Introduction

Type 2 diabetes mellitus (T2DM) and osteoporosis (OP) are two major metabolic diseases commonly seen in elderly patients. As we all know, osteoporosis is a metabolic disease that is affected by age, hormones, and other factors leading to decreased bone mass and impaired bone microstructure, resulting in increased bone fragility and even fractures. Osteoporosis and the ensuing fragility fractures are a huge economic burden for many countries [1]. Previous studies have identified many risk factors for osteoporosis and therapeutic drugs have been developed, such as bisphosphonates, teriparatide, and linagliptin [2,3,4]. In fact, only a small proportion of patients have been treated [5]. A cross-sectional study in mainland China reported that T2DM affects more than 74.22 million people in China [6]. A persistent hyperglycemic state is positively associated with osteoporosis risk [7]. Patients with long disease duration may be at higher risk of falls and fractures due to factors such as peripheral neuropathy, visual impairment, and cardiovascular disease [8]. Therefore, the management of osteoporosis and T2DM is of increasing interest. Poor nutritional status is an important risk factor for the development of T2DM and osteoporosis. A study by SHANGGUAN et al. illustrated that patients with osteoporosis in the presence of nutritional deficiencies have a higher risk of all-cause mortality [31, 32] and endocrine-metabolic diseases [31] as well as fractures [33].

Malnutrition is a risk factor for various complications and poor disease prognosis and is closely related to a patient's quality of life [34, 35]. Early assessment of nutritional status is important for diagnosing nutritional deficiencies and improving the prognosis of patients. The results of the ROC curve analysis showed that the GNRI combined with age and albumin had the best ability to predict osteoporosis compared with the PNI and CONUT scores. The GNRI, which is calculated by the combination of height, weight, and serum albumin [11], has received a lot of attention from researchers due to its ease of operation and less influence by subjective factors. The GNRI has received attention from researchers because it is easy to perform and less affected by subjective factors, and is useful for the prediction of prognosis and mortality in many diseases [36, 37].

In our study, HDL was negatively correlated with BMD at all sites (P = 0.002, P = 0.024, P = 0.002, respectively), and total cholesterol was negatively correlated with femoral neck BMD (P = 0.024) and lumbar spine BMD (P = 0.019). The results of studies on the relationship between serum lipids and BMD are currently controversial. Some studies have reported a negative correlation between HDL-C and BMD [38, 39]. Some studies have found a positive correlation between HDL-C and lumbar spine BMD and total hip BMD [40], and that HDL-C is a protective factor for OP in patients with T2DM [41]. However, some studies also reported no relationship [42]. We hypothesize that this discrepancy may be due to racial and gender differences in sample size.

It has been demonstrated that smoking and alcohol consumption affect bone metabolism and damage bone microarchitecture, although these two factors were not significant in this study. Tobacco contains high levels of tar and nicotine metabolites, and damage to bone microarchitecture is evident. Li et al. found that cadmium in smoke was shown to interfere with bone metabolism directly or indirectly in both animal and human experiments [43]. In contrast, smokers showed significantly higher levels of bone formation markers, such as osteocalcin and uncarboxylated osteocalcin, within 124 days of quitting smoking [44]. The effect of alcohol consumption on BMD is related to the amount of alcohol consumed, with moderate alcohol consumption slowing bone loss, while chronic excessive alcohol consumption may damage bone microstructure and increase fracture risk [45, 46]. An animal study showed that even without altering BMD, alcohol consumption altered the morphology and percentage of collagen in the trabeculae of the femoral neck in rats, increasing bone fragility [47].

Conclusions

In conclusion, our study found that GNRI correlated with femoral neck BMD, and lumbar spine BMD, and did not correlate with total hip BMD after controlling for age as a confounding factor. Lower GNRI was associated with the development of osteoporosis in elderly type 2 diabetic patients. GNRI combined with age and albumin had better predictive ability for osteoporosis than PNI and CONUT scores. GNRI is more suitable for nutritional assessment in elderly type 2 diabetic patients.

Limitations

There are some limitations of this study. First, it was a single-center retrospective study with an unbalanced gender distribution, which may have some bias; second, it could not prove a causal relationship between GNRI and osteoporosis in elderly patients with T2DM. Third, factors such as dietary habits, exercise intensity, and length of time receiving sunlight, which may affect bone density, were not considered.