What could a public health strategy for weight loss look like? Paul Aveyard Professor of behavioural medicine

Transcription

1 What could a public health strategy for weight loss look like? Paul Aveyard Professor of behavioural medicine

2 Conflicts of interest In this research and in other similar research, commercial weight loss companies and pharmaceutical companies have donated their products free of charge to the NHS to allow the trial to proceed.

3

4

5 WEIGHT REGAIN

6 Blood pressure Cholesterol concentration (mmol/l) lipids (males) Systolic Diastoli BMI (kg/m 2 ) Non-HDL HDL BMI (kg/m 2 )

7 Weight regain in BWMP over extended follow-up (BOCF analysis) NICE Review

8 Diabetes Prevention Program: Sustained reductions in diabetes incidence - despite weight regain Lancet, 14 (2009), pp

9 Obesity Reviews (2004) 5, 43 50

10 Mortality by trial arm in the WOSCOPS trial All-cause CVD CHD Non-cardiovascular Circulation Mar 15; 133(11):

11 Hypertension 2005:45:

12 Weight loss and CVD outcomes: SCOUT trial Independent of study treatment or later weight change, each 1 kg lost during the lead-in period resulted in: 6.2% risk reduction for Primary Outcome Event: MI, Stroke, any CVD death+ resuscitated cardiac arrest 6.4% risk reduction for death of any cause N Engl J Med Sep 2;363(10): Diabetes, Obesity and Metabolism , Issue 6, pages , June 2012.

13 EFFECTIVE AND SCALABLE INTERVENTIONS

14 SELF-HELP INTERVENTIONS

15 % What do we know about self-directed weight loss? Percentage of men and women attempting to lose weight currently Health Survey for England BMI<22 Men Women BMI>30 International Journal of Obesity :

16 P articipants: adults with BMI 25 kg/m 2 I nterventions: self-help programmes aiming to achieve weight loss through changes in diet or activity. Self-help defined as interventions that could feasibly be delivered in self-management context (e.g. used by individuals for a weight loss attempt not assisted by health care professionals, counsellors, or any other kind of person-to-person support). C omparators: another self-help intervention or a minimal control O utcomes: weight change at six and/or 12 months S tudy design: randomized controlled trials Am J Public Health Mar;105(3):e43-57.

17 Systematic review of self-help interventions 3883 results retrieved 186 full text screened 23 studies met our criteria (43 references, 9,623 participants) 18 studies included in quantitative synthesis (meta-analyses) 39 interventions: 18 tailored and interactive 6 interactive, not tailored 3 tailored, not interactive 12 fixed

18 Self-help interventions versus minimal controls (BOCF; 6 months) Intervention Control Mean Difference Mean Difference Study or Subgroup Mean SD Total Mean SD Total Weight IV, Random, 95% CI IV, Random, 95% CI Tailored and interactive Byrne 2006 McConnon 2007 Morgan 2011 Morgan 2013 Shapiro 2012 Subtotal (95% CI) % 15.0% 7.5% 10.7% 14.7% 59.9% Heterogeneity: Tau² = 2.94; Chi² = 24.96, df = 4 (P < ); I² = 84% Test for overall effect: Z = 2.11 (P = 0.04) [-4.56, -1.24] 0.30 [-0.71, 1.31] [-4.60, 1.00] [-6.55, -2.65] [-1.77, 0.37] [-3.50, -0.13] Interactive non-tailored Greene 2013 Nakata 2011 Subtotal (95% CI) Heterogeneity: Tau² = 0.00; Chi² = 0.01, df = 1 (P = 0.91); I² = 0% Test for overall effect: Z = 4.39 (P < ) Static Morgan 2013 Subtotal (95% CI) Heterogeneity: Not applicable Test for overall effect: Z = 3.25 (P = 0.001) Total (95% CI) % Heterogeneity: Tau² = 1.52; Chi² = 29.53, df = 7 (P = ); I² = 76% Test for overall effect: Z = 3.57 (P = ) Test for subgroup differences: Chi² = 1.77, df = 2 (P = 0.41), I² = 0% [-2.86 to -0.83] p = % 13.2% 28.8% 11.3% 11.3% [-2.58, -0.82] [-3.00, -0.20] [-2.42, -0.93] [-4.81, -1.19] [-4.81, -1.19] [-2.86, -0.83] Favours intervention Favours control

19 Self-help interventions versus minimal controls (BOCF; 12 months) Intervention Control Mean Difference Mean Difference Study or Subgroup Mean SD Total Mean SD Total Weight IV, Random, 95% CI IV, Random, 95% CI Tailored and interactive Haapala 2009 McConnon 2007 Morgan 2011 Patrick 2011 Shapiro % 22.5% 11.3% 28.0% 21.3% Subtotal (95% CI) % risk of bias (49% intervention versus Heterogeneity: Tau² = 0.72; Chi² = 10.64, df = 4 (P = 0.03); I² = 62% Test 70% for overall control effect: Z = 1.55 participants (P = 0.12) followed up Results sensitive to one study at high at 12 months) Removing this study reduced Total (95% CI) % Heterogeneity: Tau² = 0.72; Chi² = 10.64, df = 4 (P = 0.03); I² = 62% Test statistical for overall effect: heterogeneity Z = 1.55 (P = 0.12) to low and Test for subgroup differences: Not applicable yielded a significant effect in favour of the intervention [-4.08, -0.72] 0.70 [-0.49, 1.89] [-4.46, 0.26] [-1.48, 0.08] [-1.69, 0.89] [-1.73, 0.20] [-1.73, 0.20] Favours intervention Favours control

20 IN-PERSON DELIVERED INTERVENTION

21 Weight loss in kilograms Commercial weight loss programmes 4.5 Weight losses at programme end Commercial weight management Primary care NHS group No intervention BMJ 2011;343:d6500

22 Weight loss at one year p< Standard care Commercial programme Lancet (9801):

23 BWMP Control Mean Difference Mean Difference Study or Subgroup Mean SD Total Mean SD Total Weight IV, Random, 95% CI IV, Random, 95% CI Commercial + meal replacements Rock 2010 (JC in person) % [-9.57, -5.63] Rock 2010 (JC phone) Subtotal (95% CI) % 100.0% [-8.05, -3.95] [-8.39, -5.26] Heterogeneity: Tau² = 0.23; Chi² = 1.22, df = 1 (P = 0.27); I² = 18% Test for overall effect: Z = 8.54 (P < ) Group based commercial Heshka 2003 (WW) % [-4.14, -1.86] Jebb 2011 (WW) % [-3.00, -1.58] Jolly 2011 (RC) % [-3.15, 1.15] Jolly 2011 (SW) % [-2.81, 1.21] Jolly 2011 (WW) Subtotal (95% CI) % 100.0% [-4.58, -0.22] [-2.89, -1.54] Heterogeneity: Tau² = 0.13; Chi² = 5.00, df = 4 (P = 0.29); I² = 20% Test for overall effect: Z = 6.40 (P < ) Automated internet Hersey 2012 Subtotal (95% CI) % 100.0% [-1.37, -0.03] [-1.37, -0.03] Heterogeneity: Not applicable Test for overall effect: Z = 2.05 (P = 0.04) Primary care Jolly 2011 (GP) % 0.30 [-1.55, 2.15] Jolly 2011 (pharmacist) Munsch 2003 Nanchahal 2011 Wadden 2011 Subtotal (95% CI) Heterogeneity: Tau² = 0.35; Chi² = 6.17, df = 4 (P = 0.19); I² = 35% Test for overall effect: Z = 1.00 (P = 0.32) 17.5% 8.7% 36.7% 20.8% 100.0% 0.40 [-1.36, 2.16] [-6.16, -0.64] [-1.18, 0.58] [-2.35, 0.75] [-1.34, 0.43] Obesity Reviews 2014:15: Test for subgroup differences: Chi² = 59.27, df = 3 (P < ), I² = 94.9% Favours BWMP Favours control

24 Weight change over 2 years BI CP12 CP No. Participants Primary Analysis Weight change Month BI CP12 CP52 Standard error bars shown around mean estimates Ahern et al Lancet in press

25 Cost-effectiveness

26 Cumulative incidence per 100,000 population Cumulative incidence of obesity-related disease wk programme vs brief intervention 52-wk programme vs 12-wk programme 52-wk programme vs brief intervention

27 Direct Healthcare costs ( million) per 100,000 Cumulative total direct healthcare costs in Millions (+95%CL) avoided per 100,000 by year week programmes vs Brief intervention 52 week programme vs brief intervention 52 week programme vs 12 week programme

28 Cost-effectiveness Taking intervention costs into account, the ICER for the 12-week programme was dominant in comparison to the brief intervention for the period , resulting in 643 additional QALYs per 100,000 individuals, at a cost-saving of 68,000 per 100,000 individuals. Taking into account intervention costs, the 52-week programme resulted in 1925 additional QALYs gained per 100,000 individuals at a cost of 4 8million per 100,000 individuals. The ICER ( 2498/QALY) indicated that the 52-week programme was cost-effective compared to the brief intervention for the 2015 to 2039 period.

29 EFFECTIVE DELIVERY MECHANISMS

30 THE BWeL TRIAL Lancet Nov 19;388(10059):

31 The brief intervention Offer help Book them in Create accountability To create momentary motivation To capitalise on the moment To create lasting motivation

32 Advice increases quit attempts by 24% Offering support on how to quit increases them by 68% to 117% Direct comparison offer help vs offer advice increases quit attempts by 39% to 69%

33 JAMA Intern Med. 2013;173(6):

34 12 months prolonged abstinence 7.3% vs 1.8% Preventive Medicine 62 (2014) Addiction Jul;111(7):1257-6

35 THE CONSORT FLOW DIAGRAM

36

37 Screened and potentially eligible Inelligible Potentially eligible

38 What happened to the potentially eligible Not willing - No anon data Not willing + Anon data Not Eligible GP WD Eligible & Enrolled

39 Reasons for non-eligibility Pregnant Weight loss programme within 3 months Poor English Current weight loss programme GP visit for weight BMI<30

40 Reasons for GP exclusion Not appropriate in consultation Clinically inappropriate Other Unknown

41 BASELINE CHARACTERISTICS

42 BMI of participants

43 Age of participants

44 HOW PEOPLE FELT

45 Ratings of appropriateness by trial arm

46 Ratings of helpfulness by trial arm

47 ACCEPTING HELP TO LOSE WEIGHT AND THE IMPACT ON WEIGHT AT 1 YEAR

48

49 Of those who accepted referral No booking Did not attend Start but not complete course Complete course

50 0 Control Weight change at 3 months Intervention (95%CI -2.17; -1.35), p<

51 0 Control Weight change at 12 months Intervention (95%CI -1.97; -0.89), p<

52 100 Percentage losing 5% and 10% of baseline weight at 12 months Odds ratio 2.11 (95% CI 1.67; 2.68), p<0.001 Odds ratio 2.41 (95%CI 1.72, 3.38), p< % 10% Control Intervention

53 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Percentage of people taking action and type of action taken by 12 months in the two arms of the trial Control intervention Active intervention No action Self-help action Effective action

54 PUTTING IT ALL TOGETHER: THE INTERVENTIONS + THE DELIVERY MECHANISM

55 In preparation

56 A few slides removed as material was unpublished.

57

58 Thank you For questions or a copy of the slides paul.aveyard@phc.ox.ac.uk