More physical activity may be needed to prevent HF than some other CVD events

 The title is a reworded version of the title of the article which is designated as reference 2 in the footnotes.

There are  2 (at least 2) epidemiologic studies that indicate a linear dose-response relationship between physical activity (PA) and the risk of heart failure. While one study ( Pandy et al see below) does demonstrate a "modest" reduction in HF risk at a lower levels of PA, both studies report a more robust reduction in HF at higher exercise levels.

Pandy et al (Circulation 2015,, see ref 1 below) did a meta-analysis involving about 370  thousand subjects , 20 thousand of which developed HF over a 13 years period.They compared the HF risk in 3 categories based on level of exercise, namely 500 MET-min per week ,1000 MET-min per week and 2000 MET-min per week.

500 MET-min per week is equivalent to 2.5 hours of "moderate" exercise per week or 1.25 hours of "vigorous" exercise per week.Moderate is defined a exercise requiring 3-5.9 MET and vigorous as about 7 METS. (It requires about7 METS to run a 15 minute mile or to finish Stage 2 on the Bruce protocol treadmill exercise tests. One should be able to walk a fifteen minute mile with a 02 consumption of 5 METS.)

1000 MET-hrs per week is 5 hours of moderate or 2.5 hours of vigorous exercise per week and 2000 as twice that or 10 hours of moderate exercise per week. Yes, that does seem like a lot,

Both the 2008 US exercise  guidelines and the 2018 guidelines recommend at least 500 but state that more benefits accrue with higher levels .

Pandy reported a linear dose response with a "marked reduction in risk at very high doses of PA ( about 35%) at 2000 MET-min per week".This would be 10 hours of moderate or 5 hours ofvigorous aerobic activity.

Their data:
exercise level                                HF RF

500 MET minutes  per week            0.9 (0.87-0.92)
1000 "  " " " " " " " "                       0.81(0.77-0.80
2000 "  "   "   "" """""                     0.65 (0.58-0.73)

Quoting the authors; "Only a moderate reduction ( about 10%) risk in HF noted at the minimal ( US guidelines) recommended level,"

The authors offer a mechanistic explanation namely that CAD event risk reduction occurs at a lower level of exercise by reducing the usual suspect Risk factors (BP,Lipids,blood sugar control) while HF risk reduction occurs at a higher levels of exercise perhaps  by altering cardiac function and structure, i.e beneficial remodeling,perhaps mitigating the age related loss of left ventricular compliance

An earlier article Patel K, (Int J Cardiol 2013 see ref 2) had reached generally similar conclusions regarding the levels of exercise needed to decrease HR risk versus the amount adequate to  reduce general  CV risk,

Patel et al studied 5503 patients age 65 and older
During the 13 years of follow up incident HF developed in:
26% of those with little or no regular exercise
23% of those with "low"level of exercise
20% of those with "moderate"
19 % of those with high .

Low was defined a 1-499 Met-min per week
medium as 500-999
high as greater than 1000

The HRs for incident HF were:
low level exercise     0.87 (0.71-1.06)
medium                     0.68 (0.54-0.85) stat sig 
high                            0.60 (0.49-0.74) stat sig 

All exercise level groups had reduced HR for incident MI, stroke, and cardiovascular mortality but the lowest level group did not have a statistically significant  decrease in HF risk.

These 2 studies indicate that more exercise is associated with greater reduction in HF risk, although some  HF decrease was noted in the lowest exercise group in the Pandy study.

Now for something completely different. A different type of study published by CR de Fillippi from the Cardiovascular Health study ( reference 3) provides an interesting  insight regarding possible mechanism(s) by which more exercise is better in terms of HF risk reduction. 

They studied 2, 9333 subject free of HF at the onset and  who had normal values for two biomarkers, NT-Pro BNP and cTnT (troponin).They then measured these markers every 2 to 3 years and compared incident increase in those markers in groups divided by their exercise levels.(They quantitated exercise using  a particular system and used numerical scores to define groups and I was not able  to translate those values in to MET hrs to compare with the other 2 articles)

 They found compared with the participants with the lowest PA activity scores those with the highest (i.e. more exercise) had an low odds ratio of 0.50 (0.33-0.77) for a significant increase in NT proBNP and an OR of 0.3 for and increase in troponin.  Quoting  the authors "increased levels of NTProBNP and cTnT may reflect pathological precursors of hemodynamic stress and injury that are prevented by PA at some as yet more precisely defined level.

So how to put all this together.

An  appealing ( at least to me) mechanistic explanation is that an overall decrease in CV mortally and arguably a slight reduction in HF risk  can be brought about by modest exercise and even  perhaps slightly less  than modest levels  by reducing the usual suspect,traditional CVD risk factors.Modest here refers to about 500 MET hrs per week. but some studies have shown a decreased heart attack risk at even lower exercise levels.

A more robust risk reduction in HF risk (perhaps 20-35%) seems  to require higher levels of exercise  ( in the 1000 Met-hrs per week range and higher) by perhaps bringing about an advantageous remodeling of the heart or at least a mitigation of the deleterious remodeling of the heart than occurs with sedentary aging which in turn may predispose to the development of heart failure with preserved ejection fraction (HFpEF).

To the extent that CVD risk factor reduction can decrease heart attacks and the  accompanying reduction in   heart function some reduction in HF risk would be likely, i.e.some reduction in heart failure with reduced ejection fraction ( HFrEF). But at least half of cases of HF are not associated with decreased systolic function ( at least as measured by the ejection fraction  (EF)) but by the echocardiographic findings of decreased left ventricular compliance and distensibility which according to the work of Dr. Ben Levine and others at SW  Medical School in Dallas predispose to diastolic heart failure  (HFpEF) .That group suggests that at least 30 minutes of exercise 4 to five times a week might be sufficient to prevent that age related loss of cardiac compliance. ( see ref 4 below for more on this argument) and perhaps prevent at least some cases of HFpEF. It would be nice if as little as 30 minutes 5 times a week would be sufficient.


1)Pandy,A "Linear,dose dependent inverse association between PA(physical activity) and heart failure risk Circ. 115.132 p 1786-1794

2.Patel K. "Prevention of heart failure in older adults may require higher levels of physical activity than needed for other cardiovascular events." Int J Cardiol. 2013, Oct 168 (3) 1905-1909)

3)de Fillippi, CR "Physical activity, change in biomarkers of myocardial stress and injury, and subsequent heart failure risk in older adults. JAm Coll Cardiol.. 2012 Dec. 18;60 (24), 2539-47

4) Bhella, PS  "Impact of lifelong exercise "dose" on left ventricular compliance and distensibility"J Amer coll cardiology. 2014 , 64, no 12,p 1267  


Addendum 7/13/20 In the table of Pandy's data I had erroneously type Met hr instead of Met minutes.