LimitationsThis study is typical of

Limitations
This study is typical of most worksite health promotion studies in
that the design is cross-sectional and retrospective. The ideal
study design (i.e., a randomized, controlled study) is difficult or
impossible to conduct in a worksite population because of the
relationship between employees and employer. Another common
limitation of worksite analyses is that the HRA participants are
not representative of the entire employee population. Because of
the near universal participation rate at this company, the population
studied is very likely to be representative of the corporation
as a whole although we are assuming that those who did not
participate did so at random and that they represent nonimformative
missing data. The results are unique to this corporation,
however, and similar studies should be conducted in a variety of
worksite industries to see if the findings are replicated in different
demographic groups.
As was mentioned, waist circumference was not measured, so
BMI was used as a surrogate. The current criteria of MetS states
that if an individual has a BMI greater than 30 kg/m2
, we can
safely assume that his or her waist circumference exceeds the risk
level [25]. Nevertheless, it does leave the possibility that individuals
with BMI < 30, but with central adiposity which meets
the waist circumference risk cut point would not be identified
here. A future study will compare the use of waist circumference
and BMI as risk factors for MetS because the company added
waist circumference to its screening in 2007.
The information on medical conditions in this study relied on
self-reporting of participants. Each individual’s criteria for
reporting a certain condition may not have matched typical
diagnostic criteria for each condition. Previous studies of selfreport
data have shown that relying on self-report for medical
conditions can be a valid method [47–49], although in one study
patients reported more conditions than could be verified in
medical charts [50].
Conclusions
This study provides employers, health-care providers, and public
health professionals with more information about the extent of
MetS and its consequences in working populations. The medical
conditions arthritis, chronic pain, diabetes, heartburn, and heart
disease were significantly more prevalent in employees with MetS
than those without MetS. Indeed, individuals with MetS but no
disease in 2004 were more likely to newly report four of those
five conditions (arthritis, chronic pain, diabetes, and heart
disease) 2 years later. It was unknown whether the diseases
associated with MetS in the general population would also be
found in a working population because of the HWE [51] but it
does appear to be the case.
Moreover, this study highlights the opportunity that is available
to organizations seeking to improve the health of employees.
While employees with MetS and a medical condition hadsignificantly higher costs than other employees, the vast majority of
employees with MetS in this study had not yet developed one of
the five medical conditions studied here, and their costs were not
significantly greater than those without MetS. If individuals take
advantage of programs helping them to both maintain their low
risks and reduce their high risks, their odds of experiencing
disease will be reduced. This leads to improved vitality and
quality of life for individuals, and cost avoidance for corporations
in the form of lower health care, pharmacy, and STD costs.
Acknowledgments
The authors wish to thank Kathy Babiak, Ph.D., Alfred Franzblau,
M.D. and David Moore, Ph.D. for their valuable contributions
to the design and execution of this study and manuscript.
Source of financial support: No funding was received for this research.