Employee Wellness Program Statistics: Adoption, Evidence and ROI

If you are asking leadership to fund a wellness program, someone will ask what it returns. Search for an answer and you will find confident figures: a few dollars back for every dollar spent, large drops in sick days, big gains in productivity. Many of them are hard to trace, and some describe something other than a workplace program.

This guide sets out the figures that hold up. It covers how many employers offer workplace (corporate) wellness programs and which kinds, what randomized trials found when employers tested programs properly, where the popular return-on-investment figures come from, and the federal rules on incentives. It ends with a way to build a case your finance team can check.

DisclosureJake Jorgovan co-founded Alpha Apex Group, which is mentioned in this article.

How many employers offer wellness programs?

KFF's annual Employer Health Benefits Survey tracks these programs nationally. Its 2025 edition (opens in a new tab), published in October 2025, is based on interviews with 1,862 public and private employers. Its wellness questions go only to firms that offer health benefits, so the percentages below describe those employers.

The largest employers go further. Among firms with 5,000 or more workers, 75% offer a health risk assessment, 84% a weight loss program and 88% lifestyle or behavioral coaching. Incentives are common there too: 61% of these firms that offer a wellness program give an incentive for taking part in or completing it, against 45% of other firms with 200 or more workers.

Image

Zoom

KFF’s 2025 comparison covers firms offering health benefits. It separates smoking cessation, weight management, and lifestyle coaching by employer size. Offering a program does not establish participation, health improvement, or financial return.kff.org (opens in a new tab)

Incentives also attach to screening. Among firms with 200 or more workers, 53% of those offering a health risk assessment and 62% of those offering biometric screening use incentives or penalties to get workers to complete it. Across surveyed firms of all sizes that offer biometric screening, 21% tie a reward or penalty to meeting a biometric result, such as a cholesterol level or body weight; this figure excludes tobacco incentives.

Offering a program is not the same as people using it. A RAND study for the U.S. Departments of Labor and Health and Human Services (opens in a new tab), published in 2013, found that about half of U.S. employers offered wellness initiatives, and that fewer than half of employees at the firms it surveyed (46%) completed a screening or health risk assessment. RAND also found that incentives for completing an assessment were effective, particularly above $50.

What do randomized trials of wellness programs find?

There is little experimental evidence on wellness programs, as the authors of one of the trials below put it. A study that compares employees who join with those who don't has a built-in problem: the people who sign up may already be healthier. Randomized trials avoid it by deciding at random who is offered the program, then comparing the two groups. Two large U.S. trials have tested multi-part programs of the kind employers offer, with four published results between them.

Image

Zoom

The Illinois trial found no significant first-year difference in medical spending between treatment and control groups. The study also found that employees who chose to participate differed before the intervention, which helps explain why simple participant-versus-nonparticipant comparisons can mislead.nber.org (opens in a new tab)

The Illinois team also showed why earlier studies looked better. Before the program started, the employees who later chose to take part already had lower medical spending and healthier habits than those who didn't. In their summary of results (opens in a new tab), the researchers put it plainly: participants "already had lower health care costs before the program began." By their account, their trial's 95% confidence intervals rule out 83% of previous estimates of the effect on medical spending and absenteeism.

Two limits apply. Each trial tested one employer's program, with follow-up of up to three years, so they say little about a program that runs for a decade or targets a specific condition. And the Illinois authors note that a program could still save an employer money in another way: by attracting or keeping lower-cost employees. That saving would come from selection alone, with no change in anyone's health.

Where the big ROI figures come from

One source of the "dollars back for every dollar" figures is a 2010 meta-analysis in Health Affairs (opens in a new tab) by Katherine Baicker, David Cutler and Zirui Song. It found that medical costs fell by about $3.27 and absenteeism costs by about $2.73 for every dollar spent on wellness programs. The authors called for more work on the mechanisms and on how widely the findings apply. Nine years later, Song and Baicker's own randomized trial found no significant difference in spending or absenteeism after 18 months, and none after three years.

Two company evaluations supply other figures you may see:

• Johnson & Johnson. An evaluation in Health Affairs (opens in a new tab) compared the company with similar large employers from 2002 to 2008. Its medical spending grew 3.7 percentage points a year more slowly, with savings of $565 per employee a year (in 2009 dollars) and a return of $1.88 to $3.92 for every dollar spent. The program dates from 1979. Harvard Business Review reported in 2010 (opens in a new tab) that J&J's leaders estimated the programs had saved $250 million on health care costs over the previous decade.
• PepsiCo. An evaluation in Health Affairs (opens in a new tab) of seven years of PepsiCo's program found that the disease management part, for employees already living with a chronic condition, was associated with $136 lower health care costs per member per month, driven by 29% fewer hospital admissions. The lifestyle management part, aimed at reducing health risks, was not associated with lower costs.

These studies compare participants with non-participants or with other companies, so the selection problem the Illinois trial measured can affect them. The PepsiCo result is the most useful of them for a business case, because it separates the parts of a program. Support for people who already have a chronic condition had a measurable association with lower costs; general lifestyle programs did not.

What targeted programs show

The evidence here answers two different questions: whether an intervention prevents disease, and whether a program component is associated with lower employer costs.

Diabetes prevention. The 58% figure that some wellness statistics pages present as a program result comes from the Diabetes Prevention Program trial (opens in a new tab), published in 2002. It randomly assigned 3,234 adults with raised blood sugar to a lifestyle program (a goal of 7% weight loss and 150 minutes of activity a week), the drug metformin or a placebo. Over an average of 2.8 years, diabetes incidence was 58% lower in the lifestyle group and 31% lower in the metformin group than in the placebo group. This was a clinical trial, not a workplace program. The CDC runs a National Diabetes Prevention Program (opens in a new tab) built on that research and has a section for employers and insurers. It says people with prediabetes in a structured lifestyle change program could cut their risk of type 2 diabetes by 58%, and by 71% for people over 60.

Disease management. In the PepsiCo evaluation above, this was the only component associated with lower costs. That is evidence of employer savings from one employer's observational data; it does not test health outcomes.

What does the UK survey show about individual well-being programs?

Mindfulness apps, resilience training and stress management classes are individual-level programs: they aim to change how the worker copes with the job. A large UK survey found no association between taking part and better reported well-being. In a study of 46,336 workers in 233 UK organizations (opens in a new tab), published in the Industrial Relations Journal in January 2024, William Fleming of the University of Oxford compared workers who took part in individual-level well-being programs with those who did not. Across several measures of well-being, participants appeared no better off. The Oxford Wellbeing Research Centre describes it as thought to be the largest study of its kind.

The study is a survey at one point in time and its sample is British, so it shows an association. The researchers recommend putting more effort into working conditions; the survey itself does not measure the effect of those changes. The Oxford Wellbeing Research Centre's summary (opens in a new tab) says changes to the organization, such as scheduling flexibility, management practices, staff resources, performance review and job design, are likely to matter more. Fleming put it this way: "There's growing consensus that organisations have to change the workplace and not just the worker."

For an HR leader, that is a practical point. If employees name workload, scheduling or management as sources of stress, an app is unlikely to fix it. Address those conditions directly, and say in the business case what will change in the work itself.

What are the rules on wellness incentives?

If a wellness program's reward runs through a group health plan, such as a lower premium or deductible, the HIPAA nondiscrimination rules apply. Under those rules (29 CFR 2590.702 (opens in a new tab)):

• Participatory programs reward taking part without any health standard, such as completing a health risk assessment or joining a smoking cessation program whether or not the employee quits. They must be available to all similarly situated individuals, regardless of health status.
• Health-contingent programs require an activity (such as a walking program) or an outcome (such as a biometric result). Their combined reward can be up to 30% of the total cost of employee-only coverage, or of the coverage dependents are enrolled in if they can take part, and up to 50% where the extra amount is for a program to prevent or reduce tobacco use.
• Alternative standards protect access to the full reward. An activity-only program must offer a reasonable alternative or a waiver to anyone for whom a medical condition makes the activity unreasonably difficult or medically inadvisable. An outcome-based program must offer one to anyone who does not meet the initial target.
• Other conditions apply to every health-contingent program: a chance to qualify at least once a year, a reasonable design to promote health or prevent disease, and plan materials that say an alternative standard is available.

For a sense of scale, take KFF's average annual premium for single coverage in 2025, $9,325: 30% of it is about $2,800 a year and 50% about $4,660. Your own limit depends on your plan's cost. The regulation itself says that meeting its terms does not decide whether a practice is allowed under the Americans with Disabilities Act or other laws, so have benefits counsel review the design before launch.

How to check a wellness statistic before you use it

Before a figure goes into your proposal, run it through four questions:

A figure that fails the first question should not be in a business case. One that fails the second may still be useful, but present it as an association, and expect your CFO to ask about selection.

Building a wellness business case your CFO can check

The evidence does not say wellness programs are worthless. In the two randomized trials above, the programs did not produce statistically significant reductions in medical spending over follow-up periods of up to three years. The targeted-program evidence concerns different interventions and populations. A stronger case is narrower and is measured from the start.

1. 1

   Decide what the program is for

   Write down the outcome you want in the first year and how you will measure it: completed screenings, enrollment in a diabetes prevention program, smoking quit rates, or absence in one team. A program meant to improve retention is a different case from one meant to cut claims, and it needs turnover figures as its baseline.

2. 2

   Put the money where the evidence is

   Consider structured diabetes prevention for eligible employees and disease management support for employees with chronic conditions, and keep evidence of better health apart from evidence of employer savings. If employees name workload, scheduling or management as sources of stress, include changes to those conditions and measure their results. Treat general lifestyle offerings as a benefit employees may value; don't count on them to cut costs.

3. 3

   Run a pilot with a comparison group

   Where you can, randomly choose which eligible sites, teams or employees get the program first and which get it later. Compare outcomes by assigned group, including the people offered it who do not take part. A comparison of participants with non-participants can be skewed by differences that existed before the program, as the Illinois trial showed.

4. 4

   Set incentives inside the rules

   Decide whether the program is participatory or health-contingent, keep any health-contingent reward within the federal limits, and have counsel review it.

5. 5

   Make it easy to use and keep asking

   RAND's study lists what helped programs succeed: broad outreach and clear messaging from leaders, wellness as a priority for senior leaders, activities that are convenient and accessible for all employees, and employee feedback used to keep improving the program. A committee drawn from several departments is one way to gather that feedback.

6. 6

   Report what you find, including what did not work

   Share participation by group, outcomes against the comparison group, and employee feedback on access and usefulness. Report the full cost, including provider fees, incentives, administration and paid employee time. Use it to decide what to scale and what to stop.

When the evidence on a decision is unclear, I'd go and get the facts before committing much money. In a 2020 post, I wrote that "when the path isn't clear, it's because you lack the facts to make the right decision," and that action leads to facts. A pilot with a comparison group is how an HR team gets its own facts, instead of borrowing a vendor's.

If your leadership team asked for one number tomorrow, which of your current figures would survive the four questions above?