Relationship between physical activity and bone mineral density loss after gastrectomy in gastric cancer patients

The prevention of osteoporosis is a particularly relevant issue for gastric cancer survivors. We investigated the relationship between postoperative physical activity and the change of bone mineral density (BMD) in patients with gastric cancer. Patients who underwent radical gastrectomy for gastric cancer were enrolled in this single-center prospective cohort study. Physical activity was evaluated using the International Physical Activity Questionnaire Short Form at postoperative month (POM) 6 and patients were classified into high, middle, and low physical activity groups accordingly. The primary outcome was the change in BMD from baseline at POM 12, which was expressed as a percentage of the young adult mean (YAM). The YAM of the lumbar spine and femoral neck was measured by dual-energy X-ray absorptiometry. One hundred ten patients were enrolled in this study. The physical activity level at POM 6 was classified as high (n = 50; 45%), middle (n = 25; 23%), and low (n = 35; 32%). The mean decrease of YAM% was 5.1% in the lumbar spine and 4.2% in the femoral neck at POM 12. A multivariable-adjusted logistic regression model revealed that low physical activity at POM 6 was a significant risk factor for BMD loss at POM 12 (odds ratio, 3.76; 95% confidence interval, 1.48–9.55; p = 0.005). Low physical activity after gastrectomy is an independent risk factor for decreased BMD at POM 12. The introduction of exercise may prevent osteoporosis after the surgical treatment of gastric cancer.


Introduction
Radical gastrectomy with lymph-node dissection has been a pivotal treatment for gastric adenocarcinoma [1,2]. Patients who undergo gastrectomy will suffer from various postoperative sequelae, including abdominal and/or general symptoms, for a long time [3,4]. They also have a reduced digestive function and a reduced ability to absorb nutrients, resulting in deterioration of their nutritional status. As a result, a previous study reported that gastrectomy may trigger sarcopenia [5], anemia [6,7], or osteoporosis [8,9] in gastric cancer patients. Bone mineral density (BMD) loss, a precursor status of osteoporosis, occurs earlier in the postoperative period among these patients and eventually causes bone fractures, which may significantly decrease the postoperative quality of life (QOL) [5,6].
Maintenance of bone density requires nutrients that are necessary for bone metabolism as well as adequate loading on bone. In gastric cancer survivors, the absorption of nutrients related to bone metabolism (e.g., calcium and vitamin D) may be deteriorated, and physical activity may be decreased, both of which may contribute to BMD loss after surgery. Indeed, recent evidence suggests that exercise is effective in the prevention and treatment of osteoporosis [10][11][12]. Decreasing physical activity after gastrectomy is considered to be an important risk factor for BMD loss; however, no study has examined the association of physical activity with BMD loss among patients after gastrectomy.
We investigated the relationship between physical activity at postoperative month (POM) 6 and BMD loss at POM 12 in patients who underwent radical gastrectomy for gastric cancer. We hypothesized that decreased physical activity after gastrectomy would be an independent risk for BMD loss. The prevention of BMD loss is a particularly important issue for physicians who treat gastric cancer [13,14]. Our results would be useful information for them to understand the long-term effect of postoperative continuous exercise on the prevention of osteoporosis.

Study design and cohort development
This was a single-center, prospective cohort study. We have been conducting a prospective and observational study to assess physical activity in postoperative gastric cancer patients since September 2017. Patients with histologically proven gastric adenocarcinoma, who were scheduled to undergo radical gastrectomy including distal gastrectomy, total gastrectomy, proximal gastrectomy, or pylorus-preserving gastrectomy from September 2017 to August 2020 were eligible for inclusion in the study. Patients who had metastatic lesions, other simultaneously malignant diseases, or other severe comorbidities were excluded from this cohort. In addition, we excluded patients who received R1/2 resection and who had missing follow-up data at 6 or 12 months after gastrectomy.

Measurement of physical activity
Physical activity was assessed by a physiotherapist before surgery, and at POM 6 and 12. They used the International Physical Activity Questionnaire Short Form (IPAQ-SF), which asks about three specific types of activity that were undertaken during the previous 7 days in the four domains (leisure time, work, household activities, and transport) [15][16][17]. The items in the IPAQ-SF were structured to provide separate scores for walking, moderate-intensity, and vigorous-intensity activities. Computation of the total score for the short form requires summation of the duration (in minutes) and frequency (days) of walking, moderateintensity, and vigorous-intensity activities. According to guideline of the IPAQ scoring protocol [15,18], we classified patients' physical activity into three categories: high, moderate, or low activity. In brief, high activity is defined by satisfying the meet following two criteria: (a) vigorousintensity activity on at least 3 days achieving a minimum total physical activity of at least 1500 MET-minutes/week or (b) 7 or more days of any combination of walking, moderateintensity, or vigorous-intensity activities achieving a minimum total physical activity of at least 3000 MET-minutes/ week. Moderate activity is defined by fulfilling the following three criteria: (a) 3 or more days of vigorous-intensity activity of at least 20 min per day, (b) 5 or more days of moderate-intensity activity and/or walking of at least 30 min per day, or (c) 5 or more days of any combination of walking, moderate-intensity, or vigorous intensity activities achieving a minimum total physical activity of at least 600 METminutes/week. Low activity was defined as not fulfilling the criteria for moderate or high activity.

Measurement of BMD
The primary outcome was the percentage of change in BMD at POM 12 from baseline (before surgery). The BMD of the lumbar spine (L2-L4) and femoral neck was measured by dual-energy X-ray absorptiometry (DEXA). DEXA is the gold standard method for the measurement of BMD [19]. BMD was expressed as a percentage of the young adult mean (YAM). We calculated the percentage change in YAM (Delta YAM%) using the following equation: Delta YAM% = (YAM after postoperative year 1 − baseline YAM)/ baseline YAM × 100 (%). Considering to previous cohort studies, the definition of severe BMD loss was defined as a Delta YAM% of 8% or more in femoral neck or lumbar spine [9,20,21].

Statistical analyses
Patient characteristics are reported as descriptive statistics, with continuous variables expressed as the median and interquartile range (IQR) or the mean and standard deviation, and categorical variables expressed as the number and percentage. Continuous variables were compared using Student's t test, and categorical variables were compared using Fisher's exact test.
We used multivariable logistic regression models to identify risk factors for severe BMD loss, which was defined as a ≥ 8% decrease of YAM in the femoral neck or lumbar spine, with adjustment for confounding factors. We calculated the adjusted odds ratio (OR) of physical activity at POM 6.
All statistical tests were 2-sided, and p values < 0.05 were considered statistically significant. All statistical analyses were performed using R (version 4.0.3; https:// www.R-proje ct. org/; Lucent Technologies, Vienna, Austria). Figure 1 shows the patient flow diagram. A total of 224 patients were registered in this cohort database. After excluding ineligible patients, 110 patients were included in the analysis. Table 1 shows the patient characteristics. The physical activity level at POM 6 was classified as follows: high, n = 50 (45%); middle, n = 25 (23%); and low, n = 35 (32%). Table 2 and Fig. 2 show the delta YAM% for each physical activity level. Among the included patients, the mean Delta YAM% was 5.1% in the lumbar spine and 4.2% in the femoral neck at POM 12. Forty-two patients (38.2%) had severe BMD loss. Table 3 shows the multivariable logistic regression analyses of factors associated with severe BMD loss at 1 year after gastrectomy. In comparison to middle or high physical activity at POM 6, low physical activity was significantly associated with severe BMD loss, with a multivariable-adjusted OR of 3.76 (95% confidence interval [CI], 1.48-9.55; p = 0.005).

Discussion
The present study showed two important findings. First, at 1 year after gastrectomy, the patient's BMD decreased 5.1% in the lumbar spine and 4.2% in the femoral neck, in terms of YAM. Second, low physical activity at POM 6 was an independent risk factor for decreased BMD at POM 12.
Previous studies have discussed the association between gastrectomy and decreased BMD in terms of malnutrition after gastrectomy and hormonal disorder [8,9,13,14,22]. Some reports have focused on secondary hyperparathyroidism due to malabsorption of vitamin D and calcium as a cause of decreased BMD after gastrectomy [21,23], and others have focused on silicon, which is abundant in bones, skin, and other connective tissues [24][25][26]. Silicon is an important mineral required for bone formation, thus showing the importance of gastric exocrine and endocrine functions in the maintenance of silicon homeostasis. Indeed, it was mentioned that a decreased serum silicon level was  [26]. In addition, it has been reported that adjuvant chemotherapy and total gastrectomy might adversely affect the nutritional status, especially in patients after total gastrectomy, which in turn reduces BMD [8]. This evidence coincides with our result of marginally nonsignificant associations between severe BMD loss and lower serum albumin and total gastrectomy.
On the other hand, the importance of exercise for maintaining bone density and preventing osteoporosis has recently been reported. Although the exact mechanism of bone formation by physical activity is not yet fully  understood, Turner and Duncan reported that it is thought to induce bone anabolism and homeostasis via mechanotransduction [27,28]. Other mechanisms are thought to involve physical activity eliciting anabolic vascular effects, and the improvement of local tissue metabolism, which thereby promotes bone mineralization [12]. Although the pathophysiological mechanisms have not been completely elucidated, it has been reported that physical activity induces bone anabolism and homeostasis via mechanotransduction, or that vascular anabolism improves local tissue metabolism and promotes bone mineralization [12,27,28]. Our present results support these reports, showing that physical activity during the first 6 months after gastrectomy might be able to prevent BMD loss at the 1 year postoperatively. Few reports have prospectively measured the BMD before and after gastrectomy and analyzed the association between physical activity level and BMD loss; thus, the present results are relevant information for physicians who treat gastric cancer patients.
The present study was associated with some limitations as it was conducted at a single center with a relatively small sample size. First, there was a subject selection bias. Second, the small sample size limited the confounding factors that could be included in the multivariate model. However, the 95% CI for physical activity at POM 6 was well above 1, and the pivotal variables could be adjusted; thus, the causality seemed to be robust. Furthermore, we did not able to analyze the serum level of metabolite from vitamin D (25-hydroxy vitamin D) as a potential factor affecting bone metabolism. This was an exploratory and observational study, and the effects of exercise in the prevention of osteoporosis after gastrectomy should be investigated in a future intervention study. One strength of the present study is that information on physical activity and BMD loss was prospectively collected, which is difficult to manage in daily clinical practice.
In conclusion, a low physical activity at POM 6 was an independent risk factor for decreased BMD at POM 12. The  Table 3 Multivariable logistic regression analysis to assess the association of postoperative physical activity with severe BMD loss a The multivariable logistic regression model included age (≥ 75 years old or < 75 years old), BMI at POM 6 (< 18.5 kg/m 2 or ≥ 18.5 kg/m 2 ), serum albumin at POM 6 ( ≤ 3.5 g/dl or > 3.5 g/dl), menopausal status (menopause or non-menopause female/male), surgical procedure (total gastrectomy or other procedures), and physical activity before surgery and at POM 6 (low or middle/high) BMD, bone mineral density; BMI, body mass index; CI, confidence interval; POM, postoperative months