Bidirectional relationship between temporomandibular disorder and ankylosing spondylitis: a population-based cohort study

OBJECTIVES

This study aimed to determine the relation between temporomandibular disorder (TMD) and ankylosing spondylitis (AS) bidirectionally and ascertain the important comorbidities for AS occurrence in TMD patients.

MATERIALS AND METHODS

We conducted this population-based cohort study through Longitudinal Health Insurance Database, Taiwan. Study 1 investigated the risk of TMD in AS patients. Study 2 assessed the risk of AS in TMD patients.

RESULTS

In total, 3204 AS patients and 12,816 age-matched and gender-matched comparisons were enrolled in study 1. The TMD incidence in the AS cohort was 2.88-fold higher when compared with the comparisons (1.54 vs. 0.53 per 10,000 person-years). After adjusting for age, gender, and comorbidity, the AS cohort had a 2.66-fold (95% CI = 1.79-3.97) increased risk of TMD occurrence (P < 0.0001). The second study recruited 4998 TMD patients and 19,991 age-matched and gender-matched comparisons. Both TMD and comparison cohorts showed similar AS risk (HR = 1.49, 95% CI = 0.91-2.43, P = 0.1108) in the adjusted model. Study 2 identified a 3.66-fold increased risk of AS occurrence in TMD patients with comorbidity, including parapsoriasis, rheumatoid arthritis, osteoporosis, Cushing's syndrome, and climacteric arthritis (P < 0.012).

CONCLUSIONS

AS appears to significantly impact the occurrence of TMD. TMD might play a synergic role in AS development.

CLINICAL RELEVANCE

Clinicians have to be vigilant about the increased risk of TMD in AS patients and take care of AS disease activity and prognosis. The symptoms and signs of TMD could be a predictor of AS in patients with the aforementioned comorbidities.

Hyperuricemia precedes non-alcoholic fatty liver disease with abdominal obesity moderating this unidirectional relationship: Three longitudinal analyses

BACKGROUND AND AIMS

The temporal relationship between hyperuricemia and non-alcoholic fatty liver disease (NAFLD) is debatable. This study aimed to explore whether there exists a bidirectional or temporal relationship between them.

METHODS

A total of 11,585 participants were recruited from the Beijing Health Management Cohort during the period 2012-2016. We evaluated whether hyperuricemia was associated with NAFLD development (part I) and whether NAFLD was associated with hyperuricemia incidence (part II) using a logistic regression model. Further, the cross-lagged panel analysis model was used to simultaneously examine the bidirectional relationship between hepatic steatosis and serum uric acid (SUA) (part III). Subgroup and interaction analyses were also performed to assess whether other variables moderated those relationships.

RESULTS

In part I, multiple logistic regression indicated that baseline hyperuricemia was associated with the development of NAFLD (OR = 1.5970, p < 0.0001). In part II, multiple logistic regression showed that baseline NAFLD was not correlated with hyperuricemia incidence (OR = 0.8600, p = 0.1976). In part III, cross-lagged panel analyses indicated that the standard regression coefficient of baseline SUA to follow-up hepatic steatosis (0.1516) was significantly greater than the coefficient from the baseline hepatic steatosis to follow-up SUA (-0.0044) with p < 0.0001 for the difference. This indicated a unidirectional relationship from baseline SUA to follow-up hepatic steatosis, suggesting hyperuricemia may precede NAFLD; and this relationship was not affected by age, sex, dyslipidemia, metabolism syndrome, diabetes but was moderated by abdominal obesity.

CONCLUSIONS

This study demonstrated a unidirectional relationship from hyperuricemia to NAFLD incidence, and suggested that lowering SUA levels in hyperuricemia patients may prevent subsequent NAFLD.