Introduction
Early-onset Parkinson’s disease (EOPD) refers to patients with Parkinson’s disease (PD) having age at onset (AAO) less than or equal to 45 years but onset up to age of 50 years is included by some authors. Reference Mehanna, Smilowska and Fleisher1 EOPD comprises of about 3%–6% of all cases of PD. Reference Willis, Schootman, Kung, Racette and Willis2 Although EOPD shares many common characteristics with late-onset PD (LOPD), several features appear to cluster in earlier onset presentations, conferring a phenotypic homogeneity to early-onset cases.
EOPD patients may experience a poorer health related quality of life (QoL) than older onset counterparts due to psychosocial consequences and comorbid depression. A study comparing QoL in early onset and late onset PD found that EOPD has worse overall QoL scores, independent of presence of depression. Reference Knipe, Wickremaratchi, Wyatt-Haines, Morris and Ben-Shlomo3 Another study from Iran corroborated this finding of significantly worse depression and the “emotional” domain score of QoL in the EOPD cohort. Reference Fereshtehnejad, Hadizadeh, Farhadi, Shahidi, Delbari and Lökk4 EOPD patients with Parkin mutations have a worse QoL than the non-genetic EOPD patients, with significant contribution from non-motor symptoms (NMS) such as depression and excessive daytime sleepiness. Reference Zhou, Liu and Chen5 These findings indicate a need to systematically study the NMS that adversely impacts the QoL in patients with EOPD. Few studies from India that have examined NMS and its relation to QoL reveals an almost 100% prevalence of NMS, commonly fatigue, pain, anxiety, and urinary symptoms. Reference Kumar, Patil and Singh6,Reference Kukkle, Goyal and Geetha7,Reference de Souza, Kakode, D.’Costa and Bhonsle8 Initial studies on NMS in the Indian population did not find any difference in NMS with respect to age or age at onset. Another study found one or more NMS in all the patients and all the individual NMS domains were affecting the QoL. Reference Kumar, Patil and Singh6 Yet another study found that fatigue, lightheadedness and pain were the most prevalent NMS, with the total NMS score being the most important determinant of QoL. Reference Karri, Ramasamy and Kalidoss9 However, the EOPD population was not specifically examined in these studies. Hence, we aimed to assess the NMS in patients with EOPD and compare with healthy controls, using standardized scales, and to determine how this influenced the quality-of-life metrics.
Methods
This cross-sectional observational study was conducted in the department of Neurology at the National Institute of Mental Health and Neuro Sciences (NIMHANS), Bengaluru, India which is a tertiary care center in South India. The study was approved by the Institute Ethics Committee (IEC NO.NIMH/DO/IEC (BS & NS DIV) 2018–19). EOPD patients diagnosed using UKPDS Brain bank criteria Reference Marsili, Rizzo and Colosimo10 with age at onset (AAO) less than or equal to 45 years were included. Although a cut-off of 50 years is now recommended by the International MDS Task Force, previous studies have variably used an upper limit of 40-50 years to define EOPD. Reference Mehanna, Smilowska and Fleisher1 Clinical characterization, motor and non-motor assessments, and QoL assessment were performed using validated scales. Furthermore, 60 healthy age and gender-matched controls were also included.
Motor Assessments
Motor assessment was performed using the unified Parkinson’s disease rating scale (UPDRS) part-III (overnight levodopa OFF and supramaximal levodopa dose ON state) and Hoehn and Yahr staging. We calculated the total levodopa equivalent daily dose (T-LEDD) using the formula given by Tomlinson et al. Reference Tomlinson, Stowe, Patel, Rick, Gray and Clarke11 Hence, the T-LEDD is equal to (Levodopa × 1) + (Levodopa CR × 0.75) + (Pramipexole × 100) + (Ropinirole × 20) + (Amantadine × 1) + (Rasagiline × 100) + (Selegiline × 10) + (Levodopa X 0.33 if entacapone is used irrespective of dose) + (Levodopa × 0.5 if tolcapone is used irrespective of the dose). Reference Tomlinson, Stowe, Patel, Rick, Gray and Clarke11 None of the patients were on any newer medications such as safinamide, opicapone, and istradefylline, the LEDD of which cannot be calculated using the above formula.
The EOPD patients were categorized into three clinical phenotypes as tremor dominant (PD-TD), postural instability, and gait difficulty (PD-PIGD) and mixed type. This categorization was based on the formula, which is a “ratio of mean tremor score (sum of items 20 and 21 in UPDRS part III “OFF score” divided by 4) to the mean bradykinesia/rigidity score (sum of items 22–27 and 31 in UPDRS part III “OFF score” divided by 15).” Patients with a ratio more than 1.0 were classified into PD-TD, and ratio less than 0.8 were classified into PD-PIGD variant. Patients with mixed phenotype had ratio between 0.8 and 1.0. Reference Schiess, Zheng, Soukup, Bonnen and Nauta12
NMS Assessments
The NMS scales that were used included Epworth Sleepiness scale (ESS), Reference Walker, Sunderram, Zhang, en Lu and Scharf13,Reference Arnulf14 Non motor symptom scale (NMSS), Reference Storch, Schneider and Klingelhöfer15 NMS-Quest, Reference Romenets, Wolfson and Galatas16 Impulse control disorders questionnaire (QUIP scale), Reference Martinez-Martin, Rodriguez-Blazquez and Catalan17 Hospital anxiety and depression rating scales (HADS), Reference Zigmond and Snaith18 REM sleep behavioral disorder screening questionnaire (RBDSQ), Reference Stiasny-Kolster, Mayer, Schäfer, Möller, Heinzel-Gutenbrunner and Oertel19 Pittsburgh sleep quality index (PSQI), Reference Buysse, Reynolds, Monk, Berman and Kupfer20 Montreal cognitive Assessment (MoCA), Reference Nasreddine, Phillips and Bédirian21 and the Parkinson’s Disease Questionnaire −39 (PDQ 39). Reference Jenkinson, Fitzpatrick, Peto, Greenhall and Hyman22 All these scales were applied to all the study participants. Permission to use MDS owned scales was obtained for the purpose of the study.
The non-motor scales were applied as follows:
-
a. The PDQ-39 Summary Index (PDQSI) was used as the quality-of-life metric, calculated by dividing the sum of subdomain scores by eight. Reference Jenkinson, Fitzpatrick, Peto, Greenhall and Hyman22
-
b. A score of above 10 on ESS was considered to be a significant marker for excessive daytime sleepiness. Reference Arnulf14
-
c. Scores of 6 or more on the 13 item RBDSQ were considered to be significant and suggestive of the presence of RBD. Reference Stiasny-Kolster, Mayer, Schäfer, Möller, Heinzel-Gutenbrunner and Oertel19
-
d. Poor quality of sleep was defined as the global score of > 5 on the six-domain self-rated PSQI. Reference Buysse, Reynolds, Monk, Berman and Kupfer20
-
e. A cutoff of more than 10 was used for the diagnosis of clinically definite anxiety or depression on the HADS. A score between 8 and 10 was used to define “borderline anxiety” and “borderline depression.” Reference Zigmond and Snaith18
-
f. A score of less than 26 was used to define cognitive impairment on the MoCA. Reference Nasreddine, Phillips and Bédirian21
-
g. QUIP, which is specifically devised for use in PD, was used to assess impulse control disorders (ICDs) with very high sensitivity. Reference Martinez-Martin, Rodriguez-Blazquez and Catalan17
-
h. NMS-Quest was used as the patient-rated questionnaire for the assessment of nine non-motor domains with a maximum score of 30. Reference Romenets, Wolfson and Galatas16
-
i. NMSS for PD was used as the observer-rated scale to assess the burden of NMS which was rated as follows - 0 (no burden), 1–20 (mild burden), 21–40 (moderate burden), 41–70 (severe burden), > 70 (very severe burden). Reference Storch, Schneider and Klingelhöfer15
Statistical Analysis
The data collected were tabulated in Microsoft Excel spreadsheets and analyzed using SPSS version 16. Comparison was performed between NMS in EOPD vs NMS in healthy controls. Correlation was examined between different groups of NMS and QoL measures. The mean and standard deviation were calculated for continuous variables and expressed categorical variables as frequencies and percentages. All the variables were tested for normal distribution using the Shapiro–Wilks test. The Mann–Whitney U test was employed for analysis of continuous independent variables not following normal distribution, and the independent t-test was employed for variables that following normal distribution. For comparison between the groups, the Kruskal–Wallis test was employed. The analysis of categorical variables was done by Pearson Chi-Square test. The strength of the association was tested between the two continuous variables with Pearson’s Correlation Coefficient or Spearman rank correlation depending upon the normality of the data. A p-value of ≤ 0.05 was considered as significant.
Results
Demographics and Clinical Characteristics
A total of 124 patients and 60 controls were included in the study. The mean age of the patients was 43.98 ± 7.47 years, and the mean AAO was 37.33 ± 6.36 years. The mean age of controls was 43.25 ± 13.58. There was a male preponderance among the cases (M:F = 82:42) and healthy controls (M:F = 36:24). Majority of the patients had AAO between 40-45 years of age (45.96 %) and 30–39 years of age (45.96%). The mean duration of illness for EOPD patients was 6.62 ± 5.30 years. A positive family history was found in 15 patients (12.09 %). The mean T-LEDD was 589.73 ± 307.12 mg/day (range 60–1600 mg/day). Levodopa was prescribed for 82.25% patients. A dopaminergic agonist (DA) was prescribed in 66.12% patients, which comprised of Pramipexole (78.04%) and Ropinirole (19.96%). A DA was started as the first drug in 66.12% patients, whereas Levodopa was started as the first drug in 33.88% patients. Peak dose dyskinesias were seen in 26.62% patients.
Motor Assessment
The UPDRS part III ‘OFF’ score was 35.33 ± 13.30. whereas the UPDRS part III ‘ON’ score was 14.27 ± 9.06. The mean percentage improvement in UPDRS part III score was 59.84 ± 17.55 %. Majority of the patients (86.29%) had PD-TD, 5.64% had PD- PIGD, and 8.06% had mixed motor phenotype.
Assessment of Non-Motor Symptoms
A high proportion of patients (61.29%) were found to have poor sleep quality on the PSQI assessments. The total percentage of EOPD patients having excessive day time sleepiness was 17.74%. RBD was present in 20.26 % of the patients. On the PD-NMS questionnaire, 31.4% (n = 39) patients reported one of the symptoms pertaining to RBD. Evidence of depression on screening with HADS was present in 36.8%, whereas anxiety was present in 44%. One or more impulse control disorder (ICD) was present in 16.12%, of which dopamine dysregulation syndrome (DDS) was the most common (60%). Evidence of cognitive impairment was found in 29.6% patients.
Gastrointestinal symptoms including dysphagia and/or constipation were reported by 40.9% patients, while a large number of patients reported urinary dysfunction (73.6%). About one-fourth (25.6%) of EOPD patients reported sexual dysfunction. Other significant NMS detected included body pain (68%), anosmia (22.4%), involuntary weight gain/loss (14.67%), and excessive sweating (13.63%). Results of non-motor assessments are summarized in Table 1.
DDS = dopamine dysregulation syndrome; EOPD = early onset Parkinson’s disease; ESS = Epworth Sleepiness scale; HADS– hospital anxiety and depression scale; MoCA = Montreal cognitive assessment; NMS = Quest–Non motor symptom questionnaire; NMS Scale for PD–Non–motor scale for PD, PDQSI = Parkinson Disease QoL Summary Index; PSQI = Pittsburgh sleep quality index; RBDSQ = REM behavioral disorder screening questionnaire.
Subgroup analyses:
-
A. Age-group differences: Differences in NMS were examined between patients with AAO between<40 years and those with AAO 40 years or above. There was no significant difference between the total NMSS burden (p = 0.12), Global PSQI scores (p = 0.07), ESS score (p = 0.09), QUIP scores (p = 0.29), RBDSQ scores (p = 0.12), HADS-depression score (p = 0.91), and HADS-anxiety score (p = 0.88). However, the mean PDQ-39 Summary index score was significantly higher in patients who had AAO 40 years or more compared to those with AAO below 40 years (p = 0.02).
-
B. Comparison between early and late PD: Patients with a duration of illness greater than five years (late PD) had a significantly higher UPDRS part III “OFF” score (40.008 vs 31.148, p = 0.0002). However, the LEDD values were comparable between the two groups (p = 0.81). No difference was observed between the two groups in the ESS scores (p = 0.76), total NMS burden (p = 0.76), MoCA score (p = 0.10), PSQI score (p = 0.39), or QoL summary index scores (p = 0.75). However, the overall NMS burden correlated with longer duration of illness (p = 0.04).
-
C. Effect of use of dopamine agonist (DA) on NMS: There was no significant difference in EDS (p = 0.50), PSQI score (p = 0.21), QUIP score (p = 0.33), HADS – depression score (p = 0.73), HADS-anxiety (p = 0.85), or total NMS burden (p = 0.90).
Comparison of NMS Between Cases and Healthy Controls
Other than excessive day time sleepiness, all other non-motor scores were significantly worse in cases compared to healthy controls, and the MOCA scores were significantly better in healthy controls compared to cases (Table 2 and Figures 1 and 2).
EOPD = early onset Parkinson’s disease; ESS = Epworth Sleepiness scale; HADS– hospital anxiety and depression scale; MoCA = Montreal cognitive assessment; NMS = Quest–Non motor symptom questionnaire; NMS Scale for PD–Non–motor scale for PD, PDQSI = Parkinson Disease QoL Summary Index; PSQI = Pittsburgh sleep quality index; RBDSQ = REM behavioral disorder screening questionnaire.
Correlation Studies
-
A. Correlation between NMS and clinical characteristics : A significant positive correlation was seen between the values of ESS, PSQI (Global score), HADS (Anxiety), and QUIP score with UPDRS (OFF) score and a significant negative correlation between MoCA score and the UPDRS (OFF) score. NMSS burden correlated with longer duration of illness. A significant positive correlation was found between age at assessment and poor sleep quality, RBDSQ scores, and total NMS burden, whereas a negative correlation was found between MoCA scores and age at assessment (Table 3).
-
B. Correlation of NMS with QoL of EOPD patients: QoL was found to be dependent on multiple NMS. A positive correlation was seen between the summary index scores and ESS score, NMSS (total) score, QUIP score, and NMS Quest score (Table 4). Correlation was examined between NMS subdomains and QoL metrics. A significant positive correlation was found between PDQSI and the domains of mood/cognition, sleep/fatigue, urinary dysfunction, and “miscellaneous domains” of weight change and excessive sweating (Table 4).
AAO = age at onset; DDS = dopamine dysregulation syndrome; EOPD = early onset Parkinson’s disease; ESS = Epworth Sleepiness scale; HADS– Hospital anxiety and depression scale; MoCA = Montreal cognitive assessment; NMS = Quest–Non motor symptom questionnaire; NMS Scale for PD–Non–motor scale for PD, PDQSI = Parkinson Disease QoL Summary Index; PSQI = Pittsburgh sleep quality index; RBDSQ = REM behavioral disorder screening questionnaire.
DDS = dopamine dysregulation syndrome; EOPD = Early onset Parkinson’s disease; ESS = Epworth Sleepiness scale; HADS– hospital anxiety and depression scale; MoCA = Montreal cognitive assessment; NMS = Quest–Non motor symptom questionnaire; NMS Scale for PD–Non–motor scale for PD, PDQSI = Parkinson Disease QoL Summary Index; PSQI = Pittsburgh sleep quality index; RBDSQ = REM behavioral disorder screening questionnaire.
Quality of Life Subdomains
Eight QoL subdomains scores based on the PDQ-39 were calculated. The subdomain scores were as follows: “ADL” = 38.99 ± 22.64, ‘Mobility’= 43.34 ± 23.17, “Emotional” = 41.16 ± 23.77, “Stigma”= 41.11 ± 31.52, “Social support” = 8.63 ± 17.00, “Cognition” = 21.21 ± 15.61, “Communication” = 22.85 ± 18.49, and “Bodily discomfort” = 19.03 ± 15.92.
Discussion
This study assessed the spectrum of NMS in patients with EOPD and their influence on the QoL measures using validated scales.
A number of salient demographic features of our cohort are worth a mention. These include a significant male preponderance, which may reflect gender inequalities in accessing medical care, in addition to biological factors. Being a tertiary care center, patients with relatively advanced stage of the illness were seen, as indicated by the high mean duration of illness (6.62 ± 5.30 years), T-LEDD (555.60 ± 296.32 mg/day), and high UPDRS III scores (35.33 ± 13.30). We also observed PD-TD as the major phenotype in our cohort.
A positive family history in first-degree relatives was observed in as many as 12.09% of our patients, consistent with previous studies. Reference Schrag and Schott23,Reference Mehanna, Moore, Hou, Sarwar and Lai24 Detection of certain genetic mutations in EOPD can have prognostic value and provides clue toward the expected non-motor profile. Prominent NMS features including anxiety, depression, and dementia are seen in PARK-DJ1(PARK7), while anxiety, ICD, and apathy are common in PARK-PINK1(PARK6). Reference Liu and Le25 Parkin mutations (PARK2), which are the most commonly implicated mutation in sporadic and familial EOPD across the globe, are known to present with less frequent cognitive impairment but may have ICDs more often. Reference Liu and Le25 With expanding knowledge of monogenic forms of PD, it is likely that their NMS signatures will eventually be elucidated more completely.
We applied a combination of multiple validated NMS scales that enabled comprehensive assessment of individual NMS domains. Our study showed that a high proportion of EOPD patients who had a poor sleep quality, with EDS in 17.74% and RBD in 20.26% patients, comparable to previous studies that have reported prevalence range of 15%–26%. Reference Gjerstad, Boeve, Wentzel-Larsen, Aarsland and Larsen26 Mood-related symptoms (anxiety and depression) were also comparable to previously reported estimates of 30%–48%. Reference Ray and Agarwal27 It is likely that sleep dysfunction, anxiety, and depression in EOPD are interrelated and has a multifactorial causation, with contribution from both biological and psychosocial factors. Other studies support the observation of higher incidence of anxiety in EOPD as compared to LOPD patients. Reference Rai, Goyal and Kumar28 Anxiety may be a feature of “non-motor wearing off” phenomenon that is experienced by some patients in later stages of the disease, which is supported by our finding of a significant correlation between UPDRS part III OFF scores and HADS (anxiety) scores. There was a significant correlation between the anxiety scores and the UPDRS part III OFF scores that suggest, patients with more severe motor phenotype are more susceptible to anxiety. In contrast to the large study from India, we did not find any difference between the anxiety and depression scores and the age at disease onset. Reference Kukkle, Goyal and Geetha7 Our study showed a slightly higher prevalence of cognitive impairment of around 29%, compared to previous estimates of 10%–20% in EOPD, Reference Chaudhary, Joshi, Pathak, Mishra, Chaurasia and Gupta29,Reference Tang, Huang and Nie30 possibly reflecting a more advanced stage of the disease. It is also likely that MoCA may have overestimated the scores on cognitive impairment in the present study with English being the second language in many of our patients.
The prevalence of ICD (16 %) was consistent with previous studies which have reported a variable prevalence range of 3%–40%. Reference Lim, Tan and Ngam31 The ICD prevalence is known to be higher in EOPD patients than late-onset PD patients. Reference Vela, Martínez Castrillo and García Ruiz32 This may be related to the use of higher use of DA in EOPD, as in our cohort, compared to LOPD. However, the subgroup analysis failed to show correlation between DA use and occurrence of ICDs. Our cohort also featured several other features known to be associated with development of ICD including male preponderance and higher T-LEDD. Reference Liu and Le25 The most common ICD in our EOPD cohort was dopamine dysregulation syndrome, which is consistent with previous studies. Reference Vela, Martínez Castrillo and García Ruiz32
The incidence of gastrointestinal dysfunction, urinary dysfunction, anosmia, and autonomic dysfunction is similar to that of other studies in EOPD. Reference Mehanna, Moore, Hou, Sarwar and Lai24,Reference Špica, Pekmezović, Svetel and Kostić33 It is hypothesized that urinary dysfunction may be a component of a distinct non-motor sub-phenotype characterized by autonomic dysfunction, RBD, and depression and may be a marker of a more severe PD phenotype. Reference Horsager, Knudsen and Sommerauer34 It is likely that genetic subtypes featuring these NMS such as GBA, SNCA, and VPS-related EOPD may have been present in our cohort, but lack of genetic data at present precludes further correlation. The overall NMS burden was comparable to findings from a previous EOPD study. Reference Špica, Pekmezović, Svetel and Kostić33
EOPD patients had significantly higher NMS burden as compared to age-matched healthy controls. This corroborates with other studies that have demonstrated more frequent and severe NMS in patients with PD compared to healthy aging individuals.
Few studies have specifically looked at NMS and QoL determinants in EOPD in the Indian context. The first main publication on YOPD in India examined mainly motor features, and some NMS such as cognitive impairment and autonomic dysfunction. Reference Muthane, Swamy, Satishchandra, Subhash, Rao and Subbakrishna35 More recently, a number of studies have examined NMS in PD in the last decade and have commonly reported pain, fatigue, urinary symptoms, anxiety, and depression. Reference Kumar, Patil and Singh6,Reference Karri, Ramasamy and Kalidoss9 Depression was found to be one of the main determinants of QoL in these studies. However, these studies did not specifically examine the EOPD cohort. The pattern of NMS seen in our EOPD cohort is similar to these studies reported in literature previously. Compared to another large multicenter observational study from India that comprehensively described both motor and NMS in EOPD, we applied a wider range of scales to assess NMS and particularly focused on assessment of QoL determinants and QoL subdomains, which was not examined specifically in that study. Reference Kukkle, Goyal and Geetha7 However, the age cutoff in the present study (45 years) was different from this other major work from India. In the international context, there have been few longitudinal NMS studies, specifically focusing on EOPD, that have shed light on NMS progression over time and in different stages of the disease. Further details of comparison of major NMS studies from India and internationally can be found in Table 5.
AAO = age at onset; EOPD = early onset Parkinson’s disease; HARS = Hamilton Anxiety rating scale; HDRS = Hamilton Depression Rating Scale; LOPD = late- onset Parkinson’s disease; NMS-Quest-Non motor symptom questionnaire, MOPD = middle-onset Parkinson’s disease; SCOPA-AUT = Scales for Outcomes in Parkinson’s Disease-Autonomic questionnaire; QoL = quality of life; YOPD = young onset Parkinson’s Disease.
Quality of Life
Previous studies have shown that NMS that can adversely affect QoL include RBD, ICD, chronic pain, depression, constipation, and upper gastrointestinal dysfunction. Reference Knipe, Wickremaratchi, Wyatt-Haines, Morris and Ben-Shlomo3,Reference Muslimović, Post, Speelman, Schmand and de Haan36,Reference Mehanna and Jankovic37 Our study confirmed some findings from previous studies, as we demonstrated a significant contribution of total NMS burden and presence of ICDs to the QoL measures. In addition, EDS was identified to be another significant determinant. A recent Vietnamese study examined the association between NMS domains and QoL in EOPD and found that the domains of sleep/fatigue and Mood/cognition were the most likely to affect QoL metrics. Reference Tran, Le Ha and Nguyen38 Mood/cognition and sleep/fatigue were the domains that showed the highest degree of correlation with QoL, in concordance with the Vietnamese study. While the Vietnamese study found that 7 out of 8 NMS subdomains contributed to QoL, we found a significant role of the following 4 NMS subdomains: sleep/fatigue, mood/cognition, urinary dysfunction, and miscellaneous features. Our study is one of the first Indian studies to report data on QoL subdomains. These distinctive patterns of NMS subdomain involvement and their impact on QoL provide novel insights and scientific basis for designing interventions directed at improving the QoL of this group of patients.
A comprehensive cross-sectional evaluation of NMS was done, which was lacking in previous studies that studied only one or a few of the NMS components. However, longitudinal characterization and fluctuations of NMS were not studied and needs further exploration in future studies. The scales employed were screening tools and further confirmatory tests such as polysomnography for RBD, autonomic function testing for suspected autonomic dysfunction, and neuropsychological examination were not done. Furthermore, there was huge gender ratio difference in our study that might significantly alter the results. Genetic data of these patients will be useful to perform genotype–phenotype characterization.
Conclusions
This study done in a large cohort of EOPD patients showed a high overall non-motor symptom burden. In addition to motor disability, non-motor features of excessive day time sleepiness and presence of impulse control disorders were found to significantly influence the QoL of these patients. Within NMS subdomains, sleep/fatigue, mood/cognition, and urinary dysfunction significantly contributed to health-related QoL. Therefore, a comprehensive assessment of non-motor symptoms needs to be incorporated in routine clinical assessment of this group of patients.
Funding
None.
Competing interests
During the preparation of this work, the author(s) used ChatGPT in order to improve language and readability. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the publication.
Statement of authorship
A.P: Conceptualization, data curation, formal analysis, methodology, investigation, and original draft, N.K: Conceptualisation, data curation, formal analysis, investigation, review and editing, A.B: methodology, visualization, validation, review and editing, V.H: methodology, investigation, review and editing, R.Y: conceptualization, project administration, supervision, validation, review and editing, P.K.P: conceptualization, project administration, supervision, validation, review and editing.