Pill, IUD, Implant, Ring & Depo — Who Uses Them, and Is That Shifting?
Age-stratified use of five hormonal and LARC contraceptive methods, across a four-year retrospective calendar. Each respondent's age is fixed at the time of interview.
NSFG 2022–2023N = 5,586 female respondentsMETHX calendar · 4 years × 5 methods × 7 age groups
Age Profile — Interview Year (Current Snapshot)
For each method: share of each age group that used it during the interview year. Bars are scaled to each method's own maximum — do not compare widths across methods.
Contraceptive market overview — full bar = all respondents in age group
No method used
Other methods (condom, withdrawal, sterilization, rhythm…)
Non-EC contraceptive drug (pill, IUD, implant, Depo, ring, patch)
Not recently sexually active
Sexually active, no contraception
Other methods (condom, withdrawal…)
Non-EC drug (pill, IUD, implant, Depo…)
Methodology — overview bar
Window: Interview year only — METHX calendar months 37–48 (the 12 months immediately preceding each respondent's interview date). Method classification: A respondent is assigned to Non-EC drug if any METHX slot in months 37–48 equals pill 3, Depo 8, implant 9, IUD 19, patch 21/25, or ring 22/26. Other methods = any remaining non-zero, non-continuation code (condom, withdrawal, sterilization, rhythm, etc.). No method = all slots are code 1 or empty. Sexual activity split (toggle): Uses HADSEX and LSEXDATE (century-month of last sexual intercourse). Sexually active = HADSEX = 1 AND CMINTVW − LSEXDATE ≤ 12. Not recently active = HADSEX = 2 (never had sex) OR last sex more than 12 months before interview. This definition is specific to the interview-year window; sexual activity status in prior years (yr−3 through yr−1) cannot be reconstructed from LSEXDATE alone, as NSFG does not include a monthly sexual activity calendar.
Methodology — method bars
Window: Interview year only (same METHX months 37–48 as overview bar above). Denominator toggle:
Method users in age group — respondents who used any contraceptive in the interview year (n_users)
All respondents in age group — all respondents regardless of sexual activity or method use (n_total)
Hormonal/LARC users in age group — respondents who used pill, IUD, implant, Depo, patch, or ring in the interview year (n_drug)
Sexually active that year — respondents who reported vaginal intercourse in at least one month of the interview year (n_active, from EC-8 MONSX monthly calendar). Rate numerator = used method AND sexually active. This denominator varies year-by-year in the 4-year trend charts; the others use interview-year counts for all years.
Counting rule: A respondent is counted as a user of method X if she used it in any month of the 12-month window, regardless of how many months. Bars are scaled to each method's own maximum — widths are not comparable across methods.
Yr −3 → Interview Year: How Each Age Group Shifted
Each row is one age group. Open circle = yr −3, filled circle = interview year. Intermediate dots (yr −2, yr −1) show the trajectory shape. Line color: teal = increasing, red = declining. Rows marked ! have small sample sizes (n < 15) — interpret with caution.
Methodology — 4-year trend
Windows: yr−3 = METHX months 1–12; yr−2 = 13–24; yr−1 = 25–36; interview year = 37–48. Each window covers 12 months. Respondent age is fixed at interview time — a respondent in the 15–19 group was aged 12–16 in yr−3. Denominator: The first three options (method users / all respondents / hormonal+LARC users) use interview-year counts for all four time points — they do not adjust year-by-year for changes in sexual activity. The new "Sexually active that year" option corrects this: it uses EC-8 MONSX1–MONSX48 monthly intercourse indicators to count, for each year independently, how many respondents in the age group reported vaginal intercourse in at least one month of that 12-month window. Numerator = used method AND was sexually active in the same year. For the 15–19 group this matters substantially: in yr−3, respondents were aged 12–16 and only ~55 of 730 reported any sexual activity, versus ~187 in the interview year.
Yr −3
Interview yr
Increasing
Declining
n < 15 at either endpoint — interpret with caution
Left margin = yr−3 n · Right margin = interview yr n & Δ (pp)
What the data shows
Pill: strong youth preference, and it is eroding.
Pill is the default contraceptive for young women — 51% of teen method users rely on it versus 16% of women 45–50. But this gap is not stable: among teens, pill share dropped from 64% in yr −3 to 51% at interview; among 20–24-year-olds from 54% to 39%. Rates in older groups changed little. The age spread is widening because young women are moving away from pill faster than older cohorts are.
This decline is consistent with national trends documented across multiple sources:
· CDC/NCHS Data Brief (2024): oral contraceptive use among women 15–49 fell from ~28% (2006–2008) to ~12% (2022–2023). CDC NCHS DB-539
· New oral contraceptive prescriptions dropped from 25M+ (2019) to under 22M (2022). MDPI Society 2024
· KFF Women's Health Survey (2024): pill use among 18–49-year-olds fell from 33% (2022) to 29% (2024). KFF 2024
IUD: catching up with younger cohorts, and the trend is sustained.
IUD use peaks at ages 25–39 and has historically been low among teens — but every age group under 30 shows consistent growth over four years. Teen IUD uptake among method users more than doubled (3% → 7%); 20–24-year-olds rose from 9% to 13%. The age gradient is flattening. Among women already using a hormonal or LARC method, the IUD's share is rising across all age groups.
The long-term national rise in IUD/LARC use is well-documented:
· CDC/NCHS Data Brief #188 (2013): LARC use among women 15–44 rose from 1.5% (2002) to 7.2% (2011–2013). CDC NCHS DB-188
· CDC/NCHS Data Brief #388 (2020): LARC use reached 10.4% of women 15–49 in 2017–2019. CDC NCHS DB-388
· CDC/NCHS Data Brief #539 (2024): IUD use stands at 6.8%; LARCs collectively at 10.5% in 2022–2023. CDC NCHS DB-539
Implant: sharply concentrated in 20–24, stable over time.
Implant use drops steeply with age and is concentrated in the 20–24 group (~12% of method users). Unlike IUD, the age profile has not changed meaningfully over four years — it appears to occupy a stable demographic niche, with the 15–19 cohort showing very slight growth from a small base.
The youth concentration of implant use is consistent with national data:
· CDC/NCHS Data Brief #388 (2020): implant use among contraceptive-using teens (15–19) rose from 6% (2014) to 16% (2016) — the largest single-group increase across all ages. CDC NCHS DB-388
· PMC/NSFG analysis (2015–2017): implant use among 15–19-year-olds was ~16%, versus under 2% for women 30–49 — one of the steepest age gradients of any contraceptive method. PMC 8820401
Depo (injection): youth-concentrated, declining through the 20s.
Depo is the third most used hormonal method among teens — 12–13% of teen method users rely on it, second only to pill and implant. Use drops steeply with age: 25–29-year-olds fell from 4.4% to 2.9% over four years, the sharpest age-specific decline of any method on this page. The 15–19 and 20–24 groups are relatively stable, suggesting Depo occupies a role as a short-term option for young women before they transition to LARCs.
Ring: mid-age preference, small sample, no clear trend.
The ring is the only method here that does not peak in younger women — use is highest among 35–39-year-olds (~3% of method users). With only ~70 users per year in the sample, confidence intervals are wide and no temporal trend is detectable. Its users may skew toward women experienced with hormonal methods who prefer a non-daily option without the commitment of a LARC.
Full Data
Age at interview. In yr −3, women in the 15–19 group were aged 12–16 — many were not yet contraceptively active, so denominators are small and estimates less reliable (shown as dashed lines above). Counting rule: A respondent is counted once per year if she used that method in any month of the 12-month window (METHX1–192 calendar slots, 4 slots per month). Method codes: Pill = code 3, IUD = code 19, Implant = code 9, Ring (NuvaRing) = code 26. Drug-market denominator: women who used any of: pill, IUD, implant, Depo, ring, or patch in that year. Sexual activity split: "Sexually active" = HADSEX=Yes AND last sexual intercourse (LSEXDATE) within 12 months of interview. "Not recently sexually active" = never had sex (HADSEX=No) or last sex >12 months ago. Applies to interview-year overview bar only. Source: NSFG Female Respondent PUF · 2022–2023, N = 5,586.