2026 research update

PUBLISHED BY IS BABY COMING? · 18 OCTOBER 2026

This summary is informational, not medical advice. Decisions about your pregnancy belong with your midwife or doctor. Every claim below sits next to its source, with a note on how strong the evidence is. None of these studies tested or validates any product brand.

Dates in late pregnancy

The broadest credible synthesis remains a 2024 BMC Pregnancy & Childbirth review (48 studies; PMID 38166785). This update relies on it because the 2026 overviews restate the same data without new trials. It linked eating dates in the last weeks to more spontaneous labour starts, less need for induction, and about 1 cm more cervical dilation at admission. Effects on caesarean and instrumental delivery were not statistically significant.

How good is the evidence? Weak to moderate. The review rated most outcomes low-to-moderate certainty, only 4 of 48 studies had low risk of bias, and only 4 reported safety data at all. Dates appear safe as food, but that is not proof of safety as an intervention. There is no established amount: trials used different date types and sizes, and six large Medjool dates are not the same as six small ones. The 2011 study long cited on this site (Al-Kuran) is a prospective observational comparison, not a randomized trial.

One caution: dates are sugary. If you have gestational diabetes, fit them into your care plan with your clinician or dietitian. ACOG's gestational diabetes FAQ explains why.

Birthing balls and peanut balls

The strongest 2026 paper is Luo et al.'s PeerJ meta-analysis (23 randomized trials, 3,192 women; PMID 41943826). Using a birthing ball during labour cut pain scores by about 2 points on a 10-point scale and shortened the second stage by about 17 minutes; the pooled caesarean rate was roughly halved in relative terms. The authors rate certainty moderate-to-low, and not every pooled endpoint includes all 23 trials. The size of any benefit remains uncertain.

Peanut balls are a separate device with a smaller evidence base: trials showed shorter first and second stages but no significant effect on pain or caesarean rates, so the two are not combined here. An earlier 2025 meta-analysis (10 trials) pointed the same direction on pain. No study shows that a ball starts labour. All of this evidence is from during labour; a ball is a comfort tool, not an induction method.

ACOG Committee Opinion 766 (reaffirmed 2025) supports freedom of movement and position for low-risk women in labour, which is the closest thing to official backing a birthing ball has.

Perineal massage before birth

The most useful 2026 primary study is Zhang et al. in Midwifery (PMID 42229096): a propensity-matched cohort of 674 first-time mothers. Matched pairs are not a randomized trial, so the study shows association, not proof. Women who did structured perineal massage from 36 weeks had fewer episiotomies (17.5% vs 26.4%) and, among women without an episiotomy, fewer second-degree tears (75.9% vs 85.1%). The best results appeared after about 5 to 8 sessions. Most women who massaged still had some tearing; massage shifts the odds, it does not protect completely.

A 2026 network meta-analysis in Frontiers in Medicine (31 trials, 10,745 women; PMID 42023089) found that physical techniques (pelvic floor exercise before birth, massage and warm compresses) roughly halved some perineal-injury outcomes, while information-only programmes did nothing. Its massage-plus-warm-compress results partly come from care during labour, so they do not validate antenatal massage products. The ceiling comes from Cochrane 2024 (17 trials, 13,695 women; PMID 39465967): warm compresses during the second stage probably reduce severe tears (moderate certainty), while evidence for second-stage massage is low to very-low certainty. That review excluded techniques outside the second stage, so it does not directly evaluate massage practised before labour.

What matters is the technique, not the lubricant. If you use an oil, patch-test first and avoid known allergens (for example nut-derived oils with a nut allergy). Skip massage and ask your midwife or ob-gyn first if you have placenta previa, vaginal bleeding, any risk of preterm labour, or an active infection, and stop if it hurts or causes bleeding.

What this means for what you buy

Very little. These studies tested a fruit, a ball and a massage technique. The curated picks on the products page are separate editorial choices, with product identity, pack size and practical features checked. No research endorses a brand, because no research tested one.

If you are looking at ways to bring labour on rather than prepare for it, the natural induction evidence review grades each method, from membrane sweeps to castor oil.