SEARCH

SEARCH BY CITATION

Keywords:

  • calorimetry;
  • gelatinization;
  • image analysis;
  • in vitro digestibility;
  • potato starch

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgments
  8. References

ABSTRACT:  Starch granule microstructure affects the digestion of starch and its nutritional impact; however, the exact relationship between both factors is not clear. This study reports quantitative relationships between granule size (length and polygonal area), degree of gelatinization (DG), in vitro digestibility (by enzymatic methods), and glycemic response of potato starch granules gelatinized to various extents by heating at several constant temperatures in the range of 55 to 65 °C. DG measured by differential scanning calorimetry was closely related with heating temperature (R2= 0.997), size parameters of granules (measured by image analysis), in vitro digestion, and in vivo glycemic response (R2 of adjusted models > 0.9); shape parameters of granules (measured by image analysis) were not related with DG. Results demonstrate that DG of starch strongly affects its digestibility in vitro, and may influence the postpandrial glycemic response. Future studies should be performed to investigate the effect of potato starch gelatinization on the nutritional impact at other temperatures and in more complex matrices.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgments
  8. References

There is increased evidence that the microstructure of foods affects its nutritional value (Aguilera 2005; Brennan 2005; Parada and Aguilera 2007). As set forward by Riccardi and others (2003)“…food structure plays an important role in determining the accessibility of starch to digestion, thus influencing the postprandial blood glucose response, which modulates plasma insulin and lipid levels.” In particular, altering starch structure in foods using thermal processing changes the postprandial responses (Björck 1996; Tester and others 2004). For example, different potato-based foods have different glycemic response (listed in parenthesis assuming a glycemic index of 100 for glucose): baked potatoes (76.5 ± 8.7), instant mashed potatoes (87.7 ± 8.0), boiled red potato after cooling (56.2 ± 5.3), and fresh fried potatoes (63.6 ± 5.5) (Fernandes and others 2005).

When raw starch granules are gelatinized during heating, the disruption of starch structure increases its susceptibility to enzymatic degradation (Holm and others 1988). In many starchy foods, however, a portion of residual starch is not fully gelatinized during processing, usually due to limited water content or insufficient heating. Examples of such foods include breakfast cereal flakes and some baked products (Rashmi and Urooj 2003; Venn and Mann 2004; Sozer and others 2007). The postprandial responses of foods containing raw or partially gelatinized starches have become the subject of increasing interest in recent years. Slowly digested carbohydrates are generally considered to be beneficial for the dietary management of metabolic disorders, including diabetes and hyperlipidemia (Brand-Miller 2003; Wolever 2003; Lehmann and Robin 2007)

Different processing conditions affect the gelatinization of starch granule as measured by the enzymatic method or the loss of birefringence. For example, starch in baked products is found almost completely gelatinized in angel food cake and white bread (for example, no birefringence of granules and 97% gelatinization as measured enzymatically) or practically in native form as in sugar cookies where approximately 91% of granules were birefringent and only 4% gelatinized (Lineback and Wongsrikasem 1980).

Englyst and others (1992) classified starch based on its digestibility (measured by an in vitro enzymatic method) as: RDS, rapidly digestible starch or starch that was fully hydrolyzed within 20 min of incubation; SDS, slowly digestible starch as the starch digested during the period between 20 and 120 min, and; RS, resistant starch, the starch fraction not hydrolyzed within 120 min. In turn, RS is divided into 4 subclasses: physically protected, ungelatinized resistant granules, retrograded, and chemically modified starches (Nugent 2005).

The aim of this study was to generate a quantitative relationship between some microstructural parameters that characterize the gelatinization process of pure potato starch, the degree of gelatinization (DG) measured by DSC, and the digestibility using an in vitro technique. In particular, we were interested in the relation between the size of birefringent granules and its relation with process temperature, and how the DG affects starch digestibility, hence, its glycemic response.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgments
  8. References

Isolated potato starch (IPS)

The commercial IPS “Chuño Delicado” (packaged by Elabal Ltda., Chile), a dry powder, was purchased in a local supermarket.

Degree of gelatinization (DG) by differential scanning calorimetry (DSC)

Starch samples (approximately 7 mg) suspended in approximately 15 μL distilled water were hermetically sealed in DSC pans (Sample Pan Kit 0219–0062, Perkin Elmer Co., Norwalk, Conn., U.S.A.) and kept at room temperature for approximately 12 h. Each sample was heated for 60 min at a constant temperature selected between 55 and 65 °C in a differential scanning calorimeter DSC-7 (Perkin-Elmer Co.), and cooled to 10 °C to obtain samples with different DG (from 0% to 100%). Next, samples were scanned against a blank (empty pan) from 15 to 99 °C at a 10 °C/min scanning rate. The 1020 Series DSC 7 thermal analysis software (Perkin-Elmer Co.) was used to collect data and analyze onset (To), peak (Tp), end (Te) temperatures, and the transition enthalpy (ΔH). Equipment was calibrated using indium as the reference material. The DG of each processed sample was calculated in relation with that of native potato starch (Eq. 1) (Miah and others 2002; Sozer and others 2007). Numerical results are averages of 3 independent replicates.

  • image(1)

Measurement of size and shape of ungelatinized granules by light microscopy (microstructural analysis)

Starch dispersions (approximately 1.8 mg of starch per 15 mg of distilled water) were heated isothermally at several temperatures (55, 56, 57, 58.3, 60, 62, and 65 °C) around the onset temperature of gelatinization of IPS, 57 °C (Ratnayake and Jackson 2007) for 60 min in sealed pans using the same differential scanning calorimeter as before. After heating, the contents having different degrees of gelatinization (DGs), were viewed with a digital camera Cool Snap Pro Color, (Photometrics Roper Div. Inc., Tucson, Ariz., U.S.A.) attached to a model BX50 optical microscope (Olympus Corp., Tokyo, Japan) and illuminated with extra-bright 100 W halogen light. Birefringence was assessed by inserting a polarizing filter. A video monitor connected directly to the camera output displayed the image of birefringent granules in real time. Frames were transferred to a personal computer (Dell Inc., Round Rock, Tex., U.S.A.) via the parallel port. Each image (1392 × 1040 pixels) was saved as 8 bits TIFF image file, without compression. Shape and size parameters of the birefringent zone of granules were estimated by image analysis using the software Image-Pro PLUS, version 4.5 for Windows (Media Cybernetics Inc., Silver Spring, Md., U.S.A.). Polygon area is the total area included in the polygons defining the granule's outline divided by the total number of birefringent granules. Similar definitions apply for aspect (ratio between major axis and minor axis of the ellipse equivalent to object) and maximum length of granules (length of longest line joining 2 points of the granule's outline and passing through the centroid). Numerical results are averages of 6 independent replicates.

In vitro digestion studies

Starch digestibility was determined as described by Englyst and others (1992), with modifications, for native starch and samples heated at 57.5, 60, and 62.5 °C for 60 min. The fully gelatinized starch control treated at 100 °C for 20 min. Six grams of porcine pancreatic alpha-amylase nr 7545 (Sigma-Aldrich, St. Louis, Mo., U.S.A.) were dispersed in water (40 mL) by magnetic stirring for 10 min. The dispersion was then centrifuged for 10 min at 30000 g, and a portion of the supernatant (32 mL) was transferred to a beaker. Two milliliters of amyloglucosidase nr 9913 (Sigma-Aldrich) were diluted in 3 mL of deionized water. The solution was freshly prepared for the digestion analysis. Aliquots of guar gum solution (10 mL, 5 g/L in 0.05 M HCL) and sodium acetate solution (5 mL, 0.5 M) were added to the gelatinized starch samples (500 mg in 10 mL of distilled water) in test tubes. Guar gum provided viscosity to the starch samples. Several glass balls (7 mm diameter) and 10 mL of the enzyme solution were then added to each tube, followed by incubation in a water bath (37 °C) with agitation (170 rpm). Aliquots (0.5 mL) were taken at intervals and mixed with 4 mL of 80% ethanol, and the glucose content in the mixture measured using glucose oxidase and peroxidase assay kits nr GAGO-20 (Sigma-Aldrich). The equilibrium percentage of enzymatically hydrolyzed starch was defined as the glucose content after 60 min hydrolysis divided by that of fully gelatinized starch × 100. Numerical results are averages of 3 independent replicates.

In vivo glycemic response

The glycemic response of starch with different DG was assessed in a single individual (man, 29 years old, 60 kg weight, and 1.70 m tall) who ingested 1 sample per day after fasting until 9 am. Starch samples (50 g of starch and 500 mL of water) were prepared with different DG by heating at 55, 57, 60, and 65 °C for 60 min to yield DGs of 0%, 38.3%, 73.2%, and 99.9%, respectively. Blood glucose concentration (postprandial response) was monitored for 120 min after consumption (0, 15, 30, 45, 60, 90, and 120 min), and the maximum blood glucose concentration after meal (Max. C.) and area under curve after meal (AUC) were calculated from the incremental curves (FAO/WHO 1998). The blood glucose concentration was also measured 5 and 10 min before ingestion, and the average value taken as the base line. Blood was obtained by finger prick using Ascensia MICROLET adjustable lancing device (Bayer HealthCare, Mishawaka, Ind., U.S.A.). Blood glucose was measured using Ascensia ELITE XL diabetes care system (Bayer HealthCare), and controlled according to manufacturer's specific instructions (Granfeldt and others 2006; Burton and Lightowler 2008). Numerical results are averages of 3 independent replicates obtained in 3 different days.

Data analysis

A regression analysis was performed to relate each dependent variable (microstructural parameter, in vitro digestion, in vivo glycemic response) with the independent variable DG. Analysis was performed using specialized software Statgraphics Plus for Windows 4.0 (StatPoint Inc., Herndon, Va., U.S.A.). Equation of the best fitted model and R2 were obtained.

Results and Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgments
  8. References

DG of IPS in excess water (that is, IPS/water ratio where the raw starch melts completely during DSC analysis) is a function of time and temperature of heating (Lund 1984). Figure 1 shows that as a sample was heated (at a constant temperature) the DG increased with heating time and converged to an asymptotic value. Evidently, faster kinetics of gelatinization and higher asymptotic values of the DG were attained at higher temperatures. The final DGs obtained at 57.5, 60, and 62.5 °C were 46%, 73%, and 88%, respectively. Data for DG of IPS as a function of temperature for a constant heating period of 60 min are shown in Figure 2. Below 55 °C the DG was 0%, while for temperatures above of 65 °C the DG amounted always to 100% (see below). Thus, the temperature window between 55 and 65 °C permitted a good control of the DG and was used in the rest of the study. Similar results have been found previously (Lund 1984; Miah and others 2002).

image

Figure 1—. Changes in DG with time of potato starch suspensions during heating at constant temperature (57.5, 60, and 62.5 °C). Points are the average of 3 samples; intervals are ± SD.

Download figure to PowerPoint

image

Figure 2—. To, Tp, Te, and DG of potato starch suspensions as a function of heating temperature. Points are the average of 3 samples; intervals are ± SD for each level. Fitted equations and R2 statistics are showed in Table 1.

Download figure to PowerPoint

From DSC analysis, To, Tp, and Te for raw starch were 61.84 ± 0.55 °C, 65.55 ± 0.59 °C, and 77.40 ± 1.31 °C, respectively, while ΔH = 14.97 ± 0.42 J/g. These values are in accordance with previous reports (Yuan and other 2007). For partly gelatinized samples, Ratnayake and Jackson (2007) found that the residual gelatinization enthalpy (or the gelatinization enthalpy after heating granules at specific temperatures) of samples preheated at a constant temperature between 35 and 65 °C decreased from 22 to 0 J/g. Samples that had been heated above 65 °C exhibited no residual enthalpy.

Models adjusted to data in Figure 2 are presented in Table 1. To, Tp, and Te varied linearly with heating temperature (Table 1). In particular, the exponential model predicting DG as a function of T for 55 °C ≤T ≤ 65 °C was DG (T) = 115.12 (1 −e[−0.202(T–55)]) explained 99.7% of the variability in DG. Nonlinear relationships have been reported previously relating DG and processing temperature (Miah and others 2002); however, no appropriate mathematical models were proposed.

Table 1—.  Fitted equations and R2 statistics for To, Tp, Te, and DG, measured by DSC.
PropertyFitted equationR2
ToTo = 0.993T + 13.730.994
TpTp = 0.920T + 20.630.989
TeTe = 0.336T + 62.500.668
DGDG = 115.12 (1 −e[−0.202(T−55)])0.997

Characteristic temperature values of the gelatinization process, To, Tp, and Te, of samples heated at different temperatures (thus having different DGs) are also shown in Figure 2. Clearly, To and Tp changed with process temperature (R2= 0.994 and 0.989, for To and Tp, respectively), whereas Te varied only slightly (R2= 0.668) (Table 1). Similar results have been found in previous studies (Sozer and others 2007). In practical terms, values of To and Tp may help in estimating the DG of starch in potato-based foods, that is, given the value of To or Tp the DG may be estimated from Figure 2. One advantage is that To and Tp are not function of the amount of starch in a DSC sample (in contrast, ΔH determination depends on the starch content). The trend in To and Tp with temperature may reflect the fact that a population of starch granules has a distribution of melting temperatures and the “survivor” granules after heating at any processing temperature have gelatinization temperatures higher than those that had already gelatinized. Incidentally, Karlsson and Eliasson (2003) found that isolated potato starch gelatinized at a lower temperature and with a higher transition enthalpy than when present within potato tissue. In addition, it has been reported that transition temperatures (To, Tp, and Te) and ΔH in the presence of excess water depends on the botanic source (Liu and others 2006; Ratnayake and Jackson 2007), so values reported in this study cannot be used for other types of starches.

Microscopy analysis was used to assess the number and microstructural aspects (size and shape) of birefringent (ungelatinized) granules after heating at different constant temperatures. Figure 3 depicts how the normalized values (value at any processing temperature divided by value for 55 °C) of aspect, maximum length and polygon area of the birefringent zone of granules, and the number of birefringent granules (in the field of view) related to DG. While the aspect of granules did not change with heating temperature (or DG), polygon area showed the largest drop in normalized value with DG; from 1 to approximately 0.4. The regression equation for polygon area had twice the slope of that representing the maximum length, and hence, it was selected as the microstructural parameter of reference. The number of birefringent granules remaining after heating also exhibited a good dependence on DG (in fact, it was the only to drop to 0 at 100% gelatinization); however, it is not a geometrical parameter. Results show that not all granules melt together, in contrast, each granule (or granules fraction) has its independent temperature of melting, and this result strengthens the idea that microstructural differences between granules determines their properties. Similar results were found by Srikaeo and others (2006), where with the increased cooking time or temperature, the number of starch granules decreased following a clear tendency.

image

Figure 3—. Normalized values of maximum length, polygon area, aspect, and number of nongelatinized granules (birefringent granules) as a function of DG. Data are averages from 6 images.

Download figure to PowerPoint

Figure 4 shows how the cumulative frequency of the polygon area of the birefringent zone of granules varied after heating for 60 min at each of the 7 selected temperatures. As the heating temperature increased, the mean polygonal area of the granules (area included in the polygon defining the object's outline) decreased and the distribution of areas became narrower (that is, a steeper cumulative curve). Since higher heating temperatures resulted in larger DGs, it means that the remaining native granules (birefringent) belonged to the smaller granules within the original population. In other words, larger granules tended to gelatinize before smaller ones, a fact that is well documented in the literature (Chiotelli and Le Meste 2002; Singh and Kaur 2004; Svihus and others 2005; Kaur and others 2007).

image

Figure 4—. Cumulative frequency of polygon area of birefringent potato starch granules obtained by image analysis. Samples were heated for 60 min at the 7 listed temperatures.

Download figure to PowerPoint

In vitro enzymatic conversion of heat-treated starches is a long-standing method to determine DG (Chiang and Johnson 1977; Varriano-Martson and others 1980). The procedure is based on the fact that swelling and hydration during the gelatinization process enhances the chemical reactivity of starch granules towards amylolytic enzymes (Di Paola and others 2003). Because the enzymatic method of Englyst and others (1992) was applied on actual food matrices (white bread, cooked potato, cooked spaghetti), the kinetics of isolated potato starch were quite fast (Chung and others 2006), reaching completion before 5 min. Figure 5 depicts that the equilibrium percentage of enzymatically hydrolyzed starch (C) for samples having 0% to 100%DG (by DSC method) increased nonlinearly from almost 55% to nearly 100%. Interestingly, 0%DG did not correspond with 0% digestibility determined by the method of Englyst and others (1992); this observation accords with results found by Englyst and others (1999), where the percentage of digestible starch of raw potato starch (0%DG) was 45% and not 0%; however, Englyst and others (1999) found that not all digestible starches are rapidly digestible, while our result show that all digestible starches were rapidly digested, and this can be explained by natural differences between different starch sources. This result further emphasizes that starch heated at 55 °C (or below) could experience any structural changes, or some granules can melt, or raw starch may be digested partially (Ratnayake and Jackson 2007; Noda and others 2008).

image

Figure 5—. Percentage of enzymatically hydrolyzed starch (C) as a function of DG. Symbols are the average of 3 samples; intervals show ± SD.

Download figure to PowerPoint

Procedures to determine the glycemic response in humans are established (FAO/WHO 1998; Wolever 2003) and it is known that the variability of the glycemic response between individuals is larger than within individual subjects (Frost and Dornhorst 2000). To have preliminary data that may justify more involved experiments with humans, the glycemic response after ingestion of selected samples was recorded for 1 of the authors (JP) and results are shown in Figure 6. The average basal blood glucose concentration (BBGC) in the preingestion period (during fasting) was 4 mmol/L, so the values reported in Figure 6 actually represent the change in BBGC above this reference value during the postprandial time. Curves had a common shape, that is, an initial increase from BBGC to a maximum value and a slower decay to the base level, as reported in the literature for in vivo determination of glycemic index of rapidly digestible starch (Fernandes and others 2005; Granfeldt and others 2006). Even less processed (almost native) starch elicited a slight increase from BBGC during the first 15 min after ingestion. In fact, the mean of tmax/treturn value was 0.43 (tmax= time between ingestion and that for maximum concentration; treturn= time between ingestion and that for return to basal level). On the other hand, Englyst and others (1999) found a good correlation between rapidly available glucose in vitro (sum of RDS and free sugar glucose) and glycemic response in vivo for starchy foods, and this accords with results shown in this study; while higher is the percentage of rapidly digestible starch in a sample, higher is the glycemic response (Brennan 2005).

image

Figure 6—. Glycemic response of potato starch samples having 4 different DG (inset). Symbols represent the average of triplicates ingested in different days; intervals are ± SD.

Download figure to PowerPoint

Figure 7 shows values of Max. C. and AUC, calculated from incremental curves (Figure 6) as a function of DG. Results show a clear correlation between DG and Max. C. (R2= 0.988) and AUC (R2= 0.931) values, and this means that gelatinized fraction, that is rapidly available for digestion in our samples, is reflected in blood glucose as has been proposed by Englyst and others (1999), and that an adequate management of manufacturing process to control the DG of starchy foods is a way to control their nutritional effects. Some error bars in Figure 6 and 7 are high, and this can be explained for the variability within individual subjects reported by Frost and Dornhorst (2000); however, results suggest that triplicate measurement per subject is adequate to compensate this variation.

image

Figure 7—. In vivo glycemic response values calculated from incremental curves (Figure 6) as a function of DG. Crosses are maximum blood glucose concentration after meal (Max. C.); Circles are area under curve after meal (AUC). Symbols are the average of 3 samples; intervals show ± SD.

Download figure to PowerPoint

These preliminary results are very encouraging since they show that different DGs of isolated potato starch generate different glycemic responses (Holm and others 1988). Probably, the digestive enzyme's work is easier and more efficient than when a real food is digested due to the absence of food matrix (protein and dietary fiber, mainly), which in a more complex food acts as a barrier to enzymes, increases the tortuousity and, in general, complicates the mass transfer in the digestive tract (Björck and others 1994; Brennan and others 1996; Schneeman 2002; Tudorica and others 2002; Brennan 2005).

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgments
  8. References

Heating isolated potato starch (IPS) in excess water at temperatures close to the onset of gelatinization process (55 to 65 °C) caused a decrease in mean maximum length and polygonal area of ungelatinized starch granules. These changes of granule size (microstructural parameters) correlated well with the DG measured by differential scanning calorimetry (that is, as the DG increased, granule size decreased in absolute value). Thus, by subjecting IPS to different temperatures for 60 min it was possible to effectively control the DG of a sample because the kinetics of hydration and swelling was slow. Results demonstrated that all granules do not gelatinize simultaneously and large granules increase in size first. The increased proportion of gelatinized granules in a sample led to higher in vitro digestibility measured by the enzymatic method of Englyst and others (1992). This in turn resulted in in vivo glycemic responses varying from practically no increment in blood glucose concentration (above the basal level) for raw starch (0%DG) to one showing a prolonged response with an appreciable maximum (100%DG). Reducing the release of glucose via heat processing of starch would possibly result in a lowered insulin response and greater access to use of stored fat, thus helping in developing starchy foods for the treatment of obesity and weight control. Further studies should be performed on the relation between the food matrix (for example, starchy plant tissues) and DG of starch, and its effect on the glycemic response.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgments
  8. References

Research has been funded by the Marcel Loncin Award of the Inst. of Food Technologists to JMA and a CONICYT doctoral fellowship to JP. Authors thank Mr. Javier Bendek for performing laboratory work related to starch digestibility.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgments
  8. References