Abstract: Background and Objective: Osteoarthritis (OA) is the most common progressive joint disease considered as an inflammatory disease in the elderly and one of the main causes of physical disabilities. Limited studies have assessed the relationship between dietary patterns and knee OA. The Objective of the study was to investigate the relationship between major dietary patterns and different degrees of knee Osteoarthritis (OA) based on Kellgren and Lawrence (K and L) score in a group of Tehranian women. Materials and Methods: In this cross-sectional study, data of 231 patients was diagnosed with different grades of knee OA (age: 52.59±7.69 years). Also, anthropometric measurements and blood samples were collected in order to determine fasting blood sugar (FBS), lipid profile, hs-CRP, ALT and AST. Physical activity and food intakes information were gathered by physical activity and a semi-quantitative food frequency questionnaire (SFFQ) with 147 items. Finally, three major dietary patterns were derived from Principle Component Analysis. Results: Three major dietary patterns were extracted, "Western", "Traditional" and "Healthy". Western dietary pattern (WDP) contains high intake of junk foods, sweets-dessert-high calorie beverages, processed food, coffee and high fat dairy and low intake of fruits and vegetables. Traditional dietary pattern (TDP) characterized by high intake of red meats, organ meats, legumes, onion and potato, olive, eggs and nuts. Also, healthy dietary pattern (HDP) contains high intake of liquid oil, low fat dairy, unrefined grains, white meats and tea. Overall, total variance of the three dietary patterns accounted for 25.8%. In comparison to all patients in the lower tertile, women in the 3rd tertile of WDP were younger and less physically active. Lower physical activity, higher basal metabolic rates and fat free mass variables were observed in the 3rd tertile of the HDP (p<0.05). Among 61.5% of participants who were postmenopausal women (after stratifying based on menopausal status), participants in the higher tertile of TDP predisposed to severe degrees of knee OA (OR: 2.178, 95% CI 1.084-4.374). Conclusion: It is concluded that postmenopausal women with higher adherence to Iranian TDP in comparison to lower tertile of this dietary pattern may prone to knee OA more than two fold, due to inflammatory food groups.
INTRODUCTION
Osteoarthritis (OA) is the most common degenerative joint disease in the elderly and one of the main causes of musculoskeletal disorders, functional disabilities and morbidity1. Most common symptoms are pain, stiffness, restriction of motion and loss of function2. One of the most prevalent kind of osteoarthritis is knee OA3 which increases by getting older4. In a study on adult population in Iran, knee OA was more prevalent in rural areas of Iran than other countries5. Women are more likely to develop severe knee OA than men6. In higher degrees of knee OA, just a total joint replacement can decrease the pain, however, most of patients with knee OA are not the suitable candidate for joint replacements, since the knee operation is extremely invasive as well as expensive7. The NSAIDs have shown efficacy for OA pain management8, but have significant side effects such as gastrointestinal disorders9. Therefore, appropriate control of the environmental risk factors of OA is important to improve the management of OA10.
Several environmental factors have been associated with OA, such as nutrition and smoking. Some studies have established that the increased risk of diet-related chronic diseases is a result of inadequate nutrition11. The tendency to assess the relationship between diet and disease has changed from single foods and nutrients into the dietary patterns12. This tendency also considers synergistic effects of multiple components within the diet13. On the other hand, it is a very challenging task to accurately assess the effect of single nutrients on OA, since combination of foods are consumed14.
As OA has been categorized as an inflammatory disease, it is important to focus on the anti-inflammatory diet15. A few studies have assessed the relationship between dietary patterns and knee OA16,17. Present study, was aimed to assess the association between dietary patterns and the grades of knee OA in a Tehranian women group.
MATERIALS AND METHODS
Study population: Between November, 2015 and February, 2016, 600 women with chief complain knee pain enrolled from health house centers of Tehran Municipality. Of whom 350 women were ineligible, 250 women were accepted to attend the study. Participants were referred to the clinic of endocrinology and metabolism research Institute (EMRI) of Tehran University of Medical Sciences. The study was approved by the Ethics Committee of the EMRI (IR.Tums.EMRI.REC.1395.00130). A comprehensive questionnaire was used and informed consent was obtained from all individual participants included in the study.
Participants were selected based on the following inclusion criteria: (1) Women aged 38-70 years old with knee pain, (2) No history of diseases such as diabetes mellitus, cardiovascular disease, kidney disease, liver disease, multiple sclerosis, rheumatoid arthritis, hypothyroidism or other malignant diseases, (3) No history of other joint injuries due to trauma, accidents or exercise, a vascular necrosis, inflammatory arthropathy and infection around knee, (4) No history of joint surgery (5) No history of corticosteroid therapy during one month before entering the study and (6) No history of smoking or alcohol consumption during the past year.
Knee OA diagnosis: The OA was diagnosed based on radiological results, clinical examination or subjective sings. The Kellgren and Lawrence (K and L) score was used to detect and grade radiographic knee OA18. Based on radiographic evidences, The grades of severity were determined from 0-419.
All women underwent weight-bearing bilateral anteroposterior radiography of the knees. The knee examinations were performed by an expert rheumatologist. All radiographs were assessed by a rheumatologist and an orthopedist using the K and L radiographic features. All the radiographic interpretation was by consensus agreement. Results of radiographs categorized into the following five groups based on the K-L grade including 0 = Absence of OA, 1 = Suspected OA, 2 = Minimal OA, 3 = Moderate and 4 = Severe joint OA19. A participant was recruited for the study if radiographic knee OA showed at least one of her knee joint was graded K and L 1 or above. Finally, data obtained from 231 patients with knee OA were analyzed since nineteen women had no sign of knee OA (grade = 0). Also, women who suffered from moderate and severe knee OA merged together as grade 3.
Anthropometric measurements: Weights and heights were measured in order to calculate body mass index (kg m–2). The body composition of all women was assessed using body composition analyzer (Body Impedance Analysis = BIA) (BC-418MA-Tanita, UK) by following the manufacturer’s instructions. The device calculated fat mass and fat-free mass.
Blood sampling and biochemical parameters: Blood samples were obtained after 10-12 h of overnight fasting. All baseline blood samples were obtained between 8:00 and 10:00 am. After centrifuging, the serum was isolated and stored at a temperature of -80°C. Assessments were as follows.
Fasting blood sugar (FBS) levels were measured by a colorimetric method based on the glucose oxidase phenol 4-aminoantipyrine peroxidase (GOD/PAP) method. Triglyceride measurements were performed using the enzyme glycerol-3-phosphate oxidase Phenol 4-aminoantipyrine peroxidase (GPO-PAP) method. Direct low-density lipoprotein (LDL) and high-density lipoprotein (HDL) cholesterol were determined by the enzymatic endpoint method and enzymatic clearance assay, respectively. Randox Laboratories kits (Randox Laboratories Ltd., Ardmore, UK) were used for all evaluations. Serum high-sensitivity C-reactive protein (hs-CRP), a pro-inflammatory marker, was evaluated using a high-sensitivity immunoturbidimetric assay (Hitachi 902 analyzer, Hitachi Ltd., Tokyo, Japan). Serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were determined by an automatic analysis system (Autoanalyzer, Hitachi Ltd., Tokyo, Japan) with a Randox laboratories kit.
Semi-quantitative food frequency questionnaire: A Semi-quantitative food frequency questionnaire (SFFQ) with 147 items validated for Iranian population20 was used to assess food intake. A trained dietitian interviewed the participants and completed all of the questionnaires. Then participants were requested to respond questions about the amount of the food during the past year. Based on similarities between the nutrients, The 147 items of the SFFQ which were converted into a daily intake were categorized into 22 pre-defined food groups21.
Major dietary pattern extraction: Since OA is considered an inflammatory disease, 22 food groups were formed based on the anti-inflammatory diet21 by using Principle Component Analysis (PCA).The anti-inflammatory diet contains fresh foods, fruits and vegetables except onions and potatoes which have the alkaloid solanine, high in -3 PUFAs, whole grains, lean protein sources and spices and low in saturated fat, -6 fats, refined carbohydrates, eggs, red meat, butter and other high-fat dairy products and refined and processed foods21. We used the absolute weight in grams of each of the 22 food groups to analyze these groups using the PCA22. Daily food intakes in grams derived for each person from the 147 item SFFQ. The factor analysis method in SPSS (version 16, SPSS IL, Chicago) were used. In this method, non-correlated factors were derived by varimax option in SPSS. Since pre-selection of the number of factors is necessary for the PCA and other pattern determining methods, consisting of food-groups components with eigenvalue 1.5 and also both the scree plot and the factors themselves to determine the most meaningful set of factors demonstrating separate dietary patterns. Each factor consisted of food items with absolute factor loading greater than or equal 0.3. Factor analyses resulted in three major dietary patterns and factor loading matrices of the patterns are presented in Table 1. Positive loading demonstrated strong associations between the food groups and dietary patterns, while negative loading demonstrated negative associations. By these analyses, a three factor was selected. The three major dietary patterns were identified by using factor analyses are shown in (Table 1). Next, factor loadings were obtained among three factors (dietary patterns) and the factor score of these three dietary patterns was calculated for each individual. We named three dietary patterns (Western, Traditional, Healthy) based on the foods had loaded most on the factor and according to how the factors correlated with nutrients. Factor 1 was the most dominant dietary pattern in the population and mainly loaded of junk foods, sweets and desserts and high-calorie beverages, processed foods, high-fat dairy, coffee and low loaded on fruits and vegetables and was named as western dietary pattern (WDP), this dietary pattern explained 9.4% of variance in intake. Factor 2 consisted of high loading of red meats, organ meats, legumes, onion and potato, olives, eggs and nuts and was designated the traditional dietary pattern (TDP) and explained 8.6% of the variance in intake. Factor 3 included negative loading of solid oil and condiments and positive loading of vegetable oil, low-fat dairy, unrefined (whole) grains, white meats and tea and was designated healthy dietary pattern (HDP) and explained 7.7% of the variance in intake.
Statistical analysis: All analyses were performed using the Statistical Package for the Social Sciences (SPSS) for windows (version 16, SPSS IL, Chicago). The significance level was considered as p<0.05. Normal distribution of quantitative variables was checked by Kolmogorov-Smirnov test. Descriptive characteristics of participants were reported as means±SD for quantitative variables. In addition, One-way ANOVA test was used to compare the presence of any significant difference between groups for variables and for dietary pattern tertiles. To adjust the confounder’s effect, ANCOVA test was performed. Qualitative variables were evaluated by K-Square test. Each dietary pattern was categorized into tertiles and the first tertile was selected as the reference group. To find any relationship between dietary patterns and different grades of knee OA, we used the multinomial regression analysis including crude model and adjusted model (adjusted for confounders including age, BMI, physical activity and energy intake). In order to neutralize the effect of menopause period, we stratified data based on menopausal status (positive and negative) and then re-calculated crude model and adjusted model.
| Table 1: | Factor-loading matrix for the 3 major food patterns identified by PCA form food-frequency questionnaires |
RESULTS
Basal characteristics of participants: Finally, 231 patients were recruited for the study (age: 52.59±7.69 years) (Table 2). According to OA classification, 48% of women were suspected to knee OA, 33% were suffered from mild and 18% were diagnosed with moderate and severe knee OA.
Basal features based on dietary patterns: Characteristics of women across tertiles of each dietary pattern score are listed in Table 3. The women in the 3rd tertile of WDP were younger than those in the 1st tertile (age: 50.54±7.98). In addition, physical activity was higher in the top tertile compared to the lower tertile in the WDP. These differences were significant after adjustments for confounders including age, BMI, total energy intake and physical activity. In the TDP, women in the 3rd tertile had a significantly higher basal metabolic rate (BMR), total energy intake, fat mass, fat free mass, right leg fat free mass and left leg fat free mass than those in the 1st tertile. Among these variables, only the total energy intake remained significant after adjustments for all mentioned confounders. In addition, more adherence to TDP resulted in lower serum triglyceride levels. There were no significant associations between dietary patterns and other biochemical factors. Also, participants in the top tertile of the HDP had higher physical activity (Mets). After adjustments, the results for physical activity was significant as well. However, BMR and fat free mass variables showed significant results in which the 3rd tertile has increased in value compared to the first tertile (p>0.05).
Relationships between major dietary patterns and knee osteoarthritis classification: Analyses were calculated after stratification based on menopause status. Menopausal women who aged 57.15±5.62 years old were older than premenopausal women (45.32±4.14). In pre menopause, there was no association between dietary patterns and severity of knee OA. More adherences to the WDP showed an association with knee OA. Finally, this association was not significant after adjusting for confounders. In TDP, postmenopausal women (menopause positive) in the higher tertile were more likely to suffer from severe knee OA (OR: 2.178, 95% CI 1.084-4.374) (Table 4).
| Table 2: | Baseline characteristics and biochemical metabolic of women with mild, moderate and sever knee OA |
Numerical variables were reported as the mean±standard deviation and categorical variables were presented as number (percentage). Obesity was classified based on BMI > 30 kg m–2. OA: Osteoarthritis, BMI: Body mass index, BMR: Basal metabolic rate, FFM: Fat free mass, RL.FFM: Right leg fat free mass, LL. FFM: Left leg fat free mass, HDL: High-density lipoprotein, LDL: Low-density lipoprotein, TG: Triglyceride, FBS: Fasting blood sugar, AST: Aspartate aminotransferase, ALT: Alanine aminotransferase, hs-CRP: High sensitive C-reactive protein. N = 231, among whom 142 women of all participants are postmenopausal (61.5%) | |
| Table 3: | Characteristics of the study participants by tertiles of dietary pattern score |
p-value less than 0.05, N = 231, †p-value resulted from ANOVA test, §General linear model (ANCOVA) used for adjusting confounders including age, BMI, Physical Activity. a,b,cThe same letter considered as significant extracted by LSD procedure in POST-HOC analysis. *N = 142 (among 231 participants, 142 women were menopause). Numerical variables were reported as the mean±standard deviation and categorical variables were presented as number (percentage). Obesity was classified based on BMI>30 kg m–2. OA: Osteoarthritis, BMI: Body mass index, BMR: Basal metabolic rate, FFM: Fat free mass, RL.FFM: Right leg fat free mass, LL.FFM: Left leg fat free mass, HDL: High-density lipoprotein, LDL: Low-density lipoprotein, TG: Triglyceride, FBS: Fasting blood sugar, AST: Aspartate aminotransferase, ALT: Alanine aminotransferase, hs-CRP: High sensitive C-reactive protein |
|
| Table 4: | Multivariate non-adjusted and adjusted odds ratios and 95% confidence interval for knee OA across tertiles of dietary pattern scores by menopause status |
| aAdjusted for age, BMI, physical activity, total energy intake, *p-value<0.05 | |
DISCUSSION
In the current cross-sectional study, evidences were found to suggest that Tehranian postmenopausal women who are more adherent to TDP were at risk of higher grade of knee OA even after adjusting for confounders.
Women in higher tertile of TDP had low amounts of simple carbohydrates. It has proved that excess intake of carbohydrates results in accumulation of triglyceride in adipose tissue, it is assume that significant reduction of TG serum levels is due to lower intake of carbohydrates in TDP. Moreover, the intake of legumes led to lower serum levels of triglyceride since legumes contain fibers. The mechanisms by which dietary fiber lowers triglyceride serum levels is unknown23.
It was examined that Adherence to HDP is associated with fat-free mass. Patients following HDP consumed more protein resources including more low-fat dairy and white meats which through increasing the muscle mass can increase the fat free mass. Studies reported that the adequate protein intake can prevent developing negative nitrogen balance, therefore, adequate protein intake plays an inevitable role in maintaining the muscle mass and energy expenditure24.
Results of current study have shown that the risk of severe knee OA in postmenopausal women who adhered to TDP doubled after adjusting for confounders. The TDP mainly contains red meats and organ meats which consist of saturated fatty acids, arachidonic acid and animal protein. Food fatty acids content has a crucial role in inflammation. A study reported that there is a significant positive association between OA and food fatty acid content25. An animal study on Mouse model have shown that lower dietary intake of fat improves tissue repair and reduces the severity of OA after joint injuries26. The ω-3 to ω-6 ratio is critical in preventing inflammation. In TDP, the ratio of ω-3 to ω-6 is low and is associated with the development of higher grades of OA27.
The ω-3 PUFAs improved the wound damages, however, SFAs (Saturated Fatty Acids) or ω-6 PUFAs (Polyunsaturated Fatty Acids) independently propelled the OA into its severe grades25. Similar results indicated that there is an association between radiographic images and SFA16. In addition, TDP is high in arachidonic acid which will be converted to pro-inflammatory prostaglandins and leukotrienes. Animal foods and certain types of vegetable oils are food sources of arachidonic acid28. Firstly, arachidonic acid and saturated fatty acid which are found in red meats and organ meats contribute to inflammatory pathway and knee OA29. Secondly, prostaglandins as a member of lipid autacoids can be synthesized from arachidonic acid. Both prostaglandins and arachidonic acid regulate homeostatic functions and pathogenic mechanisms, such as the inflammatory response. Over the inflammatory reaction, prostaglandins undergo serious alterations. Before the activation of leukocytes and the infiltration of immune cells in acute inflammation, prostaglandin production will increase dramatically30. Animal proteins can increase the permeability of small bowel which leads to bacterial translocation, resulting in immune complex development in the blood stream. These large immune complexes may reside in small capillaries which lead to inflammation and injuries over the long time. It is believed that this phenomenon accounts for worsening many autoimmune and chronic inflammatory diseases, such as arthritis28. A study indicated that consumption of red meat in high amounts is related to higher circulating levels of hs-CRP (high sensitivity C-reactive protein) as an inflammatory factor31. Also, another study showed that higher consumption of red meat was related to higher risk of developing inflammation32. Results of a study revealed that humanized mice fed bioactive Neu5Gc which is high in red meat and subsequently treated with Neu5Gc antibodies developed systemic inflammation33. It has been demonstrated that the estrogen deficiency at the time of menopause boosts a woman’s risk of developing OA34. A review study35 revealed that there is a high prevalence of OA among postmenopausal women, related to the action of estrogen receptors (ERs) in joint tissues and proposed a relationship between OA and loss of ovarian function. In vitro studies36 have suggested that the direct chondro-protective role of estrogen may result from the synthesis of glycosaminoglycans, which are pivotal components of connective tissue. In addition, estrogen suppresses cyclooxygenase-2 mRNA expression both in bovine articular chondrocytes and other tissues, associated with protection against ROS-induced chondrocyte damage37.
Iran is considered as a country with high prevalence of knee osteoarthritis. Comprehending the role of diet should be one of the health care’s priorities in populations with increasing risk of chronic diseases and without appropriate medical treatment. The results of this study can be applicable in countries with similar characteristics of extracted dietary patterns to Iranian TDP. It is recommend more longitudinal research in this area in order to confirm current findings and explore potential pathophysiological mechanisms.
CONCLUSION
This mentioned study has concluded that Following a Western diet may increase the future risk of knee OA. Diet has an inevitable role in protecting the health of joints and reduction of OA incidence or progression..This study can be beneficial for women with knee OA.
SIGNIFICANCE STATEMENT
This study discovers the relationship between Iranian traditional dietary pattern and increased risk of knee OA. Dietary intake is examined through dietary pattern which is valuable compared to other forms of evaluating such as each individual components, dietary habits and so on. Moreover, the pattern that we have found relationship with knee osteoarthritis is not examined in other studies. It would be beneficial for women who follow Iranian traditional dietary pattern who live in both Iran and the other countries. This study will help the researcher to uncover the critical areas related to the different dietary patterns which is detrimental to the knee OA symptoms. Thus a new theory on the relationship between dietary patterns and different diseases may be arrived at.
ACKNOWLEDGMENTS
This cross-sectional study was financially supported by Tehran University of Medical Sciences (Funding’s ID: 94-03-30336). We thank the Osteoporosis Research Center, Endocrinology and Metabolism Clinical Science Institute, Tehran University of Medical Sciences, Tehran, Iran for allowing us to conduct this study.