study on the effect of ph360 on the framingham-based …€¦ · coaching program, in improving...

STUDY ON THE EFFECT OF ph360 ON THE FRAMINGHAM-BASED 10-YEAR GLOBAL CARDIOVASCULAR DISEASE RISK SCORE

w w w . p h 3 6 0 . m e

STUDY ON THE EFFECT OF ph360 ON THE FRAMINGHAM-BASED 10-YEAR GLOBAL CARDIOVASCULAR DISEASE RISK SCORE

Data collected between 5/1/2014 and 3/16/2018

Research conducted by:

Daniella Remy (Canada), Principal Researcher

Matt Riemann (Australia), Health Scientist

Patrick Moran (US), Statistician

Albert Garoli (Italy), Medical Consultant

Cam McDonald (Australia), Dietician

Bart Goldman (US), Medical Practitioner

w w w . p h 3 6 0 . m e

Executive Summary

The objective of this study was to assess the effectiveness of ph360 Personalized Health and Wellness

Coaching Program, in improving overall cardiovascular health and disease risk. An office-based non-laboratory

Framingham algorithm was used to evaluate each participant’s risk of cardiovascular disease (CVD) within the

next 10 years.

Between May 2014 and March 2018, 1,690 participants provided their anthropometric measurements (including

height, weight, neck circumference, waist circumference and hip circumference) and survey responses on their

blood pressure, cholesterol levels, glycemic balance, physical activity and smoking habits. In total there were

5,220 completed assessments.

Of the 1,690 subjects in this study averaging approximately 50 years of age, 86.7% were female.

The statistical analysis consisted of a a linear mixed-effect regression model with random intercepts for each

participant, adjusting for age, sex, total follow-up time, and interaction between sex and follow-up time. The

following is a summary of the findings:

• From the initial assessment through the end of follow-up, there were significant, beneficial changes in nearly

every metric assessed. This included an absolute decrease in BFI from 35.6 to 31.0 (p<.0001), a 1.5 kg/m2

decrease in BMI (p<.0001), and a 6.4 cm decrease in waist circumference (p<.0001) resulting in a significant

waist-height ratio decrease (p<.0001) from 0.53 at baseline to 0.49 at follow-up.

• There was a significant drop (p<.0001) in the Framingham 10-year CVD risk from 7.0% to 6.5% during follow-

up after adjusting for age and gender demonstrating a positive outcome on the risk of CVD by participating

in the ph360 program.

• CVD risk for women appeared to slowly and steadily decline over time, whereas there was a steeper initial

decrease in CVD risk for men during months 0-4, a plateau during months 4-8, and another steady decrease

in CVD risk from months 8-12 as found by a significant interaction (p=0.0008) between follow-up time and

gender which resulted in differential rates of decrease in Framingham risk over time between women and

men.

• The ph360 Health and Wellness Program provided significant improvements in their participating population

by reducing unwanted weight, reducing fat mass (visceral fat as evidenced by WHtR), improving lifestyle

factors associated with cardiometabolic health issues (including reducing the sedentary lifestyle, smoking

and healthy eating choices) and the overall risk of CVD.

w w w . p h 3 6 0 . m e w w w . p h 3 6 0 . m e

Methods

Participants• Subjects must have participated for at least 3 months in the program

• Subjects must have updated their data at each interval

• Some participants provided data very frequently, in which case intervals of 2-3 months apart for first year, 6

months apart thereafter were kept

• Data on participants who underwent significant health conditions were excluded (eg: pregnancy, severe

illness, hospitalization)

The remaining data consisted of 1,690 participants who provided a total of 5,220 completed assessments. Each

participant provided anywhere from 2 to 10 responses collected between May 2014 and March 2018.

In addition to basic user info (gender, date of birth), the provided data consisted of two main components:

anthropometric measurements and survey responses.

Anthropometric measurements included: height, weight, neck circumference, waist circumference and hip

circumference. Users were allowed to provide their measurements in US imperial (feet, inches, pounds) but

these were then converted into metric (centimeters and kilograms) in the user database.

Survey questions included the following:

1. Have you been told by a medical professional that you have high cholesterol?

2. Have you been told by a medical professional that you have high blood pressure?

3. Have you been diagnosed by a medical professional that you are diabetic or pre-diabetic?

4. Have you been told by a medical professional to watch the blood sugar because it tends to be too high or too

low at times?

5. What is your current level of physical activity?

• Sedentary lifestyle

• Exercising very little

• Getting moderate exercise

• Getting intense exercise

• Bodybuilding naturally

• Bodybuilding with the use of hormones

w w w . p h 3 6 0 . m e

6. Do you smoke?

• Yes, more than 1 pack a day

• Yes, daily but less than 1 pack a day

• Yes, in social environments only

• I use a vaporizer and/ chew tabacco

• No, I quit less than 2 years ago

• No, I don’t smoke at all

The second step in preparing the data was to calculate several new variables. These variables were defined as

follows:

Survey Number:

For each subject, survey date was arranged oldest to newest and assigned a factor (1, 2, 3,…,N). So, survey

number 1 is the first survey the subject took, and survey number N is the last survey the subject took, and in

this way survey number is nested within subject.

Anthropometric variables:

The measurements were used to calculate waist-to-height ratio (WHtR), body mass index (BMI), and body fat

index (BFI) as follows:

• WHtR = Waist / Height

• BMI = Weight / (Height/100)2

• BFI for men = 86.010 x log10 (Waist - Neck) - 70.041 x log10 (Height) + 36.76

• BFI for women = 163.205 x log10 (Waist + Hip + Neck) - 97.684 x log10 (Height) - 78.387

Global 10-year CVD Risk:

We calculated the office-based non-laboratory Framingham 10-year CVD risk using the data provided by our

participants. Rather than using using exact systolic blood pressure (SBP) that most users wouldn’t necessarily

have, we treated the coefficients by assuming hypertensives an SBP of 140 and non-hypertensives an SBP of

120. Also, smokers were considered anyone who smoked cigarettes in any amount, chewed tobacco or used

e-cigarettes. Details on this outcome are provided in the Exploratory Data Analysis (EDA).

w w w . p h 3 6 0 . m e w w w . p h 3 6 0 . m e

Statistical Analysis

The statistical analysis consisted of a sequence of 4 basic parts: Part I Descriptive Statistics, Part II Exploratory

Data Analysis (EDA), and Part III Regression Analysis. The objectives of each part of the analysis are as follows:

Part I, Descriptive Statistics; This portion of the analysis was intended to give us a general sense of the data and

to guide the EDA.

Part II, Exploratory Data Analysis; This portion of the analysis was intended to uncover and highlight interesting

features in the data in more detail, and this information was then used to guide subsequent analyses.

Specifically, interesting features found during the EDA, were subjected to more formal statistical testing.

Part III, Regression Analysis; To determine whether the primary outcome of CVD risk decreased over time, we

formed a linear mixed-effect regression model with random intercepts for each participant, adjusting for age,

sex, total follow-up time, and interaction between sex and follow-up time. Follow-up time was modeled using a

five-knot restricted cubic spline function to avoid a potentially biased linearity assumption, and plots of CVD risk

over follow-up were generated by gender. An identical approach was taken to measure changes in the primary

weight loss outcome, change in BFI, over time.

The analyses were all conducted using SAS 9.4.

w w w . p h 3 6 0 . m e

Results

Demographics

Table 1 and Table 2 below provide an overview of the participants in the study at baseline and follow-up.

Participants were mostly female (86.7%), ranging between 20 and 89 years. Over the course of the follow-up

period, there was a mean(SD) of 3.1(1.5) assessments per participant and a mean(SD) follow-up of 11.1(9.5)

months per participant.

There were significant improvements from baseline (first assessment) in WHtR, BMI and BFI for both males and

females across all age groups, however women seem to have a slow and steady continuous improvement while

men demonstrated slightly more erratic improvements (Table 1). Most importantly, WHtR was able to get under

the 0.5 risk cut-off for cardiovascular risk1-5. Since indices of visceral fat (WHtR and abdominal obesity), appear

to be superior to BMI to screen individuals for cardiometabolic conditions including hypertension, dyslipidemia,

prediabetes/diabetes, and subclinical inflammation, cardiovascular disease, and years of life lost6-9, the greater

reduction in WHtR and BFI are strong indicators of the positive effects of ph360’s Health and Wellness

Program.

Because of the significant risk a sedentary lifestyle10-12, diabetes13-15, and hypertension16-18 contribute to the

overall risk of cardiovascular disease, each of these were examined in the study. Results show a global reduction

in lifestyle-related risk factors in both men and women and across all age groups, especially in relation to the

length of time subjects were actively participating in the program. In other words, the longer they participated,

the greater the lifestyle improvements.

w w w . p h 3 6 0 . m e w w w . p h 3 6 0 . m e

Table 1. Summary of Means and Changes from Baseline on Anthropometric Measures.

w w w . p h 3 6 0 . m e

Table 2. Summary of Means and Changes from Baseline on Lifestyle Factors.

w w w . p h 3 6 0 . m e w w w . p h 3 6 0 . m e

Change in body comopsition over time

To determine the impact of the ph360 program on overall cardiovascular health, a global cardiovascular

disease (CVD) risk score was calculated from participants’ responses using the simple office-based non-

laboratory Framingham algorithm as outlined in Appendix A. Because participants were asked to indicate a

hypertension diagnosis using a binary response (yes/no), rather than giving a numeric systolic blood pressure

(SBP) measurement, we assigned an untreated SBP of 140 (the lower cutoff for hypertension19) for participants

with hypertension and 120 for participants without hypertension. Although this approach underestimates

the variance, it is likely to result in a conservative effect estimate given the realistic expectation that many

hypertensive participants are likely above the 140 SBP cutoff for hypertension. Additionally, we did a sensitivity

analysis by assigning an even more conservative SBP of 130 to hypertensives versus 120 for non-hypertensives.

Table 3 summarizes the mean and standard deviations of the responses at baseline and follow-up with a

significance level of p<0.05 used for all tests performed. From the initial assessment through the end of follow-

up, there were significant, beneficial changes in nearly every metric assessed. Of most interest are the absolute

decreases in BFI from 35.6 to 31.0 (p<.0001), a 1.5 kg/m2 decrease in BMI (p<.0001), and a 6.4 cm decrease

in waist circumference (p<.0001) resulting in a significant waist-height ratio decrease (p<.0001) from 0.53 at

baseline to 0.49 at follow-up. More specifically, the BFI decreased from 37.4 to 32.7 among females and from

23.5 to 19.8 among males (both p<.0001), indicating a positive weight loss outcome for both genders.

Regarding the primary outcome of Framingham 10-year CVD risk, there was a significant drop (p<.0001) from

7.0% to 6.5% during follow-up after adjusting for age and gender. This clearly shows that the ph360 Health and

Wellness Program significantly improved overall cardiovascular health.

w w w . p h 3 6 0 . m e

Table 3. Participant characteristics from participants’ first (Baseline) and last (Follow-up) responses.

w w w . p h 3 6 0 . m e w w w . p h 3 6 0 . m e

Change in Framingham CVD risk over time

To determine whether the primary outcome of CVD risk decreased over time, we formed a linear mixed-effect

regression model with random intercepts for each participant, adjusting for age, sex, total follow-up time, and

interaction between gender and follow-up time. Follow-up time was modeled using a five-knot restricted cubic

spline function to avoid a potentially biased linearity assumption, and plots of CVD risk over follow-up were

generated by gender (Figure 1). We formed an identical model to test changes in BFI (Figure 2). To measure

changes in other secondary outcome measurements (see Table 3) we formed similar models adjusting for age,

gender, and follow-up time. The significance level of p<0.05 was used for all tests performed. Results clearly

indicate a significant improvement from baseline for both men and women. Most interestingly, they seem to

improve at a steady pace and keep their improvement over time.

In the sensitivity analysis in which we substituted the more conservative SBP of 130 for those with hypertension,

CVD risk was still seen to decrease significantly (p<.0001) from a mean(SD) of 6.8(6.0) in the initial assessment

to 6.4(5.2) in the last follow-up, further confirming the above findings.

Figure 1. Change in Framingham Risk** over time with 95% Confidence Intervals.

w w w . p h 3 6 0 . m e

Figure 2. Change in BFI over time with 95% Confidence Intervals.

**Using simple office-based non-lab Framingham score:<https://www.framinghamheartstudy.org/fhs-risk-functions/cardiovascular-disease-10-year-risk/>For systolic blood pressure, those with hypertension were considered to have SBP=140 (untreated) and those without hypertension were considered to have SBP=120 (untreated)

w w w . p h 3 6 0 . m e w w w . p h 3 6 0 . m e

Conclusions

The ph360 Personalized Health and Wellness Program was significantly effective in decreasing the global risk

of cardiovascular disease as assessed using the simple office-based non-laboratory Framingham algorithm. Of

the 1,690 participants in the 3-year study, nearly all showed at least some improvement.

Contributing factors for this improvement included significant reductions in body fat (BFI), visceral adiposity

(WHtR), Weight (BMI). The recommended lifestyle changes encouraged by the ph360 program likely contributed

to the reduction of sedentary behavior, as well as lifestyle-related factors like improvements in blood pressure

and glycemic balance.

Of interest and worth further investigation, are the improvements in the older age groups. Considering the

greater difficulty in achieving weight loss results due to the hormonal and metabolic changes associated with

aging20-21 and the fact that CVD risk tends to increase with age22, the improvements made in the older age

groups stand out.

The ph360 Personalized Health and Wellness Program provided significant improvements in both males and

females across age groups, and helped maintain those positive changes over time. An interaction with gender

was also found, indicating a difference in the longitudinal improvements between males and females. Males

tended to demonstrate an initial steep decrease in CVD risk, plateau during months 4-8, and then show another

steady decrease from months 8-12. Women, on the other hand, appeared to slowly and steadily reduce their

CVD risk over time.

The strong statistically significant changes between baseline and follow-up on anthropometric measures and

lifestyle-related factors combined with the closer examination of changes over timeline intervals clearly shows

that those who participated longer in the ph360 Health and Wellness Program showed greater and more long-

lasting improvements in their overall cardiovascular risk.

w w w . p h 3 6 0 . m e

References

1. Shen, Shiwei, et al. “Waist-to-height ratio is an effective indicator for comprehensive cardiovascular health.”

Scientific reports 7 (2017): 43046.

2. Ashwell, Margaret, and Sigrid Gibson. “A proposal for a primary screening tool:Keep your waist circumference

to less than half your height’.” BMC medicine 12.1 (2014): 207.

3. Ho, Sai-Yin, Tai-Hing Lam, and Edward D. Janus. “Waist to stature ratio is more strongly associated with

cardiovascular risk factors than other simple anthropometric indices.” Annals of epidemiology 13.10 (2003):

683-691.

4. Ashwell, Margaret, and Sigrid Gibson. “Waist to height ratio is a simple and effective obesity screening tool

for cardiovascular risk factors: analysis of data from the British National Diet and Nutrition Survey of adults

aged 19–64 years.” Obesity facts 2.2 (2009): 97-103.

5. Motamed, Nima, et al. “Conicity Index and Waist-to-Hip Ratio Are Superior Obesity Indices in Predicting 10-

Year Cardiovascular Risk Among Men and Women.” Clinical cardiology 38.9 (2015): 527-534.

6. Ashwell, Margaret, et al. “Waist-to-height ratio is more predictive of years of life lost than body mass index.”

PloS one9.9 (2014): e103483.

7. Kodama, Satoru, et al. “Comparisons of the strength of associations with future type 2 diabetes risk among

anthropometric obesity indicators, including waist-to-height ratio: a meta-analysis.” American journal of

epidemiology176.11 (2012): 959-969.

8. Ashwell, M., P. Gunn, and S. Gibson. “Waist-to-height ratio is a better screening tool than waist circumference

and BMI for adult cardiometabolic risk factors: systematic review and meta-analysis.” Obesity reviews 13.3

(2012): 275-286.

9. Savva, Savvas C., Demetris Lamnisos, and Anthony G. Kafatos. “Predicting cardiometabolic risk: waist-to-

height ratio or BMI. A meta-analysis.” Diabetes, metabolic syndrome and obesity: targets and therapy 6

(2013): 403.

10. Wilmot, Emma G., et al. “Sedentary time in adults and the association with diabetes, cardiovascular disease

and death: systematic review and meta-analysis.” (2012): 2895-2905.

11. Biswas, Aviroop, et al. “Sedentary time and its association with risk for disease incidence, mortality, and

hospitalization in adults: a systematic review and meta-analysis.” Annals of internal medicine 162.2 (2015):

123-132.

12. Borodulin, Katja, et al. “Daily sedentary time and risk of cardiovascular disease: The National FINRISK 2002

Study.” Journal of Physical Activity and Health 12.7 (2015): 904-908.

13. Mottillo, Salvatore, et al. “The metabolic syndrome and cardiovascular risk: a systematic review and meta-

w w w . p h 3 6 0 . m e w w w . p h 3 6 0 . m e

analysis.” Journal of the American College of Cardiology 56.14 (2010): 1113-1132.

14. Authors/Task Force Members, et al. “ESC Guidelines on diabetes, pre-diabetes, and cardiovascular diseases

developed in collaboration with the EASD: the Task Force on diabetes, pre-diabetes, and cardiovascular

diseases of the European Society of Cardiology (ESC) and developed in collaboration with the European

Association for the Study of Diabetes (EASD).” European heart journal 34.39 (2013): 3035-3087.

15. Emerging Risk Factors Collaboration. “Diabetes mellitus, fasting blood glucose concentration, and risk of

vascular disease: a collaborative meta-analysis of 102 prospective studies.” The Lancet 375.9733 (2010):

2215-2222.

16. Ettehad, Dena, et al. “Blood pressure lowering for prevention of cardiovascular disease and death: a

systematic review and meta-analysis.” The Lancet 387.10022 (2016): 957-967.

17. Rapsomaniki, Eleni, et al. “Blood pressure and incidence of twelve cardiovascular diseases: lifetime risks,

healthy life-years lost, and age-specific associations in 1· 25 million people.” The Lancet 383.9932 (2014):

1899-1911.

18. Kahan, Thomas. “Focus on blood pressure as a major risk factor.” The Lancet 383.9932 (2014): 1866-1868.

19. Mozaffarian, Dariush, et al. “Heart disease and stroke statistics—2016 update: a report from the American

Heart Association.” Circulation 133.4 (2016): e38-e360.

20. St-Onge, Marie-Pierre, and Dympna Gallagher. “Body composition changes with aging: the cause or the

result of alterations in metabolic rate and macronutrient oxidation?.” Nutrition 26.2 (2010): 152-155.

21. Zamboni, Mauro, et al. “Health consequences of obesity in the elderly: a review of four unresolved questions.”

International journal of obesity 29.9 (2005): 1011.

22. Fowkes, F. Gerald R., et al. “Comparison of global estimates of prevalence and risk factors for peripheral

artery disease in 2000 and 2010: a systematic review and analysis.” The Lancet 382.9901 (2013): 1329-

1340.

Appendix A

The global cardiovascular disease (CVD) risk score was calculated using the simple office‑based

non‑laboratory Framingham algorithm:

https://www.framinghamheartstudy.org/fhs‑risk‑functions/cardiovascular‑disease‑10‑year‑risk/

Framingham Heart Study Simple Office‑based non‑laboratory Predictors of CVD Regression

Coefficients and Hazard Ratios

Men* (10‑year Baseline Survival: So(10) = 0.88431)

Variable Beta‡ p‑value Hazard Ratio 95% CI

Log of Age 3.11296 <.0001 22.49 (14.80, 34.16)

Log of Body Mass Index 0.79277 <.0066 2.21 (1.25, 3.91)

Log of SBP if not treated 1.85508 <.0001 6.39 (3.61, 11.33)

Log of SBP if treated 1.92672 <.0001 6.87 (3.90, 12.08)

Smoking 0.70953 <.0001 2.03 (1.75, 2.37)

Diabetes 0.53160 <.0001 1.70 (1.37, 2.11)

Women* (10‑year Baseline Survival: So(10) = 0.94833)

Variable Beta** p‑value Hazard Ratio 95% CI

Log of Age 2.72107 <.0001 15.20 (8.59, 26.87)




Smoking 0.61868 <.0001 1.86 (1.53, 2.25)

Diabetes 0.77763 <.0001 2.18 (1.63, 2.91)

* The 10‑year risk for women can be calculated as 1‑0.94833 exp(ΣßX – 26.0145)

where ß is the regression coefficient and X is the level

for each risk factor; the risk for men is given as 1‑0.88431 exp(ΣßX – 23.9388)

‡Estimated regression coefficient

16 w w w . p h 3 6 0 . m e

APPENDIX A:

The global cardiovascular disease (CVD) risk score was calculated using the simple office-based non-laboratory

Framingham algorithm:

https://www.framinghamheartstudy.org/fhs-risk-functions/cardiovascular-disease-10-year-risk/

Framingham Heart Study Simple Office-based non-laboratory Predictors of CVD Regression Coefficients and Hazard Ratios

Appendix A

The global cardiovascular disease (CVD) risk score was calculated using the simple office‑based

non‑laboratory Framingham algorithm:

https://www.framinghamheartstudy.org/fhs‑risk‑functions/cardiovascular‑disease‑10‑year‑risk/

Framingham Heart Study Simple Office‑based non‑laboratory Predictors of CVD Regression

Coefficients and Hazard Ratios

Men* (10‑year Baseline Survival: So(10) = 0.88431)

Variable Beta‡ p‑value Hazard Ratio 95% CI

Log of Age 3.11296 <.0001 22.49 (14.80, 34.16)




Smoking 0.70953 <.0001 2.03 (1.75, 2.37)

Diabetes 0.53160 <.0001 1.70 (1.37, 2.11)

Women* (10‑year Baseline Survival: So(10) = 0.94833)

Variable Beta** p‑value Hazard Ratio 95% CI

Log of Age 2.72107 <.0001 15.20 (8.59, 26.87)




Smoking 0.61868 <.0001 1.86 (1.53, 2.25)

Diabetes 0.77763 <.0001 2.18 (1.63, 2.91)

* The 10‑year risk for women can be calculated as 1‑0.94833 exp(ΣßX – 26.0145)

where ß is the regression coefficient and X is the level

for each risk factor; the risk for men is given as 1‑0.88431 exp(ΣßX – 23.9388)

‡Estimated regression coefficient

16 w w w . p h 3 6 0 . m e w w w . p h 3 6 0 . m e

study on the effect of ph360 on the framingham-based …€¦ · coaching program, in improving...

Documents