"Around 6 months" is the modern guideline. The lived picture is wider: a 10-week introduction window in which most families land, three distinct introduction patterns, and a parent-confidence curve that lags the food-variety curve by about three weeks.
The current AAP guidance is to introduce complementary foods around 6 months of age, with developmental readiness — head and trunk control, the ability to sit with support, loss of the extrusion reflex, and showing interest in food — as the practical determinant1. The CDC echoes this and is explicit that the range from about 4 to 6 months is appropriate provided readiness signs are present5, while the WHO continues to recommend exclusive breastfeeding for the first 6 months globally as a population-level guideline6. The 2020-2025 USDA/HHS Dietary Guidelines for Americans, which for the first time covered birth-to-24-months, similarly anchored introduction "around 6 months" but acknowledged the developmental range4.
For families and pediatricians, the practical question is rarely "what is the policy guideline" but "where does our family fall in the range, and is that fine?" The Wermom data offers a population view of where families actually land.
Across the full cohort, the median age at first non-milk food was week 25 — close to 5 months and 3 weeks. The 25th percentile landed at week 22 (just past 5 months) and the 75th percentile at week 28 (close to 6.5 months). The 10th and 90th percentiles bracketed weeks 19 and 32, respectively.
The shape of the curve has practical relevance for the 4-month and 6-month well-child visits. At the 4-month visit (around week 17), introduction is uncommon — fewer than 6% of the cohort had offered a non-milk food. By the 6-month visit (around week 26), roughly 60% had introduced something. By the 9-month visit (around week 39), 96% had introduced solids and 78% had logged five or more distinct foods. Pediatricians who frame the 6-month visit as "you should be starting around now" are aligned with the median, but four in ten of the families they see at that visit have not yet started — and most of those will start in the following six weeks, which is well within the developmentally normal window15.
The plurality of families followed the developmental-cue model favored by AAP guidance1. In the Wermom data, on-cue introducers logged at least two of the four canonical readiness cues — head and trunk control, sitting with support, food interest, loss of tongue-thrust — within a 2-week window before introducing a first food. Their median first-food age was week 26, with a tight interquartile band of weeks 24–27.
Two features stood out. First, this group reached "five distinct foods" within a median of 18 days after introduction — faster than either of the other two patterns. Second, their parent-confidence curve was the steepest: from baseline 2.1/5 mean confidence at week 22 to 4.2/5 at week 32. The interpretation is consistent with the developmental literature: when introduction is initiated after readiness cues are clearly present, the early feeding experience tends to be more successful, which reinforces parent confidence and supports faster variety expansion7.
For pediatricians, the on-cue group is the easiest population to support: the AAP/ESPGHAN guidance maps cleanly onto their behavior17, and the most useful 4-month-visit counseling is to enumerate the cues to watch for and to set the expectation that readiness will probably appear between weeks 22 and 28. Wermom's parent guide on introducing water alongside solids tracks this same readiness framework.
About 1 in 4 families introduced before week 24 — most commonly between weeks 18 and 23 — with cautious, single-ingredient feedings (typically iron-fortified infant cereal or single-ingredient purees) at low volumes. The most frequently logged reasons were perceived reflux relief, perceived inadequate satiety from milk feeds, and explicit pediatrician recommendation in cases of slowed weight-for-age trajectories.
Importantly, this group's behavior does not contradict guidance. The AAP and CDC both acknowledge that the developmentally appropriate window starts as early as 4 months when readiness is present15, and ESPGHAN explicitly recommends introducing complementary foods between 17 and 26 weeks, never before 17 weeks and never later than 26 weeks for non-breastfed infants7. The early-cautious pattern fits within that range.
What was notable in the data is that this group's variety expansion was slower than the on-cue group's. Their median time from first food to "five distinct foods" was 32 days, almost twice as long. Parent confidence was also lower across the introduction window — a mean of 2.7/5 at week 24 versus the on-cue group's 3.4/5. One plausible reading is that earlier introduction, when developmental cues are partially present rather than fully present, makes early feeding harder and produces more "is this normal?" moments. We did not collect outcome data to test this directly.
About 1 in 4 families introduced at or after week 28. The two most common parent-reported reasons were a goal of exclusive breastfeeding to ~6 months (aligned with WHO guidance6) and a recent family illness disruption (a sick parent, a sick infant, or a household move) that pushed introduction by a few weeks.
This group looked clinically fine in the descriptive aggregates: their growth-curve percentile tracking (where parents logged it) was indistinguishable from the on-cue group through week 36, and their variety expansion, once started, caught up by week 36 with no statistical separation from the population mean by week 39. The one practical caution worth noting is iron sufficiency: AAP and CDC guidance both emphasize that iron-fortified or iron-rich foods (meat, iron-fortified cereal, beans) should be among the first foods, because infant iron stores from birth begin to deplete by about 6 months15. Deliberate-late introducers who continue exclusive milk feeding past week 28 are the subgroup most worth explicit iron counseling. Wermom's parent guide on iron deficiency in infants 6–24 months covers this directly.
One of the highest-stakes parts of complementary feeding is allergen introduction, particularly peanut and egg. The 2017 NIAID-sponsored Addendum Guidelines and the 2019 AAP clinical report on peanut allergy prevention together synthesized the LEAP and EAT trial evidence23: for infants at high risk (severe eczema and/or egg allergy), introduce peanut between 4 and 6 months under appropriate medical supervision; for moderate-risk infants (mild-moderate eczema), introduce peanut around 6 months; for low-risk infants (no eczema, no egg allergy), introduce peanut along with other solids per family choice. The data is strong: early introduction reduces peanut allergy prevalence by approximately 80% in high-risk infants compared to delayed introduction3.
In the Wermom cohort, 71% of families had introduced peanut-containing food by week 32 and 84% by week 39 — broadly consistent with guideline uptake, but with a clear lag in the deliberate-late and early-cautious groups. Egg was introduced earlier than peanut on average (median week 28 vs. week 30), likely reflecting parent comfort with cooked egg in baby cereals and pancakes. Importantly, only 6% of cohort families with a logged eczema flag (whose infants were therefore "high-risk" by the NIAID criteria) had introduced peanut by week 24 — a gap that is clinically meaningful, because the LEAP-derived guidance specifically targets that 4-to-6-month window for the high-risk group23.
One of the most consistent patterns across all three introduction groups was the lag between food-variety expansion and parent-reported confidence. On average, families reached the "five distinct foods" threshold three weeks before they crossed the "I feel confident about feeding" self-report threshold. The gap was largest in the early-cautious group (4–5 weeks) and smallest in the on-cue group (1–2 weeks).
This has implications for the 6-month and 9-month well-child visits. Parents who present as "still figuring it out" at month 7 are statistically normal — their infant's variety has expanded ahead of their own confidence. Concrete, non-judgmental counseling that names this lag ("most families feel uncertain for about three weeks after their baby is already eating several foods well") tends to land better than reassurance phrased as "you're doing fine," which parents often interpret as dismissive.
Several limitations deserve careful attention before applying these findings clinically.
Self-selection. Wermom App users are not a representative US infant population. They skew toward parents with higher digital engagement, higher day-to-day tracking interest, and — based on app analytics — higher educational attainment and more urban geography. Introduction timing in less-tracked populations may differ; the WIC-eligible literature, for example, documents earlier introduction in some demographic subgroups4.
Definitional sensitivity. "First non-milk food" includes infant cereal mixed with milk in a bottle — a practice that some pediatricians do not consider a solid food introduction. Excluding bottle-cereal feeds shifts the population median later by approximately 8 days; the trajectory shapes are robust to this choice.
Confidence measurement. A single 1–5 Likert is a crude instrument compared to validated parent-feeding-confidence scales such as the Lansigan Confidence in Feeding Index. We chose the simpler measure for in-app completion rates; results should be interpreted as directional.
Causal inference. This is descriptive epidemiology, not a trial. We do not infer that any introduction pattern causes any outcome. We do think the population view is useful as a normalization framework — for parents wondering "are we behind?" and for clinicians who want a cohort-level reference point.
Future work. We plan a follow-on analysis pairing this cohort with parent-logged eczema, reflux, and food-reaction events to characterize the temporal relationship between introduction timing, the order in which allergens were introduced, and subsequent atopic events. That analysis is in design review as of May 2026 and will require additional clinical review before publication.
The Wermom App's feeding tracker uses the same definitions referenced in this research and surfaces your baby's introduction pattern against AAP-aligned norms.
Explore the Wermom App