 (6.4) | |
| 9 | 4490 (10.0) | 4325 (9.5) | |
| 10 | 4369 (9.6) | 4047 (8.8) | |
| 11 | 4094 (9.0) | 3766 (8.2) | |
| 12 | 3538 (7.7) | 3145 (6.9) | |
| Physician specialtyb | 0.000 | ||
| General practitioner | 11 009 (24.1) | 11 009 (24.1) | |
| Gynecologist | 26 884 (58.8) | 26 884 (58.8) | |
| Midwife | 7843 (17.1) | 7843 (17.1) | |
| Gravidity in the 10 y beforeb | 0.000 | ||
| 0 | 8348 (18.3) | 8348 (18.3) | |
| 1 | 17 152 (37.5) | 17 152 (37.5) | |
| ≥2 | 20 236 (44.2) | 20 236 (44.2) | |
| Universal health insurance coveragec | 5595 (12.2) | 6600 (14.4) | −0.065 |
| Social Deprivation Index | −0.071 | ||
| 1 (least deprived) | 8564 (18.7) | 8709 (19.0) | |
| 2 | 10 014 (21.9) | 9131 (20.0) | |
| 3 | 9524 (20.8) | 8865 (19.4) | |
| 4 | 8560 (18.7) | 8591 (18.8) | |
| 5 (most deprived) | 7194 (15.7) | 8036 (17.6) | |
| Missing | 1880 (4.1) | 2404 (5.3) | |
| Use of smoking cessation drugsd | 2507 (5.5) | 2829 (6.2) | −0.030 |
| Indicators of obesityd | 2792 (6.1) | 3680 (8.0) | −0.076 |
| Cardiovascular historyd | −0.011 | ||
| High blood pressure | 875 (1.9) | 931 (2.0) | |
| Dyslipidemia | 184 (0.4) | 220 (0.5) | |
| Myocardial infarction | 16 (<0.1) | 27 (0.1) | |
| Stroke | 149 (0.3) | 150 (0.3) | |
| Thrombosis | 114 (0.2) | 117 (0.3) | |
| Pulmonary embolism | 76 (0.2) | 91 (0.2) | |
| Gynecological historyd | −0.091 | ||
| Endometriosis | 516 (1.1) | 1022 (2.2) | |
| Heavy menstrual bleeding | 250 (0.5) | 356 (0.8) | |
| Polyp/myoma | 228 (0.5) | 308 (0.7) | |
| Endometrial hyperplasia | 81 (0.2) | 156 (0.3) | |
| Oncological historyd | −0.005 | ||
| Breast cancer | 8 (<0.1) | 9 (<0.1) | |
| Endometrial cancer | 1 (<0.1) | 0 | |
| Ovarian cancer | 8 (<0.1) | 5 (<0.1) | |
| Childbirth in the year before insertion | 18 536 (40.6) | 18 078 (39.5) | −0.014 |
Abbreviation: IUS, intrauterine system.
Threshold for standardized differences is ±0.1.
Matching variables.
Provides health insurance coverage for low-income persons.
In the preceding 10 years.
Table 2. Associations Between Receiving a 19.5-mg vs a 52-mg Levonorgestrel Intrauterine System in 2019 and the Use of Anxiolytics, Hypnotics, or Antidepressants in the 2 Years After Dispensation.
| Levonorgestrel intrauterine system | Receipt of 52-mg vs 19.5-mg system, OR (95% CI) | |||
|---|---|---|---|---|
| 19.5 mg | 52 mg | Crude | Adjusteda | |
| All | ||||
| No. | 45 736 | 45 736 | ||
| Anxiolytic use | 2222 (4.9) | 2387 (5.2) | 1.09 (1.02-1.16) | 1.05 (0.99-1.12) |
| Hypnotic use | 1184 (2.6) | 1315 (2.9) | 1.13 (1.04-1.23) | 1.09 (0.99-1.19) |
| Antidepressant use | 1627 (3.6) | 1821 (4.0) | 1.15 (1.07-1.23) | 1.13 (1.05-1.21) |
| Females without relevant medical history b , c | ||||
| No. | 31 936 | 31 936 | ||
| Anxiolytic use | 1441 (4.5) | 1548 (4.9) | 1.08 (1.00-1.17) | 1.08 (1.00-1.17) |
| Hypnotic use | 786 (2.5) | 862 (2.7) | 1.12 (1.01-1.24) | 1.08 (0.97-1.21) |
| Antidepressant use | 1031 (3.2) | 1210 (3.8) | 1.20 (1.10-1.31) | 1.19 (1.09-1.31) |
| Females aged <25 y c , d | ||||
| No. | 3782 | 3782 | ||
| Anxiolytic use | 188 (5.0) | 186 (4.9) | 1.02 (0.81-1.28) | 1.00 (0.77-1.28) |
| Hypnotic use | 84 (2.2) | 124 (3.3) | 1.58 (1.15-2.17) | 1.61 (1.04-2.48) |
| Antidepressant use | 110 (2.9) | 137 (3.6) | 1.32 (0.99-1.75) | 1.37 (0.97-1.93) |
| Females aged 25-34 y c , d | ||||
| No. | 19 455 | 19 455 | ||
| Anxiolytic use | 897 (4.6) | 947 (4.9) | 1.06 (0.96-1.17) | 1.01 (0.91-1.13) |
| Hypnotic use | 469 (2.4) | 516 (2.6) | 1.11 (0.97-1.27) | 1.06 (0.91-1.23) |
| Antidepressant use | 665 (3.4) | 695 (3.6) | 1.06 (0.95-1.19) | 1.06 (0.94-1.19) |
| Females aged ≥35 y c , d | ||||
| No. | 14 510 | 14 510 | ||
| Anxiolytic use | 779 (5.4) | 852 (5.9) | 1.11 (1.00-1.23) | 1.09 (0.98-1.22) |
| Hypnotic use | 416 (2.9) | 444 (3.1) | 1.09 (0.95-1.26) | 1.07 (0.91-1.25) |
| Antidepressant use | 598 (4.1) | 644 (4.4) | 1.10 (0.98-1.24) | 1.11 (0.98-1.26) |
Abbreviation: OR, odds ratio.
Adjusted for matching variables (age in 3 classes [13-24, 25-34, ≥35 years], month of dispensation, physician specialty, gravidity), universal health coverage, social deprivation index, vascular, oncological and gynecological history (endometriosis, heavy menstrual bleeding, endometrial hyperplasia, polyp/myoma), obesity, use of smoking cessation drugs, and childbirth in the prior year.
Subgroup of females without obesity, without history of use of smoking cessation drugs, and without arterial, venous, oncological, or gynecological history.
Only pairs for which the 2 individuals shared the same subgroup’s characteristic were retained in the analysis.
Interaction between age and anxiolytic use (P = .72), antidepressants use (P = .51), and hypnotics use (P = .04).
Discussion
This nationwide study in France found a low but statistically significant increased risk of antidepressant use associated with a 52-mg vs a 19.5-mg levonorgestrel intrauterine system; associations with the use of anxiolytics and hypnotics did not reach statistical significance. The difference in absolute percentages of antidepressant use were small and unlikely to be clinically important at an individual level. However, this result could be important on a population level. The lack of a strong dose-response relationship may suggest that levonorgestrel, and thereby serum progestin, does not markedly cause mood disturbances in females using a levonorgestrel intrauterine system. Further studies are needed.
The main limitation of the study is confounding by indication,6 which could account for the increased risk of antidepressant use. However, excluding females with previous use of psychotropic drugs, focusing on those younger than 40 years, and matching for relevant characteristics and adjusting for known and potential confounding factors should minimize bias.
Section Editors: Jody W. Zylke, MD, Deputy Editor; Kristin Walter, MD, Senior Editor.
Data Sharing Statement
References
- 1.Andersson K, Odlind V, Rybo G. Levonorgestrel-releasing and copper-releasing (Nova T) IUDs during five years of use: a randomized comparative trial. Contraception. 1994;49(1):56-72. doi: 10.1016/0010-7824(94)90109-0 [DOI] [PubMed] [Google Scholar]
- 2.Skovlund CW, Mørch LS, Kessing LV, Lidegaard Ø. Association of hormonal contraception with depression. JAMA Psychiatry. 2016;73(11):1154-1162. doi: 10.1001/jamapsychiatry.2016.2387 [DOI] [PubMed] [Google Scholar]
- 3.Bosco-Lévy P, Gouverneur A, Langlade C, Miremont G, Pariente A. Safety of levonorgestrel 52 mg intrauterine system compared to copper intrauterine device: a population-based cohort study. Contraception. 2019;99(6):345-349. doi: 10.1016/j.contraception.2019.02.011 [DOI] [PubMed] [Google Scholar]
- 4.Roland N, Baricault B, Dray-Spira R, et al. Profiles of copper intrauterine devices and levonorgestrel intrauterine systems users in France in 2019. Int J Gynaecol Obstet. 2022. Published online September 6, 2022. doi: 10.1002/ijgo.14438 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Goldstuck ND. Clarification of the role of the Jaydess(Skyla) LNG- IUS 13.5mg and Kyleena LNG-IUS 19.5mg as intrauterine contraceptive systems. Expert Rev Med Devices. 2017;14(8):593-599. doi: 10.1080/17434440.2017.1350169 [DOI] [PubMed] [Google Scholar]
- 6.Bastianelli C, Farris M, Rosato E, et al. The use of different doses levonorgestrel-releasing intrauterine system (LNG-IUS): real-world data from a multicenter Italian study. Eur J Contracept Reprod Health Care. 2022;27(1):16-22. doi: 10.1080/13625187.2021.1975269 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Sharing Statement