The reviewed studies confirmed the hypothesis of the authors. The findings of the reviewed studies showed GI can have a slight effect on inflammation. However, the complication of GI and the lack of sufficient studies with the same findings prevent us from providing a firm answer to this question. In this review, as any study had its own considerations, the authors provided a possible link between GI and inflammation as well as described the main limitations that are better to be considered during GI studies.
In 2010, findings from a multicenter diet intervention reported no significant differences between groups for CRP before and after adjustment in this study . However, more investigations within the study showed two completely different effects of GI on CRP, one positive and one negative, with two different diet patterns. The findings of this study made this hypothesis that diet patterns, especially fat values, can affect the effect of GI .
The impact of diet patterns on GI and inflammation was repeated several times in other studies. According to the reports of a cohort study, diet patterns with higher GI scores had a slightly higher reported level of TNF-a and CRP than lower GI groups . Another finding that could drive out from this study was that GI might significantly change in different diet patterns. In this study, diet patterns with sweets and desserts group had a lower GI than refined grains and breakfast cereal group which were believed to have a higher GI . This can be a significant limitation for GI and GL studies. According to the current study, considering the populations' diet pattern and the consumption of unhealthy foods with higher inflammatory index and lower GI as a confounder is recommended.
However, the main limitation of GI and inflammatory studies can be related to their small sample size along with short follow-up. The majority of studies fell in this category [31, 38]. Nevertheless, all three studies that had a well-controlled low biased risk protocol showed a significant direct relation between GI and inflammation despite a small sample size [33, 36, 39]. In all of these studies, all the participants followed a close diet pattern during the assessment that showed the priority of confounders over the sample size [33, 36, 39]. However, another study shown providing linear graphs for these small sample-sized studies can provide good information .
In another Study with a considerable population, a 137-item Food Frequency Questionnaire (FFQ) was used to determine the dietary intake and GI . In this study, which was based on the Brand-Miller GI table , a significant relation was found between IL-6 (P-ANOVA = 0.050) and TNF (P-ANOVA = 0.046) with GI at the baseline . However, after the one-year intervention, no significant differences between GI with both IL-6 and TNF resulted . The main reason for this difference can be explained by the method of this study which used low inflammatory diet patterns in the groups [41–43]. In this study, two possible anti-inflammatory patterns were compared to each other that could impact the finding.
However one of the most surprising findings was in the Bahado-Singh et al. study . This study has shown a 38.24% decrease in HS-CRP level in the low-intermediate GI group in comparison with a 15.18% decrease in the low-intermediate GI group . Despite the decrease in both groups, the decrease in the low-intermediate GI group was significantly lower than in the high GI group (p < 0.05), but there was no description of the reason for the anti-inflammatory effect of both high and low-intermediate GI diets in this study. However, despite both groups having the same diet during the assessment, this decrease can be affected by other anti-inflammatory components of both diets. Nevertheless, low GI had a higher anti-inflammatory impact.
Despite supporting data on the effect of GI on different diseases [7, 9, 10, 15–21, 44, 45], findings on inflammatory biomarkers are vastly ranged. A meta-analyzed study shows a significant difference between low and high GI groups in CRP for both models in obese people with and without diabetes . A meta-analysis study by Milajerdi et al. rejected any inflammatory effect for GI that supported the Buyken et.al study [26, 47]. This is while another study showed GI is associated with oxidative stress .
Nevertheless, the most important finding regarding this relation was reported in Yeon-soo et al. study in 2018 . In this study, an association had been found between GI and Dietary inflammatory Index (DII), which was developed by Dr. Shivapa and Dr. Hebert [50–53] to determine the dietary inflammatory potentials. This study and the effect of GI on CRP, which was reported by Schwingshackl et al. , suggests a reconsideration of the effect of GI on inflammation which was reported by Milajerdi et al.  recently.
Overall, the main weakness of GI is related to its nature which increases the bias risk (Fig. 2). Food processing, sugar-containing, other nutrients, food PH, speed of eating, blood glucose level, and insulin level are some factors that can affect the GI response of the body [12–14, 54–56]. The other critical weakness of GI is related to its pattern in food classification [12–14]. In this pattern, some pro-inflammatory foods like pizza (GI = 39), fructose (GI = 15), chocolate (GI = 40), ice cream (GI = 51), Soft drink/soda (GI = 59), potato crisps (GI = 56) have been classified as low-moderate GI foods while some fruits like pineapple (GI = 59), mango (51), watermelon (GI = 76) have higher GI [12–14]. Considering these issues, it may be possible that an unhealthy diet pattern has a lower GI than a healthy one; but research is needed to confirm this hypothesis.
Finally, by considering limitations, running a well-designed GI study has several complications that must be addressed first. Of all food components, it seems lipids have the most cofounding bias but still require more investigation. Continued follow-ups as a controlled clinical trial can be helpful too. In addition, a study to understand the association of diet patterns with GI is recommended. To understand the effect of GI on inflammation, using more homogenized populations as well as removing the effect of differences in diet patterns that can make biases has extra benefits. However, all in all, a dietary pattern high in fruit, vegetables, fish, poultry and legumes, whole grains, and low in red and processed meat, sweetened beverages, sweets, refined grains, and fried potatoes has been linked to a lower level of inflammatory biomarkers [3, 5, 6, 27].
The strength of this study is related to the point of view of the reviewers. In all steps, at least two researchers with two different opinions made the review of studies which provides a new point of view and hypotheses for further studies. Nevertheless, the main weaknesses of this study were related to the nature of GI and the lack of sufficient studies. The other limitation of this review is its close time to previous systematic studies, which is uncommon. The other limitation was the lack of any statistical analysis. Nevertheless, the authors understood current findings in GI are not suitable and homogenized for this purpose. As the effect of dietary patterns in the reviewed studies is unclear, any analysis can lead to a wrong but maybe statically specific answer on this subject.