VitD3 deficiency appears to be a widespread phenomenon worldwide currently. Numerous studies have shown that VitD3 deficiency leads to considerable metabolic changes [24–26]. Total cholesterol (TC), low-density lipoprotein (LDL), and high-density lipoprotein (HDL) changed, and HOMA-IR (an indicator of insulin resistance) elevated as the level of VitD3 decreased [27–29]. A recent study mentioned levels of VitD3 and its metabolites decreased, with a corresponding decrease in the activity of enzymes in pancreatic beta cells and insulin-responsive cells. This suggests that VitD3 may be a potential regulator of insulin secretion and a maintenance factor of pancreatic β-cell activity.
Previous studies have analyzed the relationship between VitD3 and insulin resistance in patients with pre-existing diabetes, which is flawed because the onset of type 2 diabetes is almost irreversible. In previous studies, insulin resistance has been confirmed as the pathogenesis of type 2 diabetes [11, 30, 31]. Therefore, in order to prevent and control the increase of type 2 diabetes, it is necessary to investigate the relationship between insulin resistance and VitD3 in the general population. Our findings suggest that VitD3 and insulin resistance are negatively associated after adjusting for potential confounders. The risk of insulin resistance decreased with the increase of VitD3 across the whole VitD3 level.
In general population, education, income, race, BMI, and waist circumference were all associated with the development of insulin resistance. For example, whites and other races were less likely to develop insulin resistance than Mexican Americans. The odds ratios for the insulin resistance were 0.42 (0.3, 0.58) in whites and 0.29 (0.18, 0.45) for other races compared with Mexicans. Waist circumference, BMI, and alcohol consumption were important factors in the development of insulin resistance, with each unit increase in waist circumference and BMI increasing the incidence of insulin resistance by 9% and 22%, respectively, in the unadjusted model. Compared with non-drinkers, drinkers were 26 percent more likely to develop insulin resistance. A previous animal study found that VitD3 inhibited the secretion of inflammatory factors that contribute to insulin resistance, such as interleukin-6 and tumor necrosis factor-α, that contribute to insulin resistance by activating the Jun n-terminal kinase 1 (JNK1) and IKK-β/NF-κB pathways to phosphorylate insulin receptor-1 and attenuate insulin signaling, and that VitD3 supplementation can effectively inhibit the physiological functions of interleukin-6 and tumor necrosis factor-α, thus preventing the development of insulin resistance[32, 33]. In this study, we observed consistent result in the general U.S. population, with a greater negative association between VitD3 and insulin resistance, particularly in the general U.S. population with a BMI of 24–28. According to data analysis, each 10-unit increase in VitD3 in this population was associated with a 20% reduction in the risk of insulin resistance, with an odds ratio and 95% CI of 0.8 (0.69, 0.93).
In previous studies, BMI, waist circumference, HDL, and TG were important insulin resistance factors [34, 35]. Subgroup studies have shown a stable stratified relationship between serum VitD3 and IR, and there was an interaction between VitD3 and insulin resistance at different BMI levels. The association was highest at the range of BMI 24–28. The exact mechanism needs to be elucidated by further studies.
Previous studies have investigated possible factors in the development of IR. For example, the relationship between age and IR development has been well documented . In previous studies, smoking may be a risk factor for developing IR and inhibiting insulin sensitivity [36–40]. However, smoking and insulin resistance were not statistically significant in this study, which may have to do with our population composition. We found that the non-smoking population was too sparse to reflect populational differences.
There are several limitations to this study. Firstly, the composition of the previously mentioned population, because our data came from publicly available database and were not self-selected, excluded other populations that were not eligible for inclusion in the study. For example, there were not enough non-smoking samples, which made it hard to analyze whether smoking is a risk factor for IR. Secondly, this study excluded specific populations, including unhealthy populations such as hyperlipidemia, hyperglycemia, and hyperuricemia. Hence, our study is not representative for all the populations. Third, we did not investigate the risk of IR occurrence into a genetic level. Furthermore, the use of HOMA-IR to define insulin resistance is somewhat inaccurate and difficult to replicate.
Despite the limitations of the current study, the advantages of this study remain to be highlighted. Firstly, our study is different from previous studies. Previous studies mainly focused on the insulin resistance in diabetes patients, while in this study, IR was investigated in general population with more suitability. Secondly, this study explored the five-cycle database of NHANES. NHANES is based on complex design sampling. Finally, we discovered for the first time that BMI interacts with VitD3 in the occurrence of IR which has not been mentioned previously.