FREE NEWSLETTER: Econintersect sends a nightly newsletter highlighting news events of the day, and providing a summary of new articles posted on the website. Econintersect will not sell or pass your email address to others per our privacy policy. You can cancel this subscription at any time by selecting the unsubscribing link in the footer of each email.

posted on 01 November 2015

Available Evidence Does Not Support The Conclusion That Certain Policies To Stem Obesity Would Generate Significant Savings For The Government

from the Congressional Budget Office

-- this post authored by Noelia Duchovny, Eamon Molloy, Lori Housman, and Ellen Werble

The share of the U.S. population that is obese has increased substantially since 1980, posing a significant public health challenge. Because obesity is associated with numerous diseases and higher average health care spending, lawmakers have expressed interest in developing policies that would reduce the prevalence of obesity. Determining the likely effects of such policy proposals is difficult, however. Despite a rapidly growing body of literature that explores the effects of obesity on health and health care spending, research on the effects that policy interventions aimed at weight loss would have on the federal budget is largely lacking.

The Congressional Budget Office has determined that the available evidence does not support the conclusion that certain policies to stem obesity - discussed in more detail below - would generate significant savings for the federal government. (A bibliography detailing the studies and related scholarly literature that CBO has consulted for its analysis is available on CBO's website.) Given the limitations of current research, some of which are outlined in this blog post, further well-designed studies and systematic reviews of the literature on the effects of obesity interventions and their budgetary consequences would enhance CBO's analytic capabilities in this area and could change the agency's conclusions.


Research on the increase in obesity rates and the associated consequences for people's health and health care spending is extensive. Obesity rates among U.S. adults (people age 20 or older) have more than doubled since 1980: Over one-third of adults are now considered obese - that is, they have a body mass index (BMI) of 30 or greater. (Body mass index is a measure of body fat based on height and weight; for example, a person who is five feet, six inches tall and weighs 186 pounds has a BMI of 30.) Numerous studies have demonstrated that obese people are more likely to develop serious illnesses, including heart disease, diabetes, and hypertension. In addition, the total costs of their health care are higher, on average, than those of people of normal weight (that is, with a BMI that falls between 18.5 and 25). Those differences mount with the degree of obesity; people with a BMI of 40 or higher have considerably worse health and higher health care spending, on average, than people with a BMI of 30 to 35. Furthermore, about one-third of adults are considered overweight (with a BMI that falls between 25 and 30) and are at higher risk of becoming obese than adults of normal weight.

In view of those high obesity rates and their associated effects, lawmakers are considering new policies to prevent or treat obesity. Such policies could target subgroups of the population (for instance, Medicare or Medicaid beneficiaries) or the population as a whole. Policies designed to prevent or reduce obesity often use broad measures that target the whole population - such as excise taxes on certain foods or nutrition labeling requirements - both to promote weight loss among those who are already overweight or obese and to prevent weight gain. Other policies - such as providing insurance coverage for weight-loss drugs, behavioral therapy, or bariatric surgery - focus on expanding treatment opportunities for people who are overweight or obese, relying on voluntary participation by those who meet the eligibility criteria.

The effects of such policies on overall health care spending and on health outcomes that directly affect the federal budget - such as disability and longevity - are not well understood, however. Studies that evaluate interventions tend to focus only on changes in weight, requiring researchers to draw inferences about their budgetary effects. Furthermore, those studies are largely based on evaluations of motivated and selected populations in controlled settings, a situation unlikely to be duplicated in a less structured environment. Additionally, studies on weight-loss interventions often examine populations that differ significantly from enrollees in Medicare and Medicaid - the two largest federal health care programs - and provide little information about what fraction of eligible beneficiaries might use newly covered obesity treatments.

To illustrate some of the challenges posed by these research gaps, this blog post focuses on recent proposals that aim to promote weight loss among Medicare beneficiaries who are obese, including the following:

  • New or expanded coverage for behavioral counseling, and

  • Coverage of obesity drugs.

Medicare currently covers certain treatments for obesity, including bariatric surgery in some circumstances as well as behavioral counseling by primary care practitioners. Some proposals would permit other types of providers to be paid for counseling and would cover prescription drugs for weight loss under Medicare Part D.

Expanding Medicare's coverage of services to promote weight loss would increase the program's spending initially. But improved health resulting from such interventions might reduce future health care spending, at least in part offsetting those costs to the government. Improved health might also reduce disability rates and increase people's longevity, which could also have consequences for future federal outlays.

Assessing the Budgetary Effects of Obesity Policies: Modeling Steps and Research Gaps

To determine the budgetary effects of expanding coverage under Medicare for obesity treatments, CBO would consider the following questions:

  • How many beneficiaries would participate?

  • How many providers, and of what types, would offer the treatment?

  • What share of participants would complete the full course of treatment?

  • What would be the direct costs of treatment?

  • How much weight would participants lose, and how long would that weight loss be maintained?

  • How would weight loss affect the health care spending of participants and the federal budget?

In the discussion that follows, CBO describes some of the challenges the agency faces in addressing such questions, with the hope of stimulating further research on, and analysis of, these issues by experts in the field. The post discusses the many intermediate steps involved in estimating the effects of potential policies on the federal budget and the types of research that CBO currently draws upon - and would benefit from having more of in the future.

How Many Beneficiaries Would Participate?

Data from nationally representative surveys that provide information on people's height and weight, health conditions, and participation in federal health care programs enable CBO to estimate how many people would be eligible for newly covered services. However, most studies that evaluate weight-loss interventions disclose the number of participants but not the number of people recruited to participate. As a result, CBO lacks direct evidence to determine how many eligible beneficiaries would actually use a new benefit to treat obesity.

Nonetheless, the agency anticipates that participation by eligible Medicare beneficiaries in expanded behavioral therapy programs would be low, as would their use of obesity drugs. The agency reached those conclusions on the basis of the following findings:

  • Relatively few Medicare beneficiaries currently receive intensive behavioral therapy for obesity - a benefit that has been covered since 2011. According to CBO's analysis of claims, approximately 0.5 percent of fee-for-service beneficiaries who are classified as obese used this service in 2013. Additionally, most participants used fewer visits than are recommended for a full course of treatment. Low utilization of that benefit, however, may reflect little demand, limited access to providers offering the therapy, or both. Further insights into the relative importance of demand versus supply in determining participation rates would enhance CBO's estimating capabilities.

  • Most studies have found that obesity drugs are not widely used in general, and many people who begin such treatment discontinue use soon thereafter, possibly because of adverse side effects or perceived ineffectiveness.

How Many Providers, and of What Types, Would Offer the Treatment?

Beneficiaries' participation in weight-loss interventions depends in part on their access to providers who are both authorized to offer treatment and willing to do so. Currently, primary care physicians, nurse practitioners, clinical nurse specialists, and physicians' assistants are the only providers authorized to provide intensive behavioral counseling for obesity under Medicare. The effects of policies to broaden the types of providers who could offer such services would depend on several factors:

  • The additional types of practitioners, such as dietitians, psychologists, and lay weight-loss counselors, authorized to provide the service;

  • Training and certification requirements for providers, particularly for eligible lay counselors;

  • Payment rates for treatments; and

  • Attitudes of providers toward obesity interventions.

Some evidence suggests that primary care physicians may lack the time and training to provide this service effectively. Allowing a range of other providers to be paid for obesity counseling, such as lay weight-loss counselors, for whom training is relatively quick and inexpensive, could therefore greatly expand the number of participating providers. Payment rates for behavioral counseling - currently about $25 for a 15-minute session - might also affect providers' willingness to offer that service. (Nonphysicians are more likely to accept such payment rates.) Understanding the factors determining providers' participation - especially the number and types of providers - as well as the relative effectiveness of the different types of providers would improve CBO's ability to analyze the budgetary effects of weight-loss policies.

With regard to prescription drugs for weight loss, studies suggest that providers are hesitant to prescribe obesity drugs for elderly people (age 65 or older), in particular, because of possible adverse side effects. CBO's conclusion that the use of obesity drugs under Medicare Part D is likely to be low reflects those findings. However, new insights into this issue - including whether drugs recently approved by the Food and Drug Administration or drugs currently in development would be more widely prescribed - would be particularly helpful.

What Share of Participants Would Complete the Full Course of Treatment?

Given the fraction of eligible Medicare beneficiaries who would enroll in a weight-loss program, CBO would estimate the share of initial participants who would complete the full course of treatment and the outcomes for participants who could be expected to drop out before completing treatment. (A full course of behavioral counseling may entail 20 sessions over 12 months; guidelines for prescribing weight-loss drugs typically recommend a 12-week trial and then continuing treatment, if it is considered effective.) In recent reviews of controlled studies of weight-loss interventions, about 70 percent of participants completed behavioral therapies and about 65 percent completed newly approved drug therapies. Evaluations of such initiatives usually do not follow people who drop out and often reflect the assumption that those participants maintain any weight loss achieved before leaving the study. Many dropouts can be expected to regain weight after discontinuing treatment, so assuming that they maintain their weight loss could overstate the effectiveness of the intervention.

Completion rates are likely to be lower when demonstration projects that were undertaken in controlled settings are implemented more broadly; such projects typically enroll motivated participants, use well-trained providers, and employ additional resources to retain participants. In the absence of other evidence, CBO projects higher attrition rates than occur in controlled studies and little or no permanent weight loss for those who do not complete the full course of treatment. Again, however, more knowledge about how different types of providers compare when promoting weight loss among participants might lead to modifications in the agency's projected attrition rates.

What Would Be the Direct Costs of Treatment?

The direct costs to Medicare of policies targeting obesity would depend greatly on the specific features of the legislation and the interventions involved. In general, assuming a full course of treatment, bariatric surgery is the most expensive type of personal weight-loss intervention; intensive behavioral counseling is the least expensive; and prescription weight-loss drugs - which are recommended as an adjunct to counseling, rather than as a substitute for it - fall between the two.

To estimate the per-participant cost to Medicare of a behavioral intervention, CBO would project the average number of sessions attended by participants and the cost of each session, which might depend on the type of provider. To estimate the federal cost of covering weight-loss drugs through Medicare Part D, CBO would estimate drug prices, the average number of refills, and beneficiaries' average cost sharing, as well as the costs of associated counseling. The estimate would also take into account whether there would be ongoing costs after the initial procedure or course of treatment - such as for complications related to bariatric surgery or ongoing use of weight-loss drugs.

How Much Weight Would Participants Lose, and How Long Would That Weight Loss Be Maintained?

Given the limited direct evidence about the effects of weight loss initiatives on health care spending, CBO would conduct its analysis in two steps. The agency would first estimate how much weight Medicare beneficiaries would lose as a result of the new policy and the duration of that weight loss. Then, the agency would use those estimates to calculate the budgetary implications. Weight loss is the usual goal of most obesity initiatives and is typically the main outcome researchers use to evaluate the effectiveness of new treatments. But extrapolating from those studies to estimate the effects on the Medicare population over a 10-year period is difficult for four main reasons:

  • First, because study volunteers are probably highly motivated to lose weight and might have done so without the aid of the study intervention, studies that attribute all weight loss by participants to an intervention may overstate its effectiveness. The best way to address this concern is to use a randomized controlled trial (RCT), in which researchers randomly assign participants to treatment or control groups and then estimate a treatment's effectiveness as the difference in the two groups' weight loss. Because RCTs typically provide more accurate estimates of the weight loss specifically attributable to an intervention than studies that do not use a control group, CBO places a strong emphasis on the findings from RCTs when evaluating the potential effectiveness of weight-loss interventions.

  • Second, because studies of obesity interventions use selected participants and skilled research staff, their results may not directly apply if that same intervention is offered by Medicare. Although RCTs often provide solid evidence of an intervention's effect on the study's participants, important differences may exist between those participants and the population targeted by a legislative proposal. For example, differences in age, gender, race, education, income, or motivation for weight loss may limit the applicability of that study's results to an elderly or disabled population. Similarly, studies often exclude patients with complex health problems, cognitive limitations, or other factors that may limit patients' benefits - but which are all prevalent in the Medicare population. Additionally, replicating the motivation and skill of the staff involved in a study when implementing an intervention on a much broader scale may be difficult. For those reasons, CBO anticipates that weight-loss outcomes resulting from proposed Medicare program initiatives would not match those observed in controlled studies, but the magnitude of the difference is highly uncertain and would depend in part on the policy's details.

  • Third, interventions covered by policy proposals may substantively differ from those used in studies. For example, study interventions often include expensive supplementary treatment, such as supervised exercise training and extended weight-loss maintenance programs, which increase the magnitude and maintenance of weight loss. However, because of their costs, such supplements are often excluded from large-scale translations of study interventions. CBO expects that policies offering simplified versions of study interventions would not achieve identical results.

  • Fourth, few studies on weight-loss interventions track people for more than two years, and evidence from those studies suggests that most participants eventually regain much of the weight they lost. The health benefits of weight loss probably dissipate as weight is regained, but how quickly that occurs is not well known. Given that the Congress has shown increasing interest in examining budgetary effects beyond the first decade of a policy's implementation, well-designed studies addressing the long-term effects of weight-loss interventions would be useful.

How Would Weight Loss Affect the Health Care Spending of Participants and the Federal Budget?

Despite the association between obesity and health care spending, there is surprisingly little evidence that the health care spending of obese people declines when they lose weight. CBO would take those factors into account when assessing proposals to treat obesity.

Association Between BMI and Health Care Spending. Previous studies of the impact of BMI on health care spending have generally focused on comparisons showing that, on average, adults with a BMI of 30 or more have higher overall health care spending than those of normal weight. Most of that difference is attributable to particularly high spending by adults with a BMI of 35 or more. In the case of the elderly population, in particular, CBO's analysis of survey data for 2008 to 2012 indicates that average health care spending for people with a BMI of 30 to 35 was about 10 percent higher than for those of normal weight, about 40 percent higher for those with a BMI of 35 to 40, and about 50 percent higher for those with a BMI of 40 or more. Those differences are based on analysis that holds some other factors that affect spending - such as demographic characteristics and health behavior - constant. Although average spending rises substantially for people with a BMI above 35, a relatively small share of the elderly population is in that BMI category; the share of elderly people with a BMI above 35 is roughly half the share of elderly people with a BMI between 30 and 35 (see the figure below).

Distribution of Body Mass Index and Average per Capita Health Care Spending by BMI Among Elderly Adults

Differences in health care spending across broad BMI categories can be substantial. However, because interventions typically do not move people from one category to another, those differences are unrepresentative of the potential changes in health care spending resulting from weight loss. Specifically, most behavioral or pharmacological interventions aim for weight loss of 5 percent to 10 percent of body weight. (Weight loss of 5 percent, for example, corresponds to a 5 percent decrease in BMI - from 30 to 28.5 or from 40 to 38.) The average difference in health care spending observed between a given BMI and one that is 5 percent to 10 percent lower would be less than the average difference observed between two broad BMI categories.

In its efforts to estimate the effects of policies that would expand coverage of weight-loss interventions on health care spending - and to avoid the limitations of using only the BMI categories in such analyses - CBO has studied the continuous relationship between elderly people's BMI and their overall health care spending. Spending declines as BMI moves from the underweight to normal range, remains fairly constant throughout the normal and overweight range, and grows at an increasing rate as BMI rises across the three categories of obesity (see the figure above). On average, a one-unit increase in BMI - equal to six pounds for a person who is five feet, six inches tall - is associated with an $80 increase in an elderly person's health care spending, but that change varies widely across the BMI distribution, with the largest increase in spending observed at high levels of BMI.

Effects of Weight Loss on Health and Health Care Spending. Evidence about the effects of modest weight loss on the health and health care spending of obese people is inconclusive at best. Two large-scale studies that examined the effects of weight-loss counseling on people's health and on health care spending - Look AHEAD and the Diabetes Prevention Program - resulted in clinically significant weight loss and reductions in risk factors, such as cholesterol and blood glucose levels, which in turn reduced the share of participants with diabetes. Even so, those studies did not find significant reductions in health care spending net of intervention costs or lower rates of other obesity-related health problems, such as heart attacks, diabetes-associated microvascular disease, and strokes, over a 10- to 15-year period.

Additional well-designed studies addressing this point might allow more definitive conclusions to be drawn. However, the limited evidence in the available body of literature supporting such effects may reflect several factors:

  • Unobserved differences in health risks and behavior between obese and nonobese adults, which persist even after obese adults lose weight;

  • Cumulative health effects of obesity that are not fully reversible through weight loss, which may be particularly pertinent to elderly adults who may have been obese for decades; and

  • Study samples that are too small to properly identify minor changes in the risk of obesity-related diseases that could have implications for spending.

CBO would find analyses that addressed these issues to be particularly informative.

CBO's Approach. Given the limited evidence about the ways in which an obesity intervention affects health care spending, CBO would use a multistage process to estimate how weight loss resulting from a policy that targets obesity would affect Medicare spending. That process involves making estimates and judgments about three factors:

  • The initial BMI of those who participate. Given the relationship between BMI and health care spending, weight loss of a specific percentage could yield a greater reduction in spending for a person with a BMI of 42 than for a person with a BMI of 32. Therefore, knowing the underlying BMI distribution of participants - not just the share who fall into the broad BMI categories - is important.

  • How the intervention would change the BMI of participants and how long weight loss would be sustained.

  • How average health care spending for participants with a given initial BMI would change as a result of weight loss - in other words, whether their spending would be similar to that of the average person at the new, lower BMI; closer to the average spending of a person at the initial BMI; or somewhere in between. Such a range largely reflects two uncertainties: whether obesity alone causes the differences in average health care spending between obese and nonobese adults, and the extent to which the adverse health effects associated with obesity are reversible through weight loss. (For several examples of the ranges of potential changes in average health care spending that could result from a 5 percent weight loss corresponding to different initial BMIs, see the figure below.)

Ranges of Potential Changes in Average per Capita Health Care Spending Among Elderly Adults From Weight Loss of 5 Percent, Excluding the Costs of Intervention

The evidence on which to base such judgments - the third factor, in particular - is quite limited, introducing additional uncertainty into CBO's estimates of the effects of relevant policies. The agency will continue to monitor new, related research and will incorporate any pertinent findings into its methodology.

Beyond the direct effects on federal health care spending, federal outlays (including outlays for Social Security, Medicare, and Medicaid) would also be affected if policies addressing obesity eventually led to lower disability rates or greater longevity. Whereas lower disability rates could reduce outlays, greater longevity could increase outlays. But those outcomes are highly uncertain, and further research on the potential long-term effects of policies that expand obesity treatments would be beneficial.

>>>>> Scroll down to view and make comments <<<<<<

Click here for Historical News Post Listing

Make a Comment

Econintersect wants your comments, data and opinion on the articles posted.  As the internet is a "war zone" of trolls, hackers and spammers - Econintersect must balance its defences against ease of commenting.  We have joined with Livefyre to manage our comment streams.

To comment, using Livefyre just click the "Sign In" button at the top-left corner of the comment box below. You can create a commenting account using your favorite social network such as Twitter, Facebook, Google+, LinkedIn or Open ID - or open a Livefyre account using your email address.

You can also comment using Facebook directly using he comment block below.

Econintersect Contributors


Print this page or create a PDF file of this page
Print Friendly and PDF

The growing use of ad blocking software is creating a shortfall in covering our fixed expenses. Please consider a donation to Econintersect to allow continuing output of quality and balanced financial and economic news and analysis.

Take a look at what is going on inside of
Main Home
Analysis Blog
Is Growing Household Debt An Economic Counter-Dynamic?
A New Era of Central Banking?
News Blog
Earnings And Economic Reports: Week Starting 27 February 2017
Time Crystals: How Scientists Created A New State Of Matter
Cost Of War Against ISIS Reaches 11 Billion Dollars
What We Read Today 25 February 2017
Candid Camera Classic: Dirty Clean Words
How Reliable Are GDP Consensus Forecasts?
What You Don't Know About Your Policy Can Hurt You... Financially
The Revolving Door
Historical Echoes: The Legacy Of Freedman's Savings And Trust
Infographic Of The Day: Major Cruise Ship Accidents
Early Headlines: GOP Senators Back Russia Probe, Industry Appeals Fiduciary Rule Verdict, Pot To Produce More Jobs Than MFG, New Arms Race, Aramco IPO May Disappoint Saudis And More
Who Are The More Dependent Parties In The Brexit Scenario
Super Bowl LI Drew 4th Largest TV Crowd In History
Investing Blog
Babson's Ten Commandments Of Trading Weekly Wrap-Up 24 February 2017
Opinion Blog
In Defense Of Greece: An Open Letter To The IMF
As Trump Flounders On Foreign Policy, Russia Flexes Its Nuclear Muscles
Precious Metals Blog
Deflation And Gold: A Contrarian View
Live Markets
24Feb2017 Market Close: Wall Street Rose From Session Lows To Close In The Green Near The Unchanged Line, Short-Term Indicators And Analysts Questioning Continuing Bull Run
Amazon Books & More

.... and keep up with economic news using our dynamic economic newspapers with the largest international coverage on the internet
Asia / Pacific
Middle East / Africa
USA Government



Analysis Blog
News Blog
Investing Blog
Opinion Blog
Precious Metals Blog
Markets Blog
Video of the Day


Asia / Pacific
Middle East / Africa
USA Government

RSS Feeds / Social Media

Combined Econintersect Feed

Free Newsletter

Marketplace - Books & More

Economic Forecast

Content Contribution



  Top Economics Site Contributor TalkMarkets Contributor Finance Blogs Free PageRank Checker Active Search Results Google+

This Web Page by Steven Hansen ---- Copyright 2010 - 2017 Econintersect LLC - all rights reserved