A Protocol to Deliver Intensive Behavioral Therapy (IBT) in Primary Care Settings: The MODEL-IBT Program

The Centers for Medicare and Medicaid Services (CMS) initiated coverage of intensive behavioral therapy (IBT) for obesity in 2011, providing beneficiaries 14–15 brief, individual counseling visits in 6 months. CMS offered general recommendations for delivering IBT but did not provide an evidence-based treatment protocol — the objective of the present research.

Methods:

This narrative review describes the evidence that CMS considered in developing its IBT benefit. It also examines weight losses produced by the intensive lifestyle intervention in the Diabetes Prevention Program (DPP), as well an adapted version of the DPP delivered (for the first 6 months) on the visit schedule recommended by CMS. This new protocol, which was evaluated in a recent randomized trial, provided 14 visits in the first 24 weeks, with 7 additional monthly visits through week 52.

Results:

As reported previously, the 50 participants with obesity assigned to the new IBT protocol lost a mean of 5.4% of initial weight at week 24; 46% of participants lost ≥5% of baseline weight. At 1 year, the mean loss was 6.1%, and 44% of participants lost ≥5%.

Conclusions:

With these generally favorable results, we are posting our IBT protocol on-line for practitioners and researchers to use.

Keywords: obesity, behavior therapy, lifestyle modifications, primary care

INTRODUCTION

In 2011, the Centers for Medicare and Medicaid Services (CMS) initiated coverage of intensive behavioral therapy (IBT) for obesity for beneficiaries treated in primary care settings. 1 CMS’ IBT benefit provides weekly brief (15 minute) face-to-face visits in the first month, followed by every-other-week visits in months 2–6. This represents approximately 14–15 visits during the first 6 months. Patients who lose ≥3 kg (6.6 lb) are eligible for monthly IBT visits from months 7 to 12 (to facilitate the maintenance of weight loss). Thus, the IBT benefit provides a total of 21–22 visits in 1 year for successful participants.

CMS’ IBT benefit represents a major advance in recognizing the perils of obesity and the health benefits of moderate weight loss. 2 We hope that CMS’ historic decision in covering IBT for obesity will encourage other insurers and employers to do so.

CMS’ IBT FOR OBESITY

Frequency of IBT Visits

The frequency of visits recommended by CMS is also consistent with recommendations of the 2013 Guidelines for Managing Overweight and Obesity in Adults. 2 These Guidelines recommend that patients with overweight/obesity enroll in a comprehensive, high-intensity intervention that provides ≥14 visits in 6 months. Comprehensive interventions include three central components (i.e., diet, physical activity, and behavior therapy), which are delivered by a trained interventionist who provides patients regular feedback on their progress. The Guidelines indicate that behavioral interventions also may be provided by commercial weight loss programs that have published peer-reviewed evidence of their safety and efficacy (e.g., Jenny Craig, Weight Watchers), as well as by digitally-delivered programs (e.g., Internet, smartphone, telephone) that include the provision of personalized feedback from a trained interventionist. 2

Content of IBT Visits

CMS recommends that IBT for obesity, delivered by primary care practitioners (PCPs), be consistent with a 5A approach to counseling, consisting of assess, advise, agree, assist, and arrange. 5 This approach can be used for educating patients about a variety of health behaviors (e.g., smoking, alcohol intake, etc), 5 and it provides a respectful, patient-oriented perspective for discussing the desirability of weight management. 6,7 However, the 5A approach itself does not provide a curriculum of behavior change for adopting new eating, physical activity, cognitive, and other habits, as typically required to achieve clinically meaningful weight loss. 2,8 CMS did not recommend a specific intensive behavioral weight loss program for PCPs to use, and we fear that many providers will have difficulty identifying (and delivering) an appropriate one.

The DPP provides a structured and easy-to-follow curriculum of behavior change that was developed for the original study 4 by a team led Rena Wing, Ph.D. and Bonnie Gillis, M.S., R.D., and now has been adapted for delivery in YMCAs 9 and other community settings. 10 The efficacy of the DPP program has been demonstrated in multiple studies, with mean 6-month weight losses typically of 5–7 kg (5–7% of initial weight) when delivered in academic medical centers and of 4–5 kg (4–5%) in community settings. 9–11

We believe that CMS’ IBT protocol will produce the largest weight losses when combined with the DPP intervention or a similarly structured manual, such as the LEARN Program 12 or the Look AHEAD intensive lifestyle intervention. 13 (A randomized trial ultimately is necessary to test this hypothesis, but we are confident of the outcome.) Most behavioral weight control interventions, including the DPP, are typically delivered using counseling principles and techniques that are consistent with the 5A approach.

Interventionists to Deliver IBT Visits

In its coverage decision, CMS stipulated that IBT be delivered in primary care settings by physicians, nurse practitioners (NPs), nurse specialists, and physician assistants (PAs). RDs and other auxiliary personnel potentially may provide IBT if they do so “incident to” the identified PCPs, who must be physically present at the time services are provided by an auxiliary provider). 14 (Referral to RDs and other auxiliary providers who work outside of the primary care setting is not covered by CMS.) At the time CMS issued its IBT coverage decision, we do not believe that any randomized trials had assessed weight losses produced by PCPs, in primary care settings, who followed CMS’ schedule of brief (15 minute) IBT visits. The meta-analysis conducted by LeBlanc et al 3 examined trials that were considered to be relevant to primary care but were not necessarily conducted in primary care settings (e.g., the DPP). These trials did not employ as interventionists the practitioners mandated by CMS.

CMS’ stipulation that IBT be provided by physicians, NPs, and other primary care providers was surprising, given the limited training that these professionals receive in delivering behavioral weight loss counseling, as well as their lack of time to provide such care in their already busy schedules. 15 We believe that RDs, behavioral psychologists, health coaches, and other auxiliary professionals likely would have more time and expertise for delivering IBT, as independent providers, than do PCPs (and they potentially could do so at a lower cost). (These suggestions were raised with CMS during the public comment period that preceded the coverage decision’s announcement. CMS declined to include auxiliary personnel as independent providers, preferring to keep the delivery of IBT in the primary care setting). CMS’ requirement of having physicians and other PCPs provide IBT likely has contributed to the very low (<1%) utilization of this benefit. 16

THE DIABETES PREVENTION PROGRAM

Delivering the DPP Protocol in Primary Care Settings

Prior to the release of CMS’ IBT coverage decision in 2011, our research team conducted two studies in which IBT was delivered to patients in primary care settings by medical assistants (i.e., coaches) who worked in collaboration with primary care physicians. 17–18 The coaches provided 8 brief (15 minute) counseling sessions over 6 months, working from an abbreviated/adapted version of the DPP intervention. 4 Counseling sessions in both trials were provided at weeks 0, 2, 4, 8, 12, 16, 20, and 24. Coaches conducted the sessions following a detailed protocol, and at each session participants received handouts and skill building assignments (to be completed daily before the next visit) that focused on consuming a 1200–1800 kcal/d diet and increasing their physical activity to ≥150 minutes per week. At month 6, participants in the study by Tsai et al 17 lost a mean of 4.1% of initial weight and those in the Wadden et al 18 trial lost 3.5%. Patients in the latter trial, referred to as the POWER-UP study, continued to receive monthly coaching visits through month 24, at which time their mean loss was 2.9%. 18

Results of these two trials revealed that medical assistants, with training, could provide behavioral weight loss counseling. The two studies, however, provided only moderate intensity counseling (i.e., 8 visits in 6 months), which was less frequent than the treatment schedule later recommended by CMS and by the Guidelines for Managing Overweight and Obesity in Adults (i.e., ≥14 visits in 6 months). Subsequent studies that delivered an adapted version of the DPP protocol in primary care settings showed varying weight losses, which were related to the frequency of IBT visits provided. 8,19

THE MODEL PROGRAM

Combining CMS’ IBT Benefit with the DPP Protocol

Wadden et al 20 recently completed a randomized controlled trial that provided 14 CMS-style IBT visits in the first 24 weeks, followed by an additional 7 monthly visits from weeks 24 to 52 (i.e., total of 21 visits). (All participants were provided monthly IBT visits from weeks 24–52, regardless of whether they lost ≥3 kg at week 24, as required by CMS.) The 15-minute visits were delivered following a further revision of the adapted DPP protocol used in the POWER-UP study. 18

A full description of this study’s methods, interventions, and results was published in Obesity in January 2019. 20 Our research team referred to this study (internally) as the MODEL trial, the acronym standing for Managing Obesity with Diet, Exercise, and Liraglutide. We refer to the treatment protocol as the MODEL-IBT Program.

Summary of the study’s intervention and design.

Consistent with the CMS benefit, IBT visits in our recent trial were delivered by a physician, two nurse practitioners (NPs), and two RDs, the latter who worked incident to the PCPs. All interventionists received an initial 4–6 hour training (available from the authors) which began by providing guidance on how to talk to participants about obesity in a sensitive, respectful, and collaborative manner. It also included an overview of the program’s dietary and physical activity recommendations, as well as an introduction to core behavioral strategies including goal setting, self-monitoring, stimulus control, problem solving, and cognitive restructuring. Interventionists successfully completed two role plays of delivering IBT sessions before they met with their first participants. 20 Thereafter, they received approximately monthly individual supervision of 30–60 minutes.

A total of 150 participants with obesity were randomly assigned to one of three treatment groups: 1) IBT-alone (as described above); 2) IBT combined with liraglutide 3.0 mg/d (IBT-liraglutide); and 3) IBT-liraglutide, combined for 12 weeks, with a portion-controlled diet (Multi-component group). The physician and two NPs treated a total of 90 patients and the two RDs the other 60 participants. Patients were treated by the same interventionist at all 21 visits, except as occasionally prevented by a provider’s illness or vacation. All participants were recruited and treated at a single site in an academic medical center.

Summary of study results.

At week 24, participants in the IBT-alone, IBT-liraglutide, and Multicomponent groups lost a mean of 5.4%, 10.1%, and 12.2% of initial weight, respectively. 20 A total of 56%, 86%, and 90% of participants, respectively, lost ≥3 kg, qualifying them for 6 additional monthly IBT visits. (All participants received these additional visits, regardless of weight loss, as noted previously.) Fully, 46%, 78%, and 82% of participants in the three groups, respectively, lost 5% or more of their initial weight at week 24.

A total of 91% of participants across the three treatment groups completed the 52-week study. At this time, participants in the three groups lost a mean of 6.1%, 11.5%, and 12.8% of initial weight, respectively. A total of 44%, 70%, and 74% of participants, respectively, lost 5% or more of their initial weight.

Additional analyses revealed that participants who received IBT from the physician and two NPs lost comparable amounts of weight at week 52 as those who were treated by the two RDs (mean losses of 9.8% and 9.9% respectively). Mean weight losses also did not differ significantly as a function of the interventionists’ prior experience in delivering IBT. Mean 52-week loss for participants treated by our two experienced interventionists was 10.2%, compared with 9.7% for those treated by our three novice providers.

Validation of the MODEL-IBT Program

We are encouraged by the weight losses obtained with IBT-alone in the present study. However, the study clearly needs to be replicated to determine whether the results are generalizable beyond our single site, with its skilled interventionists and selected research participants.

Shortly after we began treatment of the first participants in our trial (September 2016), Novo Nordisk began enrolling participants (in February 2017) in a randomized controlled trial in which 282 patients were recruited at 17 largely primary care sites across the US. Participants were randomly assigned to liraglutide 3.0 mg/d or placebo, both combined with 23 brief sessions of IBT over 56 weeks. The Novo Nordisk trial implemented a modified version of the DPP protocol that we used in our POWER-UP study, 18 the biggest change being the increased number of counseling visits provided during 56 weeks (i.e., 23). Our research team assisted Novo Nordisk in developing their treatment protocol, in providing study interventionists (all RDs) from the 17 sites an initial 4–6 hour training in IBT, and then offering monthly 1-hour group supervision calls for the duration of the study.

Results of the Novo Nordisk-sponsored have not been published at this time but were presented at the 2019 annual meeting of the European Congress on Obesity (ECO). 21 At week 28, participants treated by IBT alone (with placebo) lost a mean of 5.4% of initial weight, as compared with a significantly greater 8.4% for participants who received IBT plus liraglutide. Fully 58% of IBT-placebo and 82% of IBT-liraglutide participants lost ≥3 kg at this time. (The full 56-week results of the study will be published shortly.)

Weight losses in the Novo Nordisk-sponsored trial compare favorably with those from the MODEL study, 20 the latter which could be expected to produce greater losses with IBT-alone, given the investigative team’s experience with this approach. The results achieved by the RDs in the Novo Nordisk trial are encouraging, given that some dietitians had little experience with IBT, and many of their site investigators had minimal exposure to behavioral methods of weight control.

Dissemination of the MODEL-IBT Protocol

Based upon the generally favorable results of these two recent trials of IBT, delivered on the visit schedule recommended by CMS (for the first 6 months), we have decided to make our protocol available to other practitioners and investigators by posting it online at Obesity (see on-line supplemental material). We hope that the protocol’s availability will encourage further study of its benefits for both health care providers and their patients.

As seen in the supplemental, on-line materials, the MODEL-IBT Program provides 21 visits (over 1 year), which are delivered weekly from weeks 1–4 and then every other week from weeks 6–24 (at weeks 6, 8, 10, 12, 14, 16, 18, 20, 22, and 24). Visits then are provided monthly for the remainder of the year (at weeks 28, 32, 36, 40, 44, 48, and 52). For each visit, the treatment protocol provides a guide for the interventionist to conduct the session, as well as a handout for the participant. The handouts also contains the weekly skill builder assignments that participants are to complete between sessions. Perhaps the most important assignment is participants’ daily recording of their food and calorie intake, to be reviewed with the interventionist at the next visit. Recording food intake is greatly facilitated by the use of apps, such as “myfitnesspal.com” or “Lose It!” Interventionists should familiarize themselves with these and other approaches. Other key behaviors on which participants should focus regularly include monitoring physical activity and body weight. 22

The treatment guide for interventionists provides structure for conducting the session and suggested language for introducing some of the topics. Both of these may be adjusted to fit the interventionist’s style and preferences. We do, however, believe that interventionists will help participants achieve their most successful weight losses by adhering to the intervention protocol as closely as possible, as concerns monitoring food intake, physical activity, and body weight.

We developed three different treatment protocols for our recent trial: 20 1) the one that is posted on-line for IBT-alone, as well as protocols that included, 2) the addition of liraglutide 3.0 mg/d to IBT, and a 3) version that included the addition of both liraglutide 3.0 mg/d and a 12-week portion-controlled diet of 1000–1200 kcal/d (from weeks 4 to 16 of the study). Researchers interested in using the second or third protocols should contact the first author.

Consistent with the guidelines that govern the use of the original DPP materials, from which our protocol was adapted, the MODEL-IBT Program is not permitted for use by for-profit, commercial entities that wish to sell the program (or interventions directly modeled on it) to the public. It may be used by researchers and practitioners who are providing persons with overweight or obesity individual or group treatment. Practitioners are welcome to share all participant handouts with their patients.

ADVANCING CMS’ IBT BENEFIT FOR OBESITY

Consistent with CMS’ coverage for IBT, the MODEL-IBT Program was developed to be delivered by PCPs (or auxiliary health professionals working incident to these practitioners, e.g., RDs) in brief (15-minute) individual visits in primary care settings. We believe, however, that the MODEL-IBT program can be delivered by a range of health care providers (including RDs and other behavioral specialists) who work in a variety of clinical settings (including outside of primary care parctice). These practitioners may want to extend individual treatment visits to 30 minutes or more, depending on patients’ needs, as well as the reimbursement available from insurers other than CMS (which will not cover IBT outside of primary care settings). We believe that the MODEL-IBT program also can be readily adapted to provide group treatment, as eventually used in the DPP 4 and in Look AHEAD. 5 Group sessions typically last 60–90 minutes and include approximately 8–20 participants. 22

The Guidelines for Managing Overweight and Obesity 2 concluded that IBT (also known as comprehensive lifestyle modification) can be effectively delivered using either individual or group sessions and also can be provided by phone and by some digital approaches (i.e., Internet, smartphone), although digital delivery may produce somewhat smaller weight losses than traditional on-site counseling. With additional research demonstrating the efficacy of both some commercial 2 and some digitally-delivered programs, 23 we encourage CMS to consider expanding both the range of treatment providers and programs that are eligible to provide IBT – and to allow counseling to be provided outside of primary care settings. Millions of Americans with obesity could benefit from IBT. PCPs will welcome having effective, low-cost interventions to which to refer their patients with obesity (as recommended by the U.S. Preventive Services Task Force 24 ) when they are unable to provide IBT themselves.

Supplementary Material

1

Acknowledgements

The authors thank members of the Center for Weight and Eating Disorders at the University of Pennsylvania who collaborated on our recent study, Intensive behavioral therapy for obesity combined with liraglutide 3.0 mg: A randomized controlled trial. They include the publication’s other authors (Olivia A. Walsh, Robert I. Berkowitz, Ariana M. Chao, Kathryn Gruber, Sharon Leonard, Kimberly Mugler, and Zayna Bakizada), as well as research coordinators Breanna D’Antonio, Callie Fisher, and Emilie Pinkasavage. We thank Meghan Butryn, who provided a first draft of the treatment protocol for our POWER-UP study, working from materials developed by Adam Tsai, as well as the original DPP protocol, developed by Rena Wing, Ph.D. and Bonnie Gillis, M.S., R.D. We thank other members of the POWER-UP team, including David Sarwer and Sheri Volger, who contributed to the development of the treatment manual. We acknowledge Olivia Walsh for her excellent work in further revising materials from the POWER-UP study, which now comprise the treatment protocol for the MODEL-IBT Program.

The current MODEL-IBT protocol was developed without support from the National Institutes of Health or other federal or private agencies. However, the protocol is adapted from the DPP and the POWER-UP trials, both of which were supported by the NIH (i.e., Knowler et al, N Engl J Med 2002;346:393–403; Wadden et al, N Engl J Med 2011;365:1969–1979). 3,4

Funding: Completion of the clinical trial, which was implemented following the IBT protocol described in this report, was supported by an investigator-sponsored study award from Novo Nordisk to the University of Pennsylvania on behalf of Thomas A. Wadden. Development of the IBT protocol was supported by the authors. The protocol is adapted from the Diabetes Prevention Program (Knowler WC et al, N Engl J Med 2002;346:393–403) and the POWER-UP study (Wadden TA et al, N Engl J Med 2011;365:1969–79).

Footnotes

Disclosures: Thomas Wadden reports serving on scientific advisory boards for Novo Nordisk and WW (formerly Weight Watchers). Adam G. Tsai reports no disclosures. Jena Shaw Tronieri serves as a consultant to Novo Nordisk, advising the company on the implementation of intensive behavioral therapy.

References

1. Centers for Medicare and Medicaid Services. Decision memo for intensive behavioral therapy for obesity (CAG-00423N) . Published 2015. (https://www.cms.gov/medicare-coverage-database/details/nca-decision-memo.aspx?&NcaName=Intensive%20Behavioral%20Therapy%20for%20Obesity&bc=ACAAAAAAIAAA&NCAId=253. Published November 29, 2011.

2. Jensen MD, Ryan DH, Donato SM, et al. Executive summary: guidelines (2013) for the management of overweight and obesity in adults: a report of the American College of Cardiology/American Heart Association Task Force on practice guidelines and the Obesity Society published by the Obesity Society and American College of Cardiology/American Heart Association Task Force on practice guidelines. Based on a systematic review from The Obesity Expert Panel, 2013 . Obesity (Silver Spring) 2014; 22 ( suppl 2 ):S5–S39. [PubMed] [Google Scholar]

3. Leblanc ES, O’Connor E, Whitlock EP, Patnode CD, Kapka T. Effectiveness of primary care-relevant treatments for obesity in adults: a systematic evidence review for the U.S. Preventive Services Task Force . Ann Intern Med 2011; 155 :434–447. [PubMed] [Google Scholar]

4. Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin . N Engl J Med 2002; 346 :393–403. [PMC free article] [PubMed] [Google Scholar]

5. Glasgow RE, Emont S, Miller DC. Assessing delivery of the five ‘As’ for patient-centered counseling . Health Promot Int 2006; 21 :245–255. [PubMed] [Google Scholar]

6. Vallis M, Piccinini-Vallis H, Sharma AM, Freedhoff Y. Modified 5 As: minimal intervention for obesity counseling in primary care . Can Fam Physician 2013; 59 :27–31. [PMC free article] [PubMed] [Google Scholar]

7. Sherson EA, Yakes Jimenez E, Katalanos N. A review of the use of the 5 A’s model for weight loss counselling: differences between physician practice and patient demand . Fam Pract 2014; 31 :389–398. [PubMed] [Google Scholar]

8. Wadden TA, Butryn ML, Hong PS, Tsai AG. Behavioral treatment of obesity in patients encountered in primary care settings: a systematic review . JAMA 2014; 312 :1779–1791. [PMC free article] [PubMed] [Google Scholar]

9. Ackermann RT, Finch EA, Brizendine E, Zhou H, Marrero DG. Translating the diabetes prevention program into the community. The DEPLOY pilot study . Am J Prev Med 2008; 35 :357–363. [PMC free article] [PubMed] [Google Scholar]

10. West DS, Krukowski RA, Larsen CA. Treatment of obesity in community settings In: Wadden TA & Bray GA, eds. Handbook of obesity treatment (second edition). New York, NY: The Guilford Press, 2018:492–507. [Google Scholar]

11. Ali MK, Echouffo-Tcheugui J, Williamson DF. How effective were lifestyle interventions in real-world settings that were modeled on the diabetes prevention program? Health Aff (Millwood) 2012; 31 :67–75. [PubMed] [Google Scholar]

12. Brownell KD. The LEARN program for weight management . Dallas, TX: American Health Publishing Co., 2000. [Google Scholar]

13. Look AHEAD Research Group, Wadden TA, West DS, et al. The Look AHEAD study: a description of the lifestyle intervention and the evidence supporting it . Obesity (Silver Spring) 2006; 14 :737–752. [PMC free article] [PubMed] [Google Scholar]

14. Centers for Medicare and Medicaid Services. Services and supplies incident to a physician’s professional services: conditions , 42 CFR §410.26 (2011). [Google Scholar]

15. Kushner RF. Barriers to providing nutrition counseling by physicians: a survey of primary care practitioners . Prev Med 1995; 24 :546–552. [PubMed] [Google Scholar]

16. Batsis JA, Bynum JP. Uptake of the Centers for Medicare and Medicaid obesity benefit: 2012–2013 . Obesity (Silver Spring) 2016; 24 :1983–1988. [PMC free article] [PubMed] [Google Scholar]

17. Tsai AG, Wadden TA, Rogers MA, Day SC, Moore RH, Islam BJ. A primary care intervention for weight loss: results of a randomized controlled pilot study . Obesity (Silver Springs) 2010; 18 :1614–1618. [PubMed] [Google Scholar]

18. Wadden TA, Volger S, Sarwer DB, et al. A two-year randomized trial of obesity treatment in primary care practice . N Engl J Med 2011; 365 :1969–1979. [PMC free article] [PubMed] [Google Scholar]

19. Kumanyika SK, Morales KH, Allison KC, et al. Two-year results of think health! ꜟvive saludable!: a primary care weight-management trial . Obesity (Silver Springs) 2018: 26 :1412–1421. [PMC free article] [PubMed] [Google Scholar]

20. Wadden TA, Walsh OA, Berkowitz RI, et al. Intensive behavioral therapy for obesity combined with liraglutide 3.0 mg: a randomized controlled trial . Obesity (Silver Springs) 2019; 27 :75–86. [PMC free article] [PubMed] [Google Scholar]

21. Wadden TA, Tronieri JS, Sugimoto D, et al. Liraglutide 3.0 mg as an adjunct to intensive behavior therapy in individuals with obesity: SCALE IBT 56-week randomized, double-blind, placebo-controlled trial . Presented at the annual meeting of the European Obesity Congress (ECO), April 29, 2019, Glasgow, Scotland, UK. [Google Scholar]

22. Wadden TA, Bakizada Z, Wadden SZ, Alamuddin N. An overview of the treatment of obesity in adults In Wadden TA, Bray GA., eds, Handbook of Obesity Treatment (2nd edition). New York: Guilford Press, 2018, pp 283–308. [Google Scholar]

23. Thomas JG, Leahey TM, Wing RR. An automated internet behavioral weight-loss program by physician referral: a randomized controlled trial . Diabetes Care 2015; 38 :9–15. [PMC free article] [PubMed] [Google Scholar]

24. U.S. Preventive Services Task Force, Curry SJ, Krist AH, Owens DK, et al. Behavioral weight loss interventions to prevent obesity-related morbidity and mortality in adults: US Preventive Services Task Force recommendation statement . JAMA 2018; 320 :1163–1171. [PubMed] [Google Scholar]