4. Discussion

In Figure 3a and b, the changes of the nonprotein respiratory quotient can be seen for Protocol A and for Protocol B. Figure 4a for Protocol A and Figure 4b for Protocol B show the results of utilized macronutrient intakes carbohydrate CI and fat FI, as well the macronutrient oxidations for carbohydrate CarbOx and

The correlation coefficients between HOMA-IR and Wk, Fk, Rk, Rwk, Rnpk,

The results of goodness of fit of the SAM-HEM metabolic model to the known

(a) Daily metabolized carbohydrate and fat intake and oxidation in protocol A. (b) Daily metabolized

Wk 0.828219114701 0.000003557891 0.6464400352912 0.00154355168 Fk 0.873415795383 0.000000235522 0.9999138255551 0.00000000000 Rk 0.77770999896 0.000033325699 0.971975709210 0.00000000000 Rwk 0.92129422123 0.000000003126 0.967645037740 0.00000000000 Rnpk 0.935321042397 0.000000000000 0.9529177800117 0.00000000000 χ<sup>k</sup> 0.9354436104 0.000000000000 0.952802142972 0.00000000000

Protocol A Protocol B

Mean deviation in grams

Standard deviation

Correlation coefficients between HOMA-IR and Wk, Fk, Rk, Rwk, Rnpk, and χ<sup>k</sup>:

Goodness of fit of the SAM-HEM metabolic model to Wk, Fk, and Lk data.

Standard deviation

Wk 220.1045 893.8795 289.2232 512.9418 Fk 11.8265 122.9743 13.3447 94.0426 Lk 21.8630 90.8594 35.0486 40.6444

Protocol A Protocol B HOMA-IR P value HOMA-IR P value

trajectory of weight, fat weight, and lean mass are shown in Table 2.

and χ<sup>k</sup> along with their P value are shown in Table 1.

Type 2 Diabetes - From Pathophysiology to Modern Management

carbohydrate and fat intake and oxidation in protocol B.

Mean deviation in grams

fat FatOx.

Figure 4.

Table 1.

Table 2.

120

Insulin resistance is a pathogenic factor for type 2 diabetes. Insulin resistance has a deleterious impact on glucose and lipid metabolism, blood pressure, coagulation abnormality, inflammation, oxidative stress, and endothelial dysfunction. Population studies suggest that insulin resistance is an important target to reduce CVD risk [30]. A significant proportion of apparently healthy subjects are insulin resistant. About 30–40% of subjects are afflicted with insulin resistance in affluent countries, and the total number is over 1 billion worldwide [30]. HOMA-IR-estimated insulin resistance is associated with subsequent symptomatic CVD in the general population independent of all classic and several nontraditional risk factors [30]. The main result of EPBE [29] is that it clearly demonstrates the profound effect of energy perturbation on changes of insulin resistance. Insulin resistance remained slightly impaired at the end of Protocol A (CR followed by RF) as opposed to the end of Protocol B (OF followed by CR) where the insulin resistance created by OF was normalized by CR. As it is discussed by the authors of [29], the benefit of calorie restriction in terms of improvement of insulin sensitivity is firmly established in the medical literature in various disorders like binge eating with bulimia, weight cycling, obesity, and type II diabetes. In the EPBE study, euglycemic clamp measurements were performed parallel to the HOMA-IR. Observing these in parallel, an overarching picture emerges that the sugar and insulin dynamics are strongly connected to quantifiable dynamics of body composition and the fat metabolism as well as the carbohydrate- vs. fat-burning energy utilization.

Our feasibility demonstration for the main features of CPS is focused on assessing changes of insulin resistance. Using trial data from EBPE [29], we correlated HOMA-IR as a surrogate marker for insulin resistance with our surrogate markers such as R-ratio, Rw-ratio, 24 h nonprotein respiratory quotient, and fat-burning fraction. We found high correlation across the examined metabolic variables Wk, Fk, Rk, Rwk, Rnpk, and χ<sup>k</sup> with HOMA-IR along with highly significant P value for each examined variable.

The implication is that these results show strong evidence for the feasibility for our concept to a have a noninvasive long-term monitoring tool for insulin resistance for users in their natural environment. Displaying Wk, Fk, Rk, Rwk, Rnpk, and χ<sup>k</sup> on MHM and MST via our CPS can provide the needed tool to users and their providers to observe and use adaptive control strategies to improve the otherwise undetectable and invisible phenomena caused by insulin resistance and reach metabolic health.

A new method for metabolic research has been introduced here to extend the principles of indirect calorimetry to a broader application which considers serial measurements of changes of body composition and hydration status with no gas exchange measurements and is still able to estimate 24 h nonprotein respiratory quotient. For this purpose, a Lagrangian functional L was set up to establish the quantitative relationships between changes of fat mass, weight, and energy balance. Without calorie counting and just using the required input weight change ΔWk, fat mass change ΔFk, and energy balance EBk, the fat and nonfat energy balance can be estimated along with important semi-stable energy parameters of the metabolism including the weight-related alpha α^wk, the energy density parameter for weight change ϱ^ <sup>W</sup> k, the weight-related Rw-ratio Rwk, the lean mass-related alpha α^k, the energy density parameter for lean mass change ϱ^Lk, the lean mass-related R-ratio Rk, the nonprotein respiratory quotient Rnpk, and the fat-burning fraction χk. Finding proof for the quantitative relationship between insulin resistance and Rwk, Rk, Rnpk, and χ<sup>k</sup> was difficult due to lack of previous studies [24] with the

needed measurements and due to non-availability of individual data of participants of potentially qualified metabolic studies. Thompson and Slezak [25] showed first that weight and fat loss are correlated well with markers of insulin resistance/ sensitivity. Kelley et al. [22] was able to show that in vivo insulin sensitivity was related to a higher in vitro capacity for fat oxidation of skeletal muscle samples. The same author found also that the strongest predictor of improved insulin sensitivity was associated with enhanced fasting rates of fat oxidation. In this context "metabolic flexibility" in the skeletal muscles is discussed in the literature [24, 31]. One definition of metabolic flexibility is the ability to switch from fat to carbohydrate oxidation during insulin-stimulated glucose disposal. Another definition of metabolic flexibility is the capacity for the organism to adept fuel oxidation to fuel availability [31]. The opposite of metabolic flexibility is metabolic inflexibility which is an important feature of insulin resistance. In the state of insulin resistance, the fuel switching is impaired, and there is an impaired capacity to upregulate muscle lipid oxidation. Metabolic inflexibility and state of insulin resistance manifest as decreased fasting rates of fat oxidation and the lack of further suppression of fat oxidation during heightened level of insulin action postprandially [32]. A defining characteristic of metabolic inflexibility is when after a fat-rich diet, an impaired drop of overnight fasting RQ (impaired fat oxidation) can be observed. Further, insulin-resistant subjects manifest less lipid oxidation during fasting condition and greater lipid oxidation during insulin-stimulated conditions relative to non-insulinresistant subjects. The failure to augment lipid oxidation during fasting conditions likely is a key mechanism leading to lipid accumulation within skeletal muscle [32]. Supporting evidence for impaired lipolysis, diminished fat oxidation, and metabolic inflexibility was confirmed recently in obese girls with polycystic ovary syndrome and increased insulin resistance [33].

substrate utilization and fat-burning capability. This becomes significant when we want to measure metabolic flexibility and create a metric for metabolic health in general. As insulin resistance (and HOMA-IR) is connected to mortality, so is metabolic inflexibility which could be now measured outside of a metabolic laboratory. In earlier publications of ours, we found evidence for close correlation already between HOMA-IR and R-ratio in a wide variety of clinical conditions including obesity, postmenopausal state, metabolic syndrome, and prediabetes /2, 3, 24/. Data from EPBE /34/ prove now that the connection between insulin resistance and Rratio or metabolic flexibility/inflexibility exists across human physiology and pathophysiology in health or disease. Actually, EPBE /34/ helped defining the quantifiable meaning of "metabolic health," and we have now practically usable metrics for

Cyber-Physical System for Management and Self-Management of Cardiometabolic Health

It is important to consider why visceral obesity and the associated increased waist circumference are a good predictor for CVD mortality [33]. The visceral fat leads to high concentration of fatty acids which contributes to impaired liver metabolism and fatty liver. The visceral adipose tissue has been shown to be loaded with macrophages which contribute to the pro-inflammatory profile of visceral obesity which would drive endothelial dysfunction and contributes to mortality. The visceral obesity-induced lipo-toxicity eventually leads to ectopic fat depositions not just in the liver but also the heart, kidney, and also skeletal muscle [34]. For management of visceral obesity, prediabetes, metabolic syndrome, and type 2 diabetes, it is important to know that physically very active persons afflicted with these conditions experience 50% reduction of CVD risk burden [33]. Further, physical activity induces a selective mobilization of visceral adipose tissue and ectopic fat even in the absence of weight loss. Consequently, our "leap ahead" innovation to unify metabolic function assessment with cardiopulmonary fitness assessment may provide an important tool to fight for less insulin resistance and higher cardiorespiratory fitness. CPS has the promise to become a comprehensive cardiometabolic function assessment tool in freely moving individuals requiring only wearing a smart watch and using a specialized stand-up scale (high accuracy bioimpedance

Increased insulin resistance states in obesity, prediabetes, metabolic syndrome, and type 2 diabetes represent a high-risk state for CVD. Restoration of impaired insulin resistance and its manifestation of impaired glucose tolerance can significantly reduce the risk of future diabetes in prediabetics and decrease the estimated CVD risk [34]. The diabetes prevention program (DPP) [35, 36] showed a clear reduction in diabetes incidence in participants assigned to the lifestyle interventions or metformin. Actually, lifestyle intervention was about twice as effective as metformin for prevention of diabetes and was the only intervention associated with regression to normal glucose regulation. Seeing the overwhelming evidence of importance of lifestyle change, we propose to utilize a CPS-like approach as

outlined in introduction to help this process. CPS can be used to observe SVs such as weight, lean body mass, fat mass, R-ratio, Rw-ratio, calculated 24 h nonprotein respiratory quotient, and fat-burning fraction from serial measurements of weight, fat mass, and daily energy balance estimates EBk. EBk can be obtained either as per Eq. (5) with no calorie counting requirement or for enhanced accuracy with calorie counting and measurements of the physical activity energy expenditure and following Eq. (20). As a workable answer to behavioral changes mentioned in the introduction, we propose using SVs for "(A) gold standard measure" for metabolic functioning and as "(D) outcome measures for optimization" as a foreseen requisite to make breakthroughs in the fight against obesity and insulin resistance [3, 15, 16]. The predictive power of SAM-HEM can draw trajectories of SVs and allow for trend

it explaining also the title of this chapter.

DOI: http://dx.doi.org/10.5772/intechopen.84262

analyzer) for serial measurement of fat mass and weight.

123

The main likely mechanism of metabolic inflexibility is that the impaired capacity to upregulate muscle lipid oxidation in the face of high lipid supply may lead to increased muscle fat accumulation and insulin resistance [31]. Many studies have shown when people are in energy balance, the 24 h food fraction φk, fat-burning fraction χk, and nonprotein respiratory quotient Rnpk match each other [31]. With the current technology, metabolic flexibility can be studied in a metabolic chamber by measuring RQ. The testing modalities include overnight sleep study with RQ measurement or measuring RQ in response to high-carbohydrate diet or in response to high-fat diet [31]. The overnight study can show that the subject with metabolic inflexibility would burn less fat during fasting state than the individual with normal metabolism. At least 2 days of waiting is needed for seeing a clear difference in response between flexible and inflexible individuals when dietary changes are performed because adaptive mechanisms of the body prevail initially. The person with metabolic inflexibility would burn less sugar compared with a person with metabolic flexibility in response to high-carbohydrate diet. Conversely, the fat burning is better in the normal metabolism than the impaired flexibility in response to the high-fat diet. After 6–7 days, an equilibrium sets in again, and the final RQs become indistinguishable between sufferer of inflexibility and healthy, and the 24 h food fraction φ<sup>k</sup> and fat-burning fraction χ<sup>k</sup> settle close to the same value [31]. In summary, it is tempting to speculate that a CPS equipped with the capability to monitor nonprotein respiratory quotient Rnpk could detect flexibility vs. inflexibility in response to dietary challenges of the user in his or her natural environment.

It is important to point out that the energy perturbation study EPBE /34/ was done in healthy men with no confounding metabolic abnormalities. Nevertheless, the correlation analysis reveals the profound connection between insulin resistance change (as measured by HOMA-IR) and energy metabolism with manifestations of

## Cyber-Physical System for Management and Self-Management of Cardiometabolic Health DOI: http://dx.doi.org/10.5772/intechopen.84262

substrate utilization and fat-burning capability. This becomes significant when we want to measure metabolic flexibility and create a metric for metabolic health in general. As insulin resistance (and HOMA-IR) is connected to mortality, so is metabolic inflexibility which could be now measured outside of a metabolic laboratory. In earlier publications of ours, we found evidence for close correlation already between HOMA-IR and R-ratio in a wide variety of clinical conditions including obesity, postmenopausal state, metabolic syndrome, and prediabetes /2, 3, 24/. Data from EPBE /34/ prove now that the connection between insulin resistance and Rratio or metabolic flexibility/inflexibility exists across human physiology and pathophysiology in health or disease. Actually, EPBE /34/ helped defining the quantifiable meaning of "metabolic health," and we have now practically usable metrics for it explaining also the title of this chapter.

It is important to consider why visceral obesity and the associated increased waist circumference are a good predictor for CVD mortality [33]. The visceral fat leads to high concentration of fatty acids which contributes to impaired liver metabolism and fatty liver. The visceral adipose tissue has been shown to be loaded with macrophages which contribute to the pro-inflammatory profile of visceral obesity which would drive endothelial dysfunction and contributes to mortality. The visceral obesity-induced lipo-toxicity eventually leads to ectopic fat depositions not just in the liver but also the heart, kidney, and also skeletal muscle [34]. For management of visceral obesity, prediabetes, metabolic syndrome, and type 2 diabetes, it is important to know that physically very active persons afflicted with these conditions experience 50% reduction of CVD risk burden [33]. Further, physical activity induces a selective mobilization of visceral adipose tissue and ectopic fat even in the absence of weight loss. Consequently, our "leap ahead" innovation to unify metabolic function assessment with cardiopulmonary fitness assessment may provide an important tool to fight for less insulin resistance and higher cardiorespiratory fitness. CPS has the promise to become a comprehensive cardiometabolic function assessment tool in freely moving individuals requiring only wearing a smart watch and using a specialized stand-up scale (high accuracy bioimpedance analyzer) for serial measurement of fat mass and weight.

Increased insulin resistance states in obesity, prediabetes, metabolic syndrome, and type 2 diabetes represent a high-risk state for CVD. Restoration of impaired insulin resistance and its manifestation of impaired glucose tolerance can significantly reduce the risk of future diabetes in prediabetics and decrease the estimated CVD risk [34]. The diabetes prevention program (DPP) [35, 36] showed a clear reduction in diabetes incidence in participants assigned to the lifestyle interventions or metformin. Actually, lifestyle intervention was about twice as effective as metformin for prevention of diabetes and was the only intervention associated with regression to normal glucose regulation. Seeing the overwhelming evidence of importance of lifestyle change, we propose to utilize a CPS-like approach as outlined in introduction to help this process. CPS can be used to observe SVs such as weight, lean body mass, fat mass, R-ratio, Rw-ratio, calculated 24 h nonprotein respiratory quotient, and fat-burning fraction from serial measurements of weight, fat mass, and daily energy balance estimates EBk. EBk can be obtained either as per Eq. (5) with no calorie counting requirement or for enhanced accuracy with calorie counting and measurements of the physical activity energy expenditure and following Eq. (20). As a workable answer to behavioral changes mentioned in the introduction, we propose using SVs for "(A) gold standard measure" for metabolic functioning and as "(D) outcome measures for optimization" as a foreseen requisite to make breakthroughs in the fight against obesity and insulin resistance [3, 15, 16]. The predictive power of SAM-HEM can draw trajectories of SVs and allow for trend

needed measurements and due to non-availability of individual data of participants of potentially qualified metabolic studies. Thompson and Slezak [25] showed first that weight and fat loss are correlated well with markers of insulin resistance/ sensitivity. Kelley et al. [22] was able to show that in vivo insulin sensitivity was related to a higher in vitro capacity for fat oxidation of skeletal muscle samples. The same author found also that the strongest predictor of improved insulin sensitivity was associated with enhanced fasting rates of fat oxidation. In this context "metabolic flexibility" in the skeletal muscles is discussed in the literature [24, 31]. One definition of metabolic flexibility is the ability to switch from fat to carbohydrate oxidation during insulin-stimulated glucose disposal. Another definition of metabolic flexibility is the capacity for the organism to adept fuel oxidation to fuel availability [31]. The opposite of metabolic flexibility is metabolic inflexibility which is an important feature of insulin resistance. In the state of insulin resistance, the fuel switching is impaired, and there is an impaired capacity to upregulate muscle lipid oxidation. Metabolic inflexibility and state of insulin resistance manifest as decreased fasting rates of fat oxidation and the lack of further suppression of fat oxidation during heightened level of insulin action postprandially [32]. A defining characteristic of metabolic inflexibility is when after a fat-rich diet, an impaired drop of overnight fasting RQ (impaired fat oxidation) can be observed. Further, insulin-resistant subjects manifest less lipid oxidation during fasting condition and greater lipid oxidation during insulin-stimulated conditions relative to non-insulinresistant subjects. The failure to augment lipid oxidation during fasting conditions likely is a key mechanism leading to lipid accumulation within skeletal muscle [32]. Supporting evidence for impaired lipolysis, diminished fat oxidation, and metabolic inflexibility was confirmed recently in obese girls with polycystic ovary syndrome

Type 2 Diabetes - From Pathophysiology to Modern Management

The main likely mechanism of metabolic inflexibility is that the impaired capacity to upregulate muscle lipid oxidation in the face of high lipid supply may lead to increased muscle fat accumulation and insulin resistance [31]. Many studies have shown when people are in energy balance, the 24 h food fraction φk, fat-burning fraction χk, and nonprotein respiratory quotient Rnpk match each other [31]. With the current technology, metabolic flexibility can be studied in a metabolic chamber by measuring RQ. The testing modalities include overnight sleep study with RQ measurement or measuring RQ in response to high-carbohydrate diet or in response to high-fat diet [31]. The overnight study can show that the subject with metabolic inflexibility would burn less fat during fasting state than the individual with normal metabolism. At least 2 days of waiting is needed for seeing a clear difference in response between flexible and inflexible individuals when dietary changes are performed because adaptive mechanisms of the body prevail initially. The person with metabolic inflexibility would burn less sugar compared with a person with metabolic flexibility in response to high-carbohydrate diet. Conversely, the fat burning is better in the normal metabolism than the impaired flexibility in response to the high-fat diet. After 6–7 days, an equilibrium sets in again, and the final RQs become indistinguishable between sufferer of inflexibility and healthy, and the 24 h food fraction φ<sup>k</sup> and fat-burning fraction χ<sup>k</sup> settle close to the same value [31]. In summary, it is tempting to speculate that a CPS equipped with the capability to monitor nonprotein respiratory quotient Rnpk could detect flexibility vs. inflexibility in response to dietary challenges of the user in his or her natural environment. It is important to point out that the energy perturbation study EPBE /34/ was done in healthy men with no confounding metabolic abnormalities. Nevertheless, the correlation analysis reveals the profound connection between insulin resistance change (as measured by HOMA-IR) and energy metabolism with manifestations of

and increased insulin resistance [33].

122

analysis and prediction and serve as "(B) …tools allowing for planning and executing dynamic changes of behavior," as desired by behavior scientists [15, 16]. The desire for a (C) dynamic behavior change model development using self-directed behavioral change strategies can arrive with further development of CPS using control equations like the Hamilton-Jacobi-Bellman equation for dynamic optimized control [20] and with further technologies of artificial intelligence.

Glossary

ϱ^

125

A. Appendix

Measured variables

Fk fat weight

Wk weight

DOI: http://dx.doi.org/10.5772/intechopen.84262

FIk fat intake

Lk lean mass

Rk R-ratio Rwk Rw-ratio

α^<sup>k</sup> estimation of α<sup>k</sup>

α^wk estimation of αwk ϱ<sup>F</sup> ≈9:4 Kcal/g energy density for fat

ϱ^Lk estimation of ϱ<sup>L</sup>

anywhere outside of a metabolic laboratory.

<sup>W</sup> <sup>k</sup> estimation of ϱWk φ<sup>k</sup> fat intake fraction χ<sup>k</sup> fat-burning fraction

Derived or estimated variables

ΔFk fat mass change in 24 h

BMRk basal metabolic rate CIk carbohydrate calorie intake COk oxidized carbohydrate calories ECWk extracellular water mass

FOk oxidized fat calories ICWk intracellular water mass

ΔLk lean mass change in 24 h

TEEk total energy expenditure

Rnpk nonprotein respiratory quotient

ϱ<sup>L</sup> ≈1:8 Kcal/g the energy density for lean mass

ϱWk energy density for weight

MEIk metabolically utilized energy intake

PAEk physical activity energy expenditure via smart watch sensors

α<sup>k</sup> first-order term coefficient of the lean body—fat

αwk first-order term coefficient of the weight—fat

The following mathematical descriptions use elementary mathematics and a minimum set of assumptions, similar to the landmark article of [37]. These equations could be regarded as an extension of the work of [37], with my main points being that the fat-burning fraction can be calculated from serial fat and weight changes, producing the same result as 24 h indirect calorimetry. Importantly, the equations allow the clinician to determine the nonprotein respiratory quotient with serial weight and fat weight measurement, avoiding the necessity for gas exchange analysis. An important advantage of this mathematical method is that it can be used

logarithmic relationship Tylor series expansion

logarithmic relationship Tylor series expansion

ΔWk body weight change in 24 h EBk daily energy balance

Cyber-Physical System for Management and Self-Management of Cardiometabolic Health

The main contribution of this chapter to medicine and life science is that it lays out a framework using CPS to observe and monitor long-term SVs of the metabolism including markers of insulin resistance. The CPS approach may point to new and promising directions to find workable solutions to challenges of unhealthy metabolic conditions such as insulin resistance, obesity, prediabetes, metabolic syndrome, and type 2 diabetes.
