Characteristics of studies that applied a population attributable fraction (PAF) approach to estimating direct healthcare costs of physical inactivity (n=23)
| First author and year of publication | Country | Data sources | Definition of PA minimal risk counterfactual | Types of costs | Conditions included | Adjusted PAF* | Comorbidity† | Findings‡: amount (% healthcare cost), uncertainty/sensitivity analysis | Time frame | Funding/COI |
| Peer-reviewed scientific paper | ||||||||||
| Allender 200758 | UK | NHS 2002 total expenditure data, NHS 1992–1993 expenditure by disease code data | 2.5 hours MPA or 1 hour VPA/week | Inpatient, outpatient, primary and community care, pharmaceutical | IHD, stroke, breast cancer, colon cancer, diabetes | No/unclear | No | ₤1.06 billion (1.5%) Converted national estimate: $2.0 billion INT | 1 year (2002) | British Heart Foundation/no COI declared |
| Bielemann 201529 | Brazil | Brazilian Unified Health System data 2013, Brazil National Household Sample Survey 2008 | Any leisure time PA | Inpatient costs only | IHD, stroke, hypertension, breast cancer, colon cancer, diabetes, osteoporosis | Yes | No | R$141.9 million, 15% of total inpatient costs of the seven major NCDs Converted national estimate: $86.2 million INT | 1 year (2013) | No funding reported/no COI declared |
| Cadilhac 201127 | Australia | National Health Survey 2004–2005, Australian Burden of Disease data 2003, Disease Costs and Impact Study 2000–2001 | ≥5×30 min MPA or ≥3×20 min VPA/week | Annual health sector cost | IHD, stroke, cancers, fractures, depression | No/unclear | No | $A672 million (1.3%) Converted national estimate: $522.7 million INT | 1 year (2008) | VicHealth/no COI declared |
| Colditz 19996 | USA | Previously published cost estimates for each disease | Any leisure time PA | Hospital care, pharmaceutical, physician care, care in nursing home | IHD, hypertension, breast cancer, colon cancer, diabetes, osteoporotic fractures | No/unclear | No | US$24.3 billion (2.4%), statistical sensitivity analysis: US$37.2 billion (3.7%) Converted national estimate: $37.2 billion INT | 1 year (1995) | Boston Obesity Nutrition Research Centre/COI statement missing |
| Ding 20162 | 142 countries | WHO, World Bank and GBD Study data | ≥150 min MVPA/week | Total health expenditure | IHD, stroke, breast cancer, colon cancer, T2DM | Yes | Yes | $53.8 billion INT worldwide (0.64%), structural and statistical sensitivity analysis: $14.9–147.6 billion INT when using unadjusted PAFs: $123.9 ($40.9–291.2) billion INT | 1 year (2013) | No funding reported/no COI declared |
| Garrett 200460 | USA | Blue Cross databases and Behavioral Risk Factor Surveillance System | ≥5×30 min MPA or ≥3×20 min VPA/week | Inpatient and outpatient medical claim | IHD, stroke, hypertension, breast cancer, colon cancer, T2DM, osteoporotic fractures, depression, anxiety | No/unclear | No | US$83.6 million ($56/member) among US Blue Cross members | 1 year (2000) | No funding reported/Blue Cross and Blue Shield employees involved as authors |
| Janssen 201218 | Canada | Canadian Health Measures Survey 2007–2009, EBIC 2000 | 7-day accelerometry ≥150 min/week | Hospital care, pharmaceutical, physician care, care in other institution and additional care expenditure | IHD, stroke, hypertension, breast cancer, colon cancer, T2DM, osteoporosis | Yes | No | $C2.4 billion (1.4%§), statistical sensitivity analysis: $C1.6–3.1 billion Converted national estimate: $2.1 billion INT | 1 year (2009) | Public Health Agency of Canada/COI statement missing |
| Katzmarzyk 200033 | Canada | EBIC 1993, the PA Monitor Survey | Energy expenditure ≥12.6 kJ/kg/day | Hospital care, pharmaceutical, physician care and research | IHD, stroke, hypertension, breast cancer, colon cancer, T2DM, osteoporosis | No/unclear | No | $C2.1 billion (2.5%), statistical sensitivity analysis: $C1.4–3.1 billion Converted national estimate: $2.3 billion INT | 1 year (1999) | Canadian Society for Exercise Physiology and Health Canada/no COI declared |
| Katzmarzyk 200420 | Canada | CCHS 2000–2001, EBIC 1998/1993 | Expenditure ≥6.3 kJ/kg/day | Hospital care, pharmaceutical, physician care, care in other institution and additional care expenditure | IHD, stroke, hypertension, breast cancer, colon cancer, T2DM, osteoporosis | No/unclear | No | $C1.6 billion (1.5%), statistical sensitivity analysis conducted for total costs Converted national estimate: $1.7 billion INT | 1 year (2001) | Ontario Ministry of Tourism and Recreation/COI statement missing |
| Katzmarzyk 201119 | Canada | CCHS 2009, EBIC 1998 | Expenditure ≥6.3 kJ/kg/day | Hospital care, pharmaceutical, physician care, care in other institution and additional care expenditure | Coronary artery disease, stroke, hypertension, colon cancer, breast cancer, T2DM, osteoporosis | Yes | No | $C1.02 billion in Ontario, Canada | 1 year (2009) | Ontario Ministry of Health Promotion/COI statement missing |
| Krueger 201423 | Canada | NHEX, CCHS 2009, EBIC 1998, Canadian Institute for Health Information Hospital Morbidity Database | Not defined as ‘inactive’ (did not specify) | Hospital care, pharmaceutical, physician care, other healthcare professionals (excluding dental), health research and other | IHD, stroke, hypertension, breast cancer, colon cancer, T2DM, osteoporosis | Yes | No | $C3 billion (1.4%), statistical sensitivity analysis conducted for combined risk factors Converted national estimate: $2.5 billion INT | 1 year (2012) | No funding reported/no COI declared |
| Krueger 201522 | Canada | NHEX, CCHS 2012, EBIC 2008 | Leisure time energy expenditure ≥1.5 kcal/kg/day | Same as above | Same as above | Yes | No | $C3.27 billion (1.6%¶), quoted previous statistical sensitivity analysis ±17% Converted national estimate: $2.7 billion INT | 1 year (2013) | No funding reported/no COI declared |
| Krueger 201621 | Canada | NHEX, CCHS 2011–2012, EBIC 2008 | Leisure time energy expenditure ≥1.5 kcal/kg/day | Same as above | Same as above | Yes | No | $C349.6 million for British Columbia, Canada, quoted previous statistical sensitivity analysis ±17% | 1 year (2013) | Ministry of Health and Provincial Health Services Authority/COI statement missing |
| Maresova 201431 | Czech Republic | Czech Republic European Health Interview Survey 2008, WHO GBD Study, data from health insurance companies that cover 75% of all healthcare expenditures | ≥150 min/week MPA, ≥75 min/week VPA, or ≥180 min/week walking, or any combination resulting in 600 MET min over at least 3 days/week | Not specified | IHD, ischaemic stroke, breast cancer, colon cancer, T2DM | No/unclear | No | 693 million CZK (0.35%) Converted national estimate: $58.8 million INT | 1 year (2008) | University of Economics, Prague/COI statement missing |
| Martin 200125 | Switzerland | Health-enhancing PA survey 1999, a published study on costs associated with each disease, accident statistics from the Swiss Council for Accident Prevention | ≥5×30 min MPA or ≥3×20 min VPA | Not specified | CVD, hypertension, breast cancer, colon cancer, T2DM, osteoporosis, back pain, depression | No/unclear | No | 2.7 billion SFr (structural sensitivity analysis conducted) | Not specified | No funding reported/COI statement missing |
| Popkin 200657 | China | China Health and Nutrition Survey 2000, National Health Services Survey 1998 | Not specified | Total costs: inpatient, outpatient, pharmaceutical, and other | IHD, stroke hypertension, breast cancer, colon cancer, endometrial cancer, T2DM (also included costs of NCDs indirectly through overweight/obesity) | No/unclear | No | US$1.35 billion (2.4%¶) Converted national estimate: $4.3 billion INT | 1 year (2000; projected 2025 cost provided) | No funding reported/no COI declared |
| Scarborough 201134 | UK | NHS Programme Budgeting estimates, WHO GBD Project | Achieving some PA at work, home, for transport or during discretionary time | All spending in primary and secondary care services | IHD, stroke, breast cancer, colon cancer, diabetes | No/unclear | No | £0.9 billion (0.75%¶), structural sensitivity analysis: £0.9–1.0 billion Converted national estimate: $1.6 billion INT | 1 year (2006–07) | British Heart Foundation/COI statement missing |
| Zhang 201326 | China | Chinese Behavioral Risk Factors Surveillance 2007, National Health Services Survey 2003 | ≥5×30 min MPA or ≥3×20 min VPA/week | Hospital care, pharmaceutical, physician care and additional health expenditures | IHD, stroke, hypertension, cancer, T2DM (also included costs of NCDs indirectly through overweight/obesity) | Yes | No | US$3.5 billion (2.4%¶) Converted national estimate: $9.1 billion INT | 1 year (2007) | Nike Inc./no COI declared |
| Grey literature | ||||||||||
| Colman 200416 | Canada | CCHS 2000–2001, EBIC 1998 | Expenditure ≥1.5 kcal/kg/day | Hospital care, pharmaceutical, physician care, other institutions (including research), additional drug expenditure, private spending on medical care | CVD, cancer, endocrine and related diseases, musculoskeletal diseases | No/unclear | No | $C210.8 million for British Columbia, Canada | 1 year (2001) | B.C. Ministry of Health Planning/COI statement missing |
| International Sports and Culture Association and Centre for Economics and Business Research 201517 | EU-28 | WHO, Organization for Economic Cooperation and Development, Eurostat, International Development Association, EUCAN and published studies | ≥150 min MPA or ≥75 min VPA/week, or combinations | Direct costs: healthcare expenditure Indirect costs: DALYs | IHD, breast cancer, colorectal cancer, T2DM | No/unclear | No | UK: €1920 million (1.06%§); Germany: €1677 million (0.55%§); Italy: €1562 million (1.04%§); France: €1215 million (0.51%§); Spain: €992 million (1.03%§); Poland: €219 million (0.86%§). EU-28: €9.2 billion Converted national estimates: UK $2.4; Germany $2.2; Italy $2.1; France $1.5; Spain $1.5; Poland $0.5 billion INT | 1 year (2012) | International Sport and Culture Association (contributors included various organisations and companies)/COI statement missing |
| Market Economics Limited 201324 | New Zealand | Various sources including the Ministry of Health, Statistics New Zealand, District Health Board reports, and others | ≥30 min PA×5 days/week | Hospital care, pharmaceutical, outpatient, public health and other | IHD, stroke, hypertension, breast cancer, colorectal cancer, T2DM, osteoporosis, depression | No/unclear | Yes | $614 million NZD (4.6%), statistical sensitivity analysis conducted (+2%) Converted national estimate: $464.4 million INT | 2010 | Government commissioned/COI statement missing |
| Stephenson 200030 | Australia | Active Australia 1997 National PA Survey; RR from studies on PA and disease; Australian Institute of Health and Welfare’s Disease Costs and Impact Study | Inactivity ≥150 min/week | Hospital care, pharmaceutical, medical services, allied health, research, public health and other | IHD, stroke, breast cancer, colon cancer, depression | Yes | No | $A377 million (1.1%#; structural sensitivity analysis conducted) Converted national estimate: $433.2 million INT | 1 year (1993–1994) | Commonwealth Department of Health and Aged Care and Australian Sports Commission/COI statement missing |
| Public Health England 2016 | UK | Programme budgeting data released by NHS England in 2010–2014 | Not specified | Not specified | IHD, stroke, breast cancer, colon cancer, diabetes | No/unclear | No | £455 million for England, UK (0.3%#) Converted national estimate: $657.8 million INT | 1 year (2013–2014) | Public Health England/COI statement missing |
*Adjusted PAF: whether PAF used was based on relative risks adjusted for confounders. Yes=explicitly described adjustment in the paper; No/unclear=did not describe adjustment in the paper, we could not use a consistent methodology to determine whether the PAF was crude or adjusted but not stated.
†Comorbidity: whether the potential double counting among comorbidities was addressed (yes/no).
‡Findings: interpreted as the total amount of direct healthcare cost that was associated with physical inactivity (all findings referred to the general population with the exception of Garrett 2004,60 which referred to all Blue Cross members ≥18 years). % interpreted as the percentage of overall healthcare cost that was spent on diseases that were attributable to physical inactivity. In most cases, the percentages were reported in the original studies; in some cases, the author (DD) calculated or recalculated the percentages based on national healthcare expenditure data from the WHO (available at http://apps.who.int/nha/database/ViewData/Indicators/en).
§Recalculated and corrected by the authors of the current review.
¶Calculated or recalculated percentages.
A, Australian dollars; C, Canadian dollars; CCHS, Canadian Community Health Survey; COI, conflict of interest; CVD, cardiovascular disease; CZK, Czech Koruna; DALYs, Disability Adjusted Life Years; EBIC, Economic Burden of Illness in Canada; EU-28, 28 member countries of the European Union; GBD: Global Burden of Disease; IHD, ischaemic heart disease; INT, international dollars; MET, metabolic equivalents; min, minutes; MPA, moderate physical activity; MVPA, moderate-to-vigorous physical activity; NCD, non-communicable disease; NHEX, National Health Expenditure Database for Canada; NHS, National Health Service; NZD, New Zealand dollars; min, minutes; PA, physical activity; PAF, population attributable fraction; £, pounds sterling; R, Brazil real; T2DM, type 2 diabetes mellitus; RR, relative risks; SFr, Swiss francs; VPA, vigorous physical activity.