Are serum lipids involved in primary frozen shoulder? A case-control study.

BACKGROUND

Hyperlipidemia is a proposed, but unproven, risk factor for primary frozen shoulder. The purpose of this study was to evaluate the association between serum lipid profiles and primary frozen shoulder.

METHODS

This was a case-control study. The case group comprised 300 patients diagnosed with frozen shoulder from October 2009 to April 2013. Patients with diabetes, thyroid disease, or previous shoulder surgery or trauma were excluded. The control group comprised 900 age and sex-matched persons with normal shoulder function who visited our health promotion center for general check-ups during the same period. We calculated the odds ratios and 95% confidence intervals to identify any association between serum lipid level and primary frozen shoulder, using conditional logistic regression analysis. We evaluated continuous data on the serum levels of total cholesterol, calculated low-density lipoprotein, measured low-density lipoprotein, high-density lipoprotein, triglyceride, and non-high-density lipoprotein cholesterol. We also evaluated categorical data on hyper-cholesterolemia, hyper-low-density lipoproteinemia (calculated and measured), hyper-high-density lipoproteinemia, hyper-triglyceridemia, and hyper-non-high-density lipoprotein cholesterolemia.

RESULTS

Univariate analysis of the continuous data showed total cholesterol (odds ratio, 1.010 [95% confidence interval, 1.006 to 1.013]; p < 0.001), calculated low-density lipoprotein (odds ratio, 1.008 [95% confidence interval, 1.004 to 1.012]; p < 0.001), measured low-density lipoprotein (odds ratio, 1.007 [95% confidence interval, 1.003 to 1.011]; p = 0.001), high-density lipoprotein (odds ratio, 1.015 [95% confidence interval, 1.006 to 1.024]; p = 0.001), and non-high-density lipoprotein cholesterol (odds ratio, 1.007 [95% confidence interval, 1.004 to 1.011]; p < 0.001) to be significantly associated with primary frozen shoulder. Univariate analysis of categorical values showed hyper-cholesterolemia (odds ratio, 1.789 [95% confidence interval, 1.366 to 2.343]; p < 0.001), calculated hyper-low-density lipoproteinemia (odds ratio, 1.609 [95% confidence interval, 1.210 to 2.138]; p = 0.001), measured hyper-low-density lipoproteinemia (odds ratio, 1.643 [95% confidence interval, 1.190 to 2.269]; p = 0.003), hyper-high-density lipoproteinemia (odds ratio, 1.440 [95% confidence interval, 1.062 to 1.953]; p = 0.019), and hyper-non-high-density lipoprotein cholesterolemia (odds ratio, 1.645 [95% confidence interval, 1.259 to 2.151]; p < 0.001) to be significantly associated with primary frozen shoulder.

CONCLUSIONS

We conclude that hypercholesterolemia and inflammatory lipoproteinemias, particularly hyper-low-density lipoproteinemia and hyper-non-high-density lipoprotein cholesterolemia, have a significant association with primary frozen shoulder. Further research is needed to evaluate whether a non-optimal serum lipid level is a cause, a related co-factor, or a result of primary frozen shoulder.