Effect of Streptozotocin on Plasma Insulin Levels of Rats and Mice: A Meta-analysis Study

BACKGROUND: In the studies focusing on diabetic organisms, Streprozotocine (STZ) is a frequently used agent to induce diabetes in rats and mice. However the current studies do not represent practical importance of their statistical findings. For showing practical importance of the differences in plasma insulin levels of diabetic rats and mice induced by STZ, there should be a statistical synthesis regarding statistical findings of the studies. AIM: The purpose of this study is to make a meta-analysis of the studies on the effect of STZ on plasma insulin levels in diabetic rats and mice. MATERIALS AND METHODS: In this study 39 effect sizes (37 studies) about levels of plasma insulin were analyzed by calculating individual effect sizes (d) and mean effect size. RESULTS: The effect sizes were between -13.7 and +65.3 and the mean effect size value (+9.33) represented a large effect indicating that STZ was an effective agent to significantly decrease plasma insulin levels of diabetic rats and mice. CONCLUSION: It can be said that the differences in plasma insulin levels between STZ-applied and no application groups has a practical importance in making animal model of diabetes.


Introduction
Nowadays, diabetes is frequently seen in society, its prevalence is about 382 million people around the world [1]. Diabetes is characterized by insufficient secretion rate of insulin or lack of insulin activity [2,3]. Diabetes is associated with different health problems including cardiovascular diseases, neuropathy, retinopathy, ulcers and amputations [4,5]. Treatment of diabetes is a complex issue but some animal models were developed to understand the management as diabetes is a chronic condition [6,7]. Over 30 years, alloxan, streptozotocin (STZ, 2-deoxy-2-(3-(methyl-3-nitrosoureido)-D-glucopyranose), highfat diet-fed and nicotinamid are used for establishing experimental diabetes models of animal [8].
Streptozotocin is still commonly used agent to induce diabetes in rats and mice [9][10][11][12]. STZ is produced by Streptomycetes sachromogenes and STZ causes to abnormal B-cell functions by imparing glucose oxidatation and decreasing insulin biosynthesis and secretion [13,14]. Szkudelski stated that STZ dose range is larger than alloxan and other agents and only one dose is enough to induce diabetes [15]. Decrease in plasma insulin levels in animal models after STZ application is used a sign for inducement of diabetes [16][17][18]. In spite of reporting significant differences in plasma insulin levels after STZ application, majority of the studies using STZ do not report practical importance or effect sizes of the differences. But there is a need to show practical importance for future decisions on dose and time of STZ application.
Based on this idea, the purpose of this study is to make a meta-analysis of the studies on the effect of STZ on plasma insulin levels in diabetic organisms.

Materials and Methods
In this study, meta-analysis approach was used to evaluate practical importance of the differences regarding plasma insulin levels of STZinduced diabetic rats and mice. Meta-analysis is different from a review including summarizing existent literature, since meta-analysis involves statistically synthesizing results of different studies [19,20]. For meta-analysis in this study, Cohen's d effect size values were calculated for 37 studies and mean effect size value was found for deciding about average effect size value as an indicator of mean practical importance of the differences in plasma insulin levels induced by STZ.

Selection of the Publications
In selection process of the publications PubMed, Google Scholar, Proquest and National Theses Database System were searched by using key words "Plasma insulin levels, STZ, Rats". The time restriction for the publications was [2005][2006][2007][2008][2009][2010][2011][2012][2013][2014][2015]. In National Theses Database System no thesis was found about the keywords it might be related to system error while Proquest search showed 89 theses. However, one thesis was found appropriate. When Pubmed was searched 526 results were found. As the highest publication number, Google scholar search results gave 3960 publications.
After adding the publications to the pool, checking abstracts and content of the publications were conducted. Eventually it was determined that 37 studies reported change in plasma insulin levels of diabetic organisms and they reported 39 differences for effect size calculations across different doses of STZ. Descriptive knowledge about the publications is represented in Table 1. The titles of them can be seen in appendix (Table 3).

Calculation of Effect Sizes and Analysis
In this study plasma insulin levels measured in control and STZ groups were considered for calculating effect size values. The effect size of differences regarding plasma insulin levels were accepted as an indicator of practical importance of the differences, therefore one Cohen d formula was used to calculate effect sizes [21,22]. After individual effect sizes per difference in each publication were calculated, mean effect size value was obtained by adding all effect sizes and dividing total effect size score into number of individual effect sizes. Hence just only one value regarding effect of STZ on plasma insulin levels was gathered.

Results
Results of the study showed that only 4 of the all individual effect sizes indicated negative values while the rest of effect sizes (n=35) was positive. Moreover one small and 38 large effect sizes were seen in the calculations. Descriptive values regarding Plasma Insulin Levels in control and STZ-induced diabetes groups, Unit of Plasma Insulin Levels and Individual Effect Sizes were shown in Table 2.
As seen in the Table 2, the individual effect sizes were between -13.7 to +65.3. The mean effect size value was found as +9.33.

Discussion
The results of this study made it clearer that STZ-application is an effective way of decreasing significantly plasma insulin levels of rats and mice.
Mean effect size value calculated from the publications showed that practical importance of STZinduced decrease in plasma insulin levels had a large effect. In other words effect size value of +9.33 refers to a large effect size [23]. Therefore the mean value of the STZ applied group is over 90 percentile of the no treatment group or control group.
The results of the study are in line with the findings of the current research studies using STZ for inducing diabetes in rats and mice [9,10]. Sai Varsha, Thiagarajan, Manikandan and Dhanasekaran applied STZ (35mg/kg) to Male albino Wistar rats, the authors observed plasma insulin decrease in rats after 72 hours [24].
The findings of this study contribute to our understandings about practical importance of differences in plasma insulin levels induced by STZ. When looked at the number of the publications in this study, it can be seen that decisions are based on differences in the publications over 35. Hence the findings of this study make our inferences about plasma insulin level differences induced by STZ more valid rather than relying on only one study's finding. At the same time findings of the study has a potential for informing researchers about dose and duration of STZ application to change plasma insulin levels of diabetic rats and mice. As another implication of this study, the publications analyzed in this study show characteristics of current practice about using STZ, therefore the effect sizes reported in this study also inform practice using STZ in diabetes studies.
In spite of strong sides of this study, it can be suggested that number of the publications using STZ might be increased in future studies to improving quality of inferences and to make the analysis more comprehensive. At the same time, other publications involving reports and unpublished documents should also be investigated for determining effect sizes regarding the differences about plasma insulin levels induced by STZ. Finally future studies might look at the studies published before 2005.