Floss, Electric Toothbrush or Waterpik?

About 31.6% of US adults aged 30 and over report flossing daily: that is the NHANES 2011–2014 estimate from 8,356 adults.^[10] Meanwhile, 42% of dentate US adults aged 30+ have periodontitis, and 7.8% have the severe form.^[9] The question “which device is best” is downstream of a much larger problem — most people are not using any of them consistently.

Daily flossing was lower among current tobacco users (OR 0.82, 95% CI 0.68–0.99) and no different in people with diabetes (OR 0.75, 0.52–1.08).^[10] These are two groups with the most to gain from good periodontal care, not two groups already doing more of it.

The 2014 Cochrane review included 56 trials, 51 with usable data, and 4,624 participants, with a minimum four weeks of unsupervised use.^[1] Moderate-certainty evidence favoured powered brushes: plaque SMD −0.50 (95% CI −0.70 to −0.31; 40 trials, n=2,871) short term and −0.47 (−0.82 to −0.11; 14 trials, n=978) long term. Those translate to roughly 11% and 21% less plaque on the Quigley-Hein (Turesky) index.

Gingivitis SMD was −0.43 (−0.60 to −0.25; 44 trials, n=3,345) short term and −0.21 (−0.31 to −0.12; 16 trials, n=1,645) long term — about 6% and 11% on the Löe-Silness index. Rotation-oscillation brushes had the largest evidence base (27 trials) and significant reductions at both time points. Adverse effects were localised and temporary.

The caveats matter: plaque heterogeneity was high (I² 83% and 86%), only 5 of 56 trials were at low risk of bias, and 46 were unclear. The review states plainly that “the clinical importance of these findings remains unclear.” Better scores are not automatically fewer lost teeth.

The longer, weaker, more interesting signal comes from an 11-year Study of Health in Pomerania cohort of 2,819 adults. Powered-brush users had less progression of mean probing depth (β −0.09, 95% CI −0.16 to −0.02) and clinical attachment loss (β −0.19, −0.32 to −0.07), 17.7% less DMFS progression and 19.5% more teeth retained.^[7] This is observational: powered-brush users were younger, and people who buy an electric toothbrush differ from people who do not in ways no model fully adjusts for.

There is a counterweight. In the network meta-analysis restricted to periodontal maintenance patients, four of five comparisons found no clinical difference between a manual and a powered brush.^[5]

Cochrane's 2019 review included 35 RCTs and 3,929 adults, with a minimum four weeks of follow-up.^[2] Three limitations frame every result: no trial assessed interproximal caries, most did not assess periodontitis, and most participants started with little gum inflammation. There was little room to improve.

Floss plus brushing: low-certainty evidence of reduced gingivitis at one month, SMD −0.58 (95% CI −1.12 to −0.04; 8 trials, 585 participants), with similar findings at three and six months. Bleeding-site and plaque results were inconsistent, at very low certainty.

Interdental brushes: low-certainty evidence of less plaque, SMD −1.07 (−1.51 to −0.63; 2 trials, 93 participants) — the largest effect in the review. Reduced gingivitis came from a single trial of 62 people (MD −0.53, −0.83 to −0.23; very low certainty), with no clear difference in bleeding sites. Head to head, interdental brushes may reduce gingivitis more than floss at one and three months (low certainty), with no difference in probing depth.

Water flossers: very low-certainty evidence of reduced gingival index at one month, SMD −0.48 (−0.89 to −0.06; 4 trials, 380 participants), but not at three or six months. There was no reduction in bleeding sites at one month (MD −0.00, −0.07 to 0.06) or three months; nor in plaque at one, three or six months (one month SMD −0.16, −0.41 to 0.10).

The review concludes that floss or interdental brushes added to toothbrushing may reduce gingivitis or plaque or both, and that interdental brushes may be more effective than floss. Evidence for cleaning sticks and oral irrigators is “limited and inconsistent”; overall, “the effect sizes observed may not be clinically important.” An independent systematic review of oral irrigation, covering 7 studies, likewise found no beneficial effect on visible plaque as an adjunct to brushing, though a positive trend for gingival health.^[4]

Then the author-affiliation check. The most-cited recent water-flosser trial enrolled 105 participants for four weeks and reported a significantly greater fall in whole-mouth bleeding on probing with a water flosser than floss, 0.41 versus 0.19.^[6] It was executed through a Water Pik, Inc. research grant; the sponsor supplied the devices; and the first author was a Water Pik, Inc. employee during the study. It ran at a contract research centre, and the authors declare no conflicts of interest.

None of that makes the result wrong. It does mean the water flosser's best evidence and its manufacturer are not independent of one another — the same pattern we found in the glycocalyx supplement literature and nattokinase literature.

In 2016 an Associated Press investigation reported weak evidence for flossing, and the US dietary guidelines quietly dropped their flossing line.^[17] The distinction is important: Cochrane found weak, short, poorly-blinded trials measuring surrogate outcomes — absence of good evidence, not evidence of absence.

Flossing trials cannot be blinded. They run for weeks rather than decades and have never been powered for cavities or tooth loss; a trial randomising people to never clean between their teeth for ten years would not be approved. The earlier 2011 Cochrane review found reduced gingivitis at one month (SMD −0.36, −0.66 to −0.05), three months (−0.41, −0.68 to −0.14) and six months (−0.72, −1.09 to −0.35), weak and very unreliable plaque evidence, and no trial reporting caries at all.^[3]

The ADA still recommends brushing twice daily and cleaning between the teeth once daily with floss or another interdental cleaner. Its Seal category for powered interdental cleaners requires evidence that the product plus brushing beats brushing alone.^[16] The EFP's S3-level guideline builds the first step of periodontal therapy around supragingival biofilm control, behaviour change and risk-factor control before instrumentation.^[8] A guideline recommendation and the evidence underneath it are two different things. It is worth saying which is which.

Step one — treating gum disease really does change the blood. In the landmark randomised trial, 120 patients with severe periodontitis received intensive periodontal treatment or community-based care.^[11] At 24 hours, flow-mediated dilatation was worse in the intensive group (absolute difference 1.4%, 95% CI 0.5 to 2.3; p=0.002), with higher CRP, IL-6, soluble E-selectin and von Willebrand factor: an acute inflammatory insult from treatment itself.

By day 60 the intensive group was ahead (difference 0.9%, 95% CI 0.1 to 1.7; p=0.02), and by day 180 clearly ahead (2.0%, 95% CI 1.2 to 2.8; p<0.001). Improvement tracked the improvement in periodontal measures (r=0.29, p=0.003).

A meta-analysis of 25 trials in 1,748 patients found treatment reduced hs-CRP by 0.50 mg/L (95% CI −0.78 to −0.22), IL-6 by 0.48 ng/L, TNF-α by 0.75 pg/mL and fibrinogen by 0.47 g/L, with larger effects in patients who already had cardiovascular disease or diabetes.^[12] In a 12-month randomised trial of 264 people with type 2 diabetes and moderate-to-severe periodontitis, intensive treatment lowered HbA1c by 0.6 percentage points (95% CI 0.3–0.9; p<0.0001) versus control treatment.^[13]

Step two — but the causal chain does not close. The American Heart Association concluded that observational studies support an association between periodontal disease and atherosclerotic vascular disease independent of known confounders, but do not support a causative relationship. Despite short-term improvements in systemic inflammation and endothelial dysfunction, “there is no evidence that they prevent ASVD or modify its outcomes.”^[14] The EFP/World Heart Federation consensus reached a similar place: association, plausible mechanisms, no demonstrated effect on hard cardiovascular outcomes.^[15]

Two Mendelian randomisation studies used genetic variants to test causality without the lifestyle confounding that plagues this field, and found nothing. In 44,221 stroke cases and 122,733 coronary artery disease cases, genetically predicted periodontitis was not associated with any stroke (OR 0.99, 95% CI 0.97–1.02), ischaemic stroke (1.00, 0.97–1.03), coronary artery disease (1.01, 0.99–1.03) or carotid intima-media thickness (β −0.002, −0.004 to 0.001).^[18] An independent MR analysis covering coronary artery disease, heart failure, atrial fibrillation and stroke reached the same conclusion, suggesting shared risk factors and comorbidities explain the observed associations.^[19]

Step three — and the cohort that is quoted most is still a cohort. In 247,696 Korean adults followed a median 9.5 years, with 14,893 major cardiovascular events, each additional daily tooth-brushing was associated with 9% lower risk and regular professional cleaning with 14% lower risk after multivariable adjustment.^[20] People who brush three times a day and see a hygienist every year differ systematically from people who do not. No adjustment fixes that.

Clean between your teeth because it is good for your mouth; the evidence that it is good for your arteries runs through professional treatment of established disease, and even there it stops at the biomarker.

The brush is the upgrade with the best evidence: an oscillating-rotating powered brush, two minutes, twice a day.

Then clean between your teeth once daily with whichever device you will genuinely use. Open gaps that take an interdental brush are where the largest measured plaque effect sits. For tight gaps, floss. For braces, implants, bridges, limited dexterity, or if you have simply never flossed, a water flosser is reasonable. Its measured benefit over brushing alone is small and short-term, but it is used — and an unused floss box has an effect size of zero.

Technique and consistency dominate device choice in every review. Persistent bleeding, receding gums, loose teeth or deep pockets on a dental chart need professional assessment and treatment. That is where the documented hs-CRP, endothelial and HbA1c benefits were measured; home devices were not tested for that.

For the toothpaste half of the question — fluoride concentration, and spit, don't rinse — see our fluoride toothpaste and water review.