Employee wellness programs may not actually improve health or save money, according to a new study published in JAMA.

Researchers followed 33,000 employees of a large retailer over 18 months to determine the effects of a company wellness program. These staffers were compared to a control group of 29,000 workers with no wellness program.

One benefit was that individual employees participating in wellness programs reported better health behaviors, like regular exercise and active weight management. Nearly 70% of those in the wellness program group said they routinely exercise, 8.3 points higher than the control group. A similar percentage in the wellness group (69.2%) actively managed their weight, compared to 54% in the control group.

However, there were no significant differences in more measurable factors, like clinical measures of health, healthcare spending and use, and employment outcomes. Many companies implement wellness programs expecting to see returns in both spending and employment.

The researchers measured cholesterol, glucose, blood pressure and BMI in both groups, and found the effects of the wellness program were minimal. They also looked at medical spending, including on doctor visits, hospitalization and medicines, finding little difference between the groups after 18 months. Changes in absenteeism, performance reviews and tenure were also minimal.

The study’s authors noted that their findings contrast with previous studies on wellness programs, which often found reduced employee absenteeism and healthcare spending. They noted that those studies were often based on observational data and may have been prone to selection bias. This study, using a randomized selection process, is likely more accurate, they contended.

Many companies have implemented wellness programs based on studies that found positive returns on health spending and employment, but the report published in JAMA suggested that companies may not see returns on their investment in wellness programs in the short term.

While this report only found changes in health behaviors, the authors suggested that improvements in health spending and clinical health measures could happen eventually if the positive health behaviors, like exercise and weight management, continue.