1. Background
Persian medicine (PM) is recognized as one of the major traditional medical systems worldwide (1). Within PM, Mizaj (temperament) is a fundamental concept underlying health preservation, disease prevention, and therapeutic decision-making. Although the term “Mizaj” literally means “mixture,” in PM it refers to the qualitative outcome of interactions among the four primary elements: water, earth, fire, and air. Each element is characterized by active (hotness/coldness) and receptive (dryness/wetness) qualities (2). These interactions result in distinct Mizaj categories that explain individual differences in physical, physiological, and behavioral characteristics (3, 4). Mizaj is considered both an inherent constitutional trait (Mizaj-e-jebelli) and a dynamic characteristic that may change over time under the influence of environmental and lifestyle factors. These factors include residential climate, dietary habits, sleep patterns, the level and types of physical activity, stress management and psychosocial conditions, patterns of social interaction, and maintenance of essential principles along with appropriate elimination of waste materials (Setteh-e-zarurieh) (5, 6).
In clinical practice, PM specialists determine Mizaj through a comprehensive evaluation of 10 categories of indicators collectively known as Ajnas-e-Ashara (the 10 essential determinants of temperament). These indicators include tactile quality, body habitus (obesity-leanness), hair characteristics, skin color, physical body structure, responsiveness to external factors, sleep-wake patterns, quality of waste elimination, bodily functions, and mental and emotional states (7).
In recent decades, an increasing body of quantitative and qualitative research has expanded the operational understanding of both general and organ-specific Mizaj assessment based on classical PM sources (8, 9); however, important practical limitations remain.
To improve standardization, Mojahedi et al. developed a 10-item self-report Mizaj questionnaire in 2014 (10), followed by a 20-item questionnaire introduced by Salmannezhad et al. in 2018 to address limitations of the earlier tool (11). Nevertheless, inconsistencies between questionnaire-based assessments have been reported, occasionally yielding discordant or even contradictory results. This variability represents a methodological challenge for both clinical practice and research and underscores the need to evaluate the level of agreement among different approaches to Mizaj assessment.
2. Objectives
The present study was designed to evaluate the level of agreement among commonly used Mizaj assessment methods in PM, including clinical judgment by a PM physician and two standardized self-report questionnaires. By examining the extent to which these approaches yield comparable classifications, the findings may clarify their methodological alignment and potential applicability in both research and clinical settings. This study was designed as an exploratory inter-method agreement analysis and does not constitute a validation or superiority study.
3. Methods
3.1. Study Design and Participants
This descriptive-analytical, comparative cross-sectional study was conducted from October 2023 to February 2025. The study included healthcare workers aged 20 to 60 years from a medical center in Tehran who attended annual health monitoring appointments at occupational health units. After the study objectives were fully explained and written informed consent was obtained, participants were recruited using convenience sampling. The inclusion criteria were age 20 to 60 years, both sexes, no definite history of serious or life-threatening diseases, and absence of motor, cognitive, verbal, or hearing impairments that would limit communication with the volunteer. The exclusion criteria were any debilitating underlying disease and any cognitive, speech, or communication impairment that could interfere with the interview process.
3.2. Data Collection
During the sampling period, a PM specialist conducted an approximately 1-hour interview with each participant. The interview included verification of the inclusion criteria, collection of a detailed medical history, and a clinical Mizaj assessment (CMA) of each volunteer’s Mizaj at the time of the visit. This assessment was based on the Ajnas-e-Ashara (10 criteria), and each indicator within Ajnas-e-Ashara was scored on a 3-point scale (0 = absent, 1 = moderate, 2 = prominent) for both active and receptive qualities. CMA was performed before questionnaire completion to minimize the risk of incorporation bias.
After the clinical assessment, participants were provided with two standardized self-report Mizaj questionnaires to complete; therefore, the physician was blinded to the questionnaire results during the clinical evaluation.
The 10-item Mojahedi Mizaj Questionnaire (MMQ) (Supplementary files) has primarily been validated for the 20- to 40-year age group. The MMQ consists of eight items (Questions 1 - 8) assessing Hotness-Coldness and two items (Questions 9 - 10) assessing Wetness-Dryness (Supplementary 1). Reported sensitivity ranges from 0.52 to 0.65, and specificity ranges from 0.67 to 0.97. The content validity index (CVI) ranges from 0.70 to 1.00, and the Cronbach alpha coefficient is 0.71. Each item has three response options scored as 1 (Cold for Questions 1 - 8 and Wet for Questions 9 - 10), 2 (Balanced for Questions 1 - 10), and 3 (Hot for Questions 1 - 8 and Dry for Questions 9 - 10). Mizaj classification is determined based on the sum of scores and predefined cutoff points. For Hotness-Coldness (sum of items 1 - 8), a score of 14 or lower indicates Cold Mizaj, 15 to 18 indicates Balanced Mizaj, and 19 or higher indicates Hot Mizaj. For Wetness-Dryness (sum of items 9 - 10), a score of 3 or lower indicates Wet Mizaj, 4 indicates Balanced Mizaj, and 5 or higher indicates Dry Mizaj (10).
The Salmannezhad 20-item Mizaj Questionnaire (SMQ) (Supplementary files) has been validated for individuals aged 20 to 60 years. The SMQ consists of 15 items (Questions 1 - 15) evaluating Hotness-Coldness and five items (Questions 16 - 20) assessing Wetness-Dryness. Reported sensitivity and specificity range from 0.63 to 0.80 and 0.57 to 0.85, respectively. Cronbach alpha coefficients range from 0.69 to 0.74. The scale-level content validity index (S-CVI) is 0.93, and the intraclass correlation coefficient (ICC) for total scores between two assessment stages is 0.901 (95% CI, 0.841 - 0.938). Each item has five response options scored as 1 or 2 (Cold for Questions 1 - 15 and Wet for Questions 16 - 20), 3 (Balanced for Questions 1 - 20), and 4 or 5 (Hot for Questions 1 - 15 and Dry for Questions 16 - 20). Mizaj classification is determined based on total scores and predefined cutoff points. For Hotness-Coldness (sum of items 1 - 15), a score of 46 or lower indicates Cold Mizaj, 47 to 49 indicates Balanced Mizaj, and 50 or higher indicates Hot Mizaj. For Wetness-Dryness (sum of items 16 - 20), a score of 14 or lower indicates Wet Mizaj, 15 to 16 indicates Balanced Mizaj, and 17 or higher indicates Dry Mizaj (11). Both questionnaires were designed to determine "Sahih Mizaj" (healthy temperament) and cannot identify "Sue Mizaj" or imbalanced temperament.
Subsequently, agreement analysis was performed based on pairwise comparisons among the three methods.
3.3. Sample Size
The sample size in the parent study was calculated to examine the association between Mizaj and selected biomedical indicators, resulting in a target sample of 383 participants. The present agreement analysis was conducted as a secondary analysis using the same dataset. Although the sample size was determined by the parent study, a post hoc adequacy check using the formula proposed by Donner and Eliasziw (1992) (12) showed that N = 350 was sufficient to estimate a kappa of 0.6 with a precision of ± 0.15 at the 95% confidence level (required n = 349), which is adequate to distinguish moderate from fair agreement.
3.4. Statistical Analysis
Data were analyzed using SPSS version 27. The kappa coefficient was used to assess agreement between Mizaj assessment methods, and chi-square tests were used to examine associations between categorical variables. In this study, P values less than 0.05 were considered statistically significant.
4. Results
A total of 437 individuals were screened, of whom 350 met the eligibility criteria and completed the study protocol. Among the 350 participants, 85 (24.3%) were female and 265 (75.7%) were male.
According to the MMQ, "Balance in active and receptive qualities" was the most frequent Mizaj (29.1%), whereas "Cold and Wet" was the least frequent (2.6%). According to the SMQ, "Hot and Wet" was the most frequently reported Mizaj (26.9%), whereas "Dry" was the least frequent (4.6%). CMA similarly identified "Balance in active and receptive qualities" as the most common category (26.6%) and "Cold" as the least common category (4%).
All participants were aged 20 - 60 years. When grouped according to PM age categories as "Adolescent" (20 - 30 years), "Young Adult" (31 - 40 years), and "Middle-aged" (41 - 60 years), the 41- to 60-year age group had the highest frequency, with 235 participants (67.2%). In this group, "Hot and Wet" was the dominant Mizaj, with a prevalence of 23.8%. Most participants (82.9%) had a university education (Table 1).
Table 1.Demographic Characteristics and Temperament Distribution of Study Participants a
| Characteristic | No. (%) | H | C | W | D | H&W | H&D | C&W | C&D | B |
|---|---|---|---|---|---|---|---|---|---|---|
| Gender | ||||||||||
| Female | 85 (24.3) | 9.4 | 7.1 | 17.6 | 7.1 | 16.5 | 3.5 | 16.5 | 8.2 | 14.1 |
| Male | 265 (75.7) | 13.2 | 6.4 | 12.5 | 6.8 | 22.6 | 12.1 | 3.8 | 2.6 | 20 |
| Age (y) | ||||||||||
| 20 - 30 | 34 (9.7) | 20.6 | 11.8 | 5.9 | 17.6 | 8.8 | 5.9 | 2.9 | 8.8 | 17.6 |
| 31 - 40 | 81 (23.1) | 9.9 | 4.9 | 18.5 | 7.4 | 18.5 | 11.1 | 7.4 | 3.7 | 18.5 |
| 41 - 60 | 235 (67.2) | 11.9 | 6.4 | 13.2 | 5.1 | 23.8 | 10.2 | 7.2 | 3.4 | 18.7 |
| Level of education | ||||||||||
| Under diploma and diploma | 60 (17.1) | 13.3 | 6.6 | 18.3 | 5 | 18.4 | 10 | 6.7 | 3.3 | 18.4 |
| University degree | 290 (82.9) | 12.1 | 6.6 | 12.8 | 7.2 | 21.7 | 10 | 6.9 | 4.1 | 18.6 |
| Mizaj assessment method | ||||||||||
| MMQ | 11.4 | 4.6 | 15.1 | 14 | 11.1 | 9.1 | 2.6 | 2.9 | 29.1 | |
| SMQ | 14 | 8.3 | 8.3 | 4.6 | 26.9 | 10.6 | 14 | 6.3 | 7.1 | |
| CMA | 12 | 4 | 14.3 | 4.9 | 18.9 | 10.8 | 4.3 | 4.3 | 26.6 |
a Abbreviations: MMQ, Mojahedi Mizaj Questionnaire; SMQ, Salmannezhad Mizaj Questionnaire; CMA, Clinical Mizaj Assessment; H, Hot; C, Cold; D, Dry; W, Wet; B, Balance temperament.
Table 2 presents the results of the pairwise comparative analysis of the three Mizaj assessment methods and the percentage agreement among the results of these methods for the nine Mizaj types. Data in this table were initially analyzed using the chi-square test. Because all items in the table were statistically significant (P < 0.001), the kappa coefficient was subsequently used to assess the level of agreement.
Table 2.Pairwise Comparisons Between Three Mizaj Assessment Methods a
| Method of Assessment and Mizaj | No. (%) | Count of Agreement (%) | Kappa Value (95% CI) |
|---|---|---|---|
| Hotness | |||
| MMQ vs CMA | MMQ: 112 (32); CMA: 148 (42.29) | 79 (22.57) | 0.383 (0.273 - 0.493) |
| SMQ vs CMA | SMQ: 182 (52); CMA: 148 (42.29) | 122 (34.86) | 0.512 (0.422 - 0.602) |
| MMQ vs SMQ | MMQ: 112 (32); SMQ: 182 (52) | 89 (25.43) | 0.347 (0.255 - 0.439) |
| Balance in Hot and Cold | |||
| MMQ vs CMA | MMQ: 202 (57.71); CMA: 157 (44.86) | 114 (32.57) | 0.263 (0.153 - 0.373) |
| SMQ vs CMA | SMQ: 70 (20); CMA: 157 (44.86) | 48 (13.71) | 0.202 (0.110 - 0.294) |
| MMQ vs SMQ | MMQ: 202 (57.71); SMQ: 70 (20) | 53 (15.14) | 0.132 (0.058 - 0.206) |
| Coldness | |||
| MMQ vs CMA | MMQ: 36 (10.29); CMA: 45 (12.86) | 23 (6.57) | 0.512 (0.368 - 0.656) |
| SMQ vs CMA | SMQ: 98 (28); CMA: 45 (12.86) | 36 (10.29) | 0.397 (0.287 - 0.507) |
| MMQ vs SMQ | MMQ: 36 (10.29); SMQ: 98 (28) | 30 (8.57) | 0.350 (0.242 - 0.458) |
| Dryness | |||
| MMQ vs CMA | MMQ: 92 (26.29); CMA: 70 (20) | 46 (13.14) | 0.441 (0.329 - 0.553) |
| SMQ vs CMA | SMQ: 72 (20.57); CMA: 70 (20) | 36 (10.29) | 0.382 (0.262 - 0.502) |
| MMQ vs SMQ | MMQ: 92 (26.29); SMQ: 72 (20.57) | 42 (12) | 0.366 (0.252 - 0.480) |
| Balance in Dry and Wet | |||
| MMQ vs CMA | MMQ: 158 (45.14); CMA: 151 (43.14) | 89 (25.43) | 0.241 (0.137 - 0.345) |
| SMQ vs CMA | SMQ: 105 (30); CMA: 151 (43.14) | 52 (14.86) | 0.081 (-0.023 to 0.185) |
| MMQ vs SMQ | MMQ: 158 (45.14); SMQ: 105 (30) | 61 (17.43) | 0.162 (0.060 - 0.264) |
| Wetness | |||
| MMQ vs CMA | MMQ: 100 (28.57); CMA: 129 (36.86) | 67 (19.14) | 0.388 (0.286 - 0.490) |
| SMQ vs CMA | SMQ: 173 (49.43); CMA: 129 (36.86) | 95 (27.14) | 0.358 (0.262 - 0.454) |
| MMQ vs SMQ | MMQ: 100 (28.57); SMQ: 173 (49.43) | 74 (21.14) | 0.282 (0.188 - 0.376) |
a Abbreviations: MMQ, Mojahedi Mizaj Questionnaire; SMQ, Salmannezhad Mizaj Questionnaire; CMA, Clinical Mizaj Assessment; CI, confidence interval.
Across the entire table, the highest kappa values (kappa = 0.512) were observed in two comparisons: between the CMA and the SMQ for Hotness (95% CI, 0.422 - 0.602) and between the CMA and the MMQ for Coldness (95% CI, 0.368 - 0.656).
For the CMA-SMQ comparison, the highest agreement was observed for Hotness (kappa = 0.512), whereas the lowest agreement was observed for the Balance in Dry and Wet (kappa = 0.081).
For the CMA-MMQ comparison, the highest agreement was observed for Coldness (kappa = 0.512), whereas the lowest agreement was observed for the Balance in Dry and Wet (kappa = 0.241).
In comparisons between the two questionnaires (SMQ and MMQ), the highest agreement was observed for Dryness (kappa = 0.366), whereas the lowest agreement was observed for the Balance in Hot and Cold (kappa = 0.132).
5. Discussion
The present study was designed to evaluate the level of agreement between commonly used methods of Mizaj assessment in PM, including clinical judgment by a PM physician (CMA) and two standardized self-report questionnaires (MMQ and SMQ). By determining the extent to which these approaches yield comparable temperamental classifications, the findings help clarify their methodological alignment and clinical applicability.
According to the interpretation criteria proposed by Landis and Koch (1977) (13), kappa values between 0.41 and 0.60 represent moderate agreement, values between 0.21 and 0.40 represent fair agreement, and values between 0.00 and 0.20 represent slight agreement. In this study, most coefficients fell within the slight to moderate ranges, reflecting heterogeneous alignment between assessment approaches across different temperament dimensions. This indicates that, although certain dimensions show partial overlap, none of the approaches are fully interchangeable. This pattern suggests that clinical and self-report assessments capture complementary but distinct aspects of the temperament construct.
When comparing CMA with the MMQ, the highest agreement was observed for the Coldness dimension (kappa = 0.512; 95% CI, 0.368 - 0.656), representing moderate agreement. This suggests a moderate level of correspondence between MMQ and CMA classifications of Cold temperament, possibly because both approaches emphasize relatively observable physiological and somatic indicators. By contrast, the lowest agreement in this pair occurred for Balance in Dry and Wet (kappa = 0.241), which falls within the fair agreement range. Although this level of agreement was somewhat stronger than that observed between CMA and SMQ for the same dimension, it remained limited. This relatively weak correspondence may be due to restricted item coverage in the MMQ, which primarily evaluates wetness and dryness through indices such as obesity or leanness and skin texture (10), and may therefore fail to capture subtler clinical signs recognized by physicians during face-to-face assessments.
When comparing CMA with the SMQ, the highest concordance was observed for the Hotness dimension (kappa = 0.512; 95% CI, 0.422 - 0.602), indicating moderate agreement. This finding suggests that both methods identify individuals with Hot temperament features in a relatively similar manner. One possible explanation is that manifestations of hotness, such as thermal sensation, behavioral tendencies, and other perceptible features, may be more readily recognized by both respondents and clinicians. In contrast, the weakest alignment in this pair was found for Balance in Dry and Wet (kappa = 0.081; 95% CI, -0.023 to 0.185), corresponding to slight agreement. The confidence interval overlapping zero further indicates that this agreement may not be statistically meaningful. This weak correspondence suggests that questionnaire-based self-perceptions are limited in capturing the clinical cues physicians use to judge moisture balance. Possible contributors include self-report bias, uncertainty in respondents' self-perception, and a mismatch between qualitative clinical markers and questionnaire item content.
Agreement between the two questionnaires (MMQ and SMQ) was the lowest among all pairwise comparisons, ranging from kappa = 0.132 to 0.366, corresponding to slight to fair agreement. The strongest alignment occurred for Dryness (kappa = 0.366), whereas the weakest was found for Balance in Hot and Cold (kappa = 0.132). Although both tools are grounded in similar theoretical concepts derived from PM, they appear to differ in item weighting, response scaling, and interpretive thresholds. This finding indicates that standardized questionnaires developed for Mizaj assessment should not be considered interchangeable, as they may diverge considerably in their classification outcomes. The consistently modest kappa values point to both methodological and conceptual differences among the tools. Clinical evaluation incorporates multisensory and interpretive cues, including visual observation, palpation, and narrative context (14-16). In contrast, questionnaires operationalize temperament through structured self-report domains, which may improve reproducibility but can also oversimplify complex temperamental nuances.
As a result, reliance on questionnaire data alone may lead to misclassification, particularly in borderline or mixed cases. These observations support a complementary rather than competitive role for clinical and questionnaire-based assessments. In practice, questionnaires may serve as efficient screening instruments, whereas comprehensive clinical evaluation remains essential for confirmation and therapeutic decision-making. The broader implications are consistent with evidence from psychometric studies showing that even widely validated temperament-related tools, such as the Persian TCI 140, exhibit differential factor structures and internal consistencies across cultural contexts, underscoring the inherently multidimensional nature of temperament measurement (17).
Although assessment agreement was the primary focus of the present study, the findings related to Mizaj distribution and demographic patterns also offer useful contextual insight. The most frequent categories in both CMA and MMQ were Balance in active and receptive qualities, whereas the SMQ more often identified Hot and Wet temperament. Similar distributions have been reported among patients with metabolic syndrome in Qazvin, although differences in sampling strategy and health status may explain part of the divergence. Another Iranian study by Parvizi et al. (18) in Shiraz identified Wet as the dominant temperament, a difference that may be influenced by climate and dietary habits. Variations from the findings reported by Aziz et al. in India (19) may likewise reflect ecological conditions as well as differences in the studied populations (20).
With respect to age, classical PM doctrines associate youth with Hot and Wet temperament and aging with Cold and Dry temperament. Although some age-related shifts were observed in the present study, the trends were less distinct than traditional expectations. This discrepancy may be related to modern lifestyle and psychosocial factors, such as sedentary behavior, dietary modification, sleep disturbance, and mental stress (20-23). These secondary observations further illustrate the complex interaction between inherited constitution and environmental modulation in contemporary populations.
5.1. Limitations
This study has several limitations. The evaluation was conducted by a single PM specialist, reflecting routine clinical practice; although this limits estimation of interrater reliability, future multi-rater studies should address this gap. Because no independent external gold standard exists for Mizaj assessment, the findings should be interpreted as reflecting agreement between methods rather than diagnostic validity. Although incorporation bias is a concern in agreement studies, it was mitigated here because clinical assessment was completed before, and independently of, questionnaire administration. The study population consisted predominantly of male healthcare workers, which may limit generalizability. Finally, as with all kappa-based analyses, the observed coefficients may have been influenced by the prevalence and distribution of temperament categories.
5.2. Conclusions
The findings of the present study show that inter-method agreement in Mizaj assessment is limited and generally remains within the slight to moderate range. The physician's clinical judgment showed closer alignment with the Salmannezhad questionnaire for Hotness and with the Mojahedi questionnaire for Coldness, while very weak or near-zero agreement in some domains, particularly Balance in Dry and Wet, indicates that certain aspects of Mizaj remain difficult to standardize across instruments. Further refinement of self-report content, larger validation studies, and cross-cultural calibration of diagnostic criteria are needed to improve methodological coherence. Integrating multiple sources of data, including clinical observation, physiological indicators, and psychometric profiles, may provide a more faithful operationalization of the multifaceted concept of Mizaj in both research and clinical practice.
