The present study was a hospital based, cross-sectional study conducted in the Department of Medicine, SP Medical College, Bikaner, a Tertiary Care Center in Northwest Rajasthan, India. A sample size of 947 was calculated based on a diabetic prevalence rate of 10.4% in India, according to IDF atlas 2019 [1] with a permissible error of 20% using the statistical formula, n=4pq/L2 (p=prevalence; q=100-prevalence; L=p×permissible error). In the present study, 960 type 2 diabetic patients aged ≥18 years attending Diabetes Center and Medicine Outpatient Department (OPD) were interviewed over one and a half year period from July 2018 to December 2019. Ethical clearance from the Institutional Review Board {No: F.29. (Acad) SPMC/2019/3868} and informed consent from each subject were taken.

Inclusion criteria: Patients with T2DM aged 18 years or more attending the Medicine OPD and diabetic clinic for follow-up or general treatment, with at least one year duration of diabetes (one year was to ensure that patients have attended at least one session of the diabetes education program) who were willing to participate in the study were included.

Exclusion criteria: Individuals with diabetes other than T2DM (type1 diabetes, pancreatic diabetes, secondary diabetes, gestational diabetes), age younger than 18 years, and those unable to answer the questionnaire because of deafness, dementia, or psychosis, those with diabetes-related complications (such as diabetic retinopathy, nephropathy, neuropathy, and diabetic foot, because they are more aware of diabetes and its complications) were excluded from this study.

This was a questionnaire-based cross-sectional study. A structured, bilingual questionnaire derived from a validated set of questionnaires were used. Each participant was interviewed face-to-face by the principal investigator and enough time was provided. The questionnaire was divided into six main categories, including, sociodemographic information, anthropometric measurement, diabetes history, KAP-related information.

Sociodemographic information included age, gender, marital status (married, unmarried, widowed/divorced), residential area (urban, rural), education (illiterate, primary, high school, graduate), occupation (unemployed, labourer, retired, homemaker, service, business), type of family (nuclear, joint) and socioeconomic status. Illiteracy and type of family were defined according to standard definition [9]. Residential areas were defined according to the census of India 2011 [10]. Modified BG Prasad scale was used to measure the socioeconomic status and to classify the status into five classes [11].

Anthropometric measurements including height, weight, and BMI were done using standard technique and appropriate formula. A BMI of 23 and 27.5 was taken as the cut-off for the overweight and obese, as defined by the World Health Organisation (WHO) for this population [12]. Diabetes history included duration of diabetes, family history of diabetes, and diabetic educational programs attended or not.

The questions regarding knowledge were derived from the Chennai Urban Rural Epidemiology Study [5]. A total of eight questions were included. The scoring was done as follows: (a) For closed questions, correct answers were graded as one and incorrect answers (inclusive of “don’t know”) as zero; (b) For multiple choice questions, score one for each correct response and zero for incorrect answers (inclusive of “don’t know”); (c) Thus, the least possible score was ‘0’ if all answers were incorrect and the maximum score was ‘24’ if all answers were correct; (d) A composite score in percentage was calculated by dividing each subject’s score by the maximum possible score; e.g., if an individual’s score was ‘15’, then the composite score would be 15/24×100=62.5%.

The questions regarding attitude and practice were derived from Asmelash et al., study [13]. Eight questions were included for each attitude and practice regarding diabetes. Each positive response was assigned a score of ‘1’ and each negative response with a score of ‘0’. For the eight attitudes and practice related questions the maximum attainable score was ‘8’ and the minimum score was ‘0’. The composite score in percentage was then derived. Poor KAP corresponded to a score of (<Mean±1SD); average KAP corresponded to a score between (Mean±1SD); good KAP corresponded to a score of (>Mean±1SD) [3].

Questions used for obtaining data regarding knowledge of diabetes were as follows [5]:

Questions evaluating attitude towards diabetes were as follows [13]:

Questions evaluating the practice of diabetes control and management were as follows [13]:

The data were analysed using Statistical Package for the Social Sciences (SPSS) version 16.0. Data were presented as mean, standard deviation, and proportion. ANOVA and t-test were used to test the equality of means between various groups. Pearson’s correlation coefficient was used to measure the strength of association between KAP. Multiple linear regression analysis was used to determine the variable associated with diabetes-related KAP. A p-value of less than 0.05 was considered statistically significant.

A total of 960 patients with Type 2 DM were included in the study. The mean age of study participants was 55.45±11.64 years. Among them, a male (57.5%) preponderance was observed. The majority of the respondents(60.83%) were urban inhabitants. The majority (80%) had some formal education with only 20% being illiterate. A higher proportion of the respondents (68.33%) lived in joint families. More than half (58.33%) of the respondents belonged to socioeconomic class II. The mean duration of diabetes was 5.73±2.94 years. More than half (54.17%) of the respondents had a duration of diabetes 5-10 years. The majority (79.17%) had a family history of T2DM. Forty-five percent of the respondents attended the diabetes education program once and 13.33% of the respondents never attended any diabetes education program. The mean BMI of the study participants was 25.66±5.64 and almost half (48.33%) were overweight and obese [Table/Fig-1].

The mean KAP scores of the respondents were 67.46±16.81, 63.75±15.52, and 49.17±15.29, respectively. Among the respondents, the level of knowledge was poor in 23.33%, average in 65.84%, and good in 10.83% of the respondents. The levels of the attitude of the study participants were found to be poor in 16.67%, average in 74.17%, and good in 9.16% respondents. The levels of practices were also described accordingly as poor in 12.5%, average in 78.33%, and good in 9.17% respondents.

The knowledge of diabetes was found to be better among the respondents aged less than 30 years (p=0.012), but attitude and practices were not significantly different among various age groups. Males had better knowledge compared to female counterparts (p=0.033), while females had a better attitude (p=0.008). Respondents living in the urban areas showed a better attitude and practices compared to those living in rural areas. In general, respondents from upper socioeconomic class, with higher educational background, with longer duration of diabetes, with a family history of diabetes, those who attended diabetes education program and with higher BMI demonstrated significantly greater scores in terms of KAP (p<0.05). Respondents with business as their occupation had better KAP scores (p<0.05) compared to other occupational groups. Respondents living in joint families had a higher attitude score while those living in nuclear families had higher practice scores (p<0.05) [Table/Fig-2]. On correlation analysis, knowledge was positively correlated with attitude (r=0.73, p=0.001) and practices (r=0.81, p=0.001): and attitude were positively correlated with practices (r=0.77, p=0.001).

Multiple linear regression analysis showed that the predictors that influenced KAP scores were age, gender, education, occupation, socioeconomic status, type of family, family history, diabetes duration, and educational program [Table/Fig-3].

The KAP regarding diabetes is greatly influenced by socioeconomic conditions, cultural beliefs, and habits. Understanding these variables help to design prevention and management strategies for diabetes [3]. Insufficient knowledge and awareness regarding complications of diabetes may lead to a high economic burden in terms of the management of complications [14]. In the present study, the respondents’ knowledge was assessed based on their understanding of diabetes, which included risk factors, symptoms, prevention, and complications options. The diabetes-related knowledge level was found to be average in 65.84%, good in 10.83%, and poor in 23.33% respondents, respectively. Similar findings were reported by Fatema K et al., Rahaman KS et al., and Farzana S et al., [3,14,15]. On the contrary, Khaznadar AA et al., and Tejaswi P et al., reported good KAP scores among patients with T2DM [16,17]. Whereas other studies have reported that knowledge about diabetes is generally poor among diabetic patients [18-20]. The majority of the participants had an average attitude (74.17 %) and practice (78.33%) score. There were a small number of patients, who were able to manage their health conditions to avoid further complications. These findings are in line with Islam FM et al., study [21]. Whereas, other authors (Tejaswi P et al., and Alsous M et al.,) reported a good attitude and practice score towards diabetes [17,22].

Better knowledge is associated with a better attitude (r=0.73, p=0.001) and practice (r=0.81, p=0.001) and better attitude is associated with better practice (r=0.77, p=0.001). This suggests that the higher their knowledge the better their attitude towards diabetes. These findings were consistent with other previous studies [3,17,22,23], while, Banu H et al., reported significant strong correlation between knowledge and attitude, significant weak correlation between knowledge, and practice and significant moderate correlation between attitude and practice [24].

In the current study, KAP was better among the younger age group, which could be explained by better internet accessibility by younger people in the present era. This finding is similar to other previous studies [16,25] and is contradictory to the study of Niroomand et al. in which KAP improved with age [26]. We observed a gender gap in knowledge and attitude regarding diabetes. There were inequalities in health services and education, with the literacy and access to healthcare facilities of females were lower than that of their male counterparts. Males showed significantly higher levels of knowledge, while females showed better attitude scores compared to males. This finding is similar to previous studies where a wide gender gap was observed in KAP regarding diabetes [3,15,20,21,27]. On the contrary, Murata G et al., showed no significant gender difference [28].

Subjects living in urban areas had better knowledge and attitude regarding diabetes. This finding is similar to previous studies [16,29]. This may be due to better education and easy access to healthcare facilities in urban areas. We observed that business persons had better KAP score than other occupational groups, which might be due to better access to healthcare facilities. This finding was contradictory with findings of Khaznadar AA et al., and Murugesan N et al., where professionals or those with executive jobs had higher scores [16,30]. It is important to note that subjects from obese and overweight groups had significantly better KAP scores (p<0.05) compared to the normal and underweight groups. This may be because obese and overweight subjects are more counselled by healthcare professionals because obesity is a significant cardiovascular risk factor. Similar findings were reported by Niroomand M et al., and Fatema K et al., [3,26]. This is contrary to a study done by Farzana S et al., [15].

Good KAP was seen among respondents with higher socioeconomic status. Similarly educational status had a significant association with good KAPs. This could be explained by the fact that subjects with higher socioeconomic status have better access to education and healthcare facilities and educated subjects are able to read necessary information easily compared to the illiterates. These findings were consistent with other studies [3,17,18,22,31]. Contrary to this, Khaznadar AA et al., found that mean practice scores were higher among illiterate subjects [16]. In the present study, the duration of diabetes was a significant predictor of KAP. This finding was consistent with the study done by Rahaman KS et al., and Niroomand et al., and contradictory to the study done by Khamseh ME et al., [14,26,32]. This may be due to regular counselling and contact with diabetes specialists.

Attending a diabetes education program was significantly associated with better KAP s in the present study. This finding was consistent with a study done by Rahaman KS et al., [14]. It is worth to mention that providing education to vulnerable groups could become a costeffective public health strategy and proper diabetes education programs could provide awareness about diabetes even for less educated people. A family history of diabetes was associated with better KAP, which are consistent with the findings of Rahaman KS et al., andNiroomand M et al., and contradictory with the findings of Balagopal P et al., [14,26,33]. Receiving information from the family members with diabetes might influence the patient’s KAP, yet, such information is not always reliable [26]. Subjects living in joint families had a higher attitude score whereas those living in the nuclear families had higher practice scores (p<0.05) but no difference in knowledge scores. This could be explained by the fact that subjects living in nuclear families have fewer responsibilities and hence enough time to take care of their diabetes while subjects living in a joint family have a better approach regarding diabetes due to information received from family members. No studies comparing the type of family and KAP regarding diabetes was found in literature.

Few limitations of this study need mention. Since, the current study was based on a single center, the results may not be truly representative of all diabetic patients in this region. The use of questionnaires for measuring KAP levels could be another limitation because of the risk of social desirability bias.

Poor KAP levels were found particularly in rural habitats, illiterates, unemployed, and low income groups, and they need greater attention in diabetes education and counselling. There is a need to carry out extensive awareness programs, after identifying the suitable measures to disseminate the message to the general public. It is possible to improve practice by providing adequate information, increasing the availability of educational materials, and proper guidance towards diabetes management. The information collected in this study will also be helpful in the dissemination of knowledge on the preventive aspects of diabetic complications. In conclusion, a good knowledge is associated with better attitude and practices. Consequently, active participation and empowerment of the people with diabetes by continuous education and support is mandatory using available mass media.