1Senior Specialist registrar pediatric, Aljalila children's Specialty Hospital, Dubai- 25314, UAE.
2National Nutrition Institute, 16- Kasr Al-Ainy St., Cairo, Egypt.
3Fellow of Pediatrics at Shebin El-Kom Teaching Hospital, Shebin El-Kom 32511, Egypt.
4Fellow of Pediatrics and Neonatology at Damanhour Teaching Hospital, Damanhour-22742 Egypt.
5Faculty of Medicine, Department of Physiology, Al-Azhar University (Assiut), Assiut- 71524 Egypt.
6Specialist neonatologist, Dubai Hospital, Dubai- 25314, UAE
7Faculty of Medicine, Department of Medical biochemistry, Al-Azhar University, Cairo, Egypt
8AlGhad College for Applied Medical Sciences in Najran- 55461, Saudi Arabia
9Faculty of Nursing, Department of critical care and emergency nursing, Minia University, Egypt.
Mohamed Zaeim Hafez Ahmed, Faculty of Medicine, Department of Physiology, Al-Azhar University (Assiut), Assiut- 71524 Egypt.
Mohamed Zaeim Hafez Ahmed, et,al. Serum Zinc and Selenium Levels: Are they Have a Vital Role in Influencing the Etiopathogenesis of Children with Complex Febrile Seizures?. J. Clin. Pediatr. Care. Vol. 3 Iss. 1. (2025) DOI: 10.58489/2836-8630/014
© 2025 Mohamed Zaeim Hafez Ahmed, this is an open-access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Complex febrile seizures, Serum zinc, Selenium, Children, Oxidative stress
Objective: to determine the impact of serum zinc and selenium levels on the etiology and risk of febrile seizures in children.
Methods: From August 2024 to June 2025, a case control study was carried out at al-Azhar University Hospital for 80 children with complex febrile seizures (CFS) and 40 healthy children to evaluate the impact of serum zinc and selenium on the etiopathogenesis of children with febrile seizures.
Results: According to serum selenium levels, there was no discernible difference between CFS patients and healthy controls (P=0.214). However, CFS patients had a considerably lower serum zinc level (92.18 ± 15.47) than healthy controls (134.93 ± 17.06) (P<0.001). At AUC 0.580 (P=0.156), the sensitivity was 91.9%, with a cutoff threshold of serum selenium levels of >49.9. Additionally, the sensitivity was 96%, specifically 89.8% at AUC 0.974 (P<0.001), and the cutoff value for serum zinc levels was >91.1.
Conclusion: Because zinc lowers oxidative stress and raises the seizure threshold, it may be a major factor in preventing febrile seizures. Supplementing with zinc may help shield kids from the negative consequences of febrile seizures and enhance the overall effectiveness of treatment.
Also referred to as febrile fits or febrile convulsions, febrile seizures are the most prevalent kind of seizure in children. It is described as a seizure that typically occurs in children between the ages of 6 and 60 months along with a fever that is typically higher than 100.4 F (or 38 C) [1-3]. About 2% to 5% of children have FS, which is divided into two simple and complex types [4-6].
According to certain attributes, FSs are divided into simple FSs and complicated FSs [7]. Simple FS is characterized by seizures that do not recur within 24 hours, have generalized features, and last less than 15 minutes [8-10]. On the other hand, extended focal seizures that happen more than once in a 24-hour period are a hallmark of complex FS [11,6].
According to [12-15], certain vaccinations, viral infections, developmental delay, discharge from a neonatal unit after 28 days, day care attendance, and possibly iron and zinc deficiencies are risk factors for febrile seizures. Seizures with a fever have multiple causes. Frivolous seizures are generally thought to be caused by a developing central nervous system's (CNS) susceptibility to the effects of fever, along with underlying genetic predispositions and environmental factors [16,1,17].
Once disorders of the central nervous system have been ruled out, febrile seizures are diagnosed. Determining the infection causing the fever is the main goal of clinical evaluation. To rule out secondary causes of seizures, neurological examinations are required. Several aspects should be covered in the evaluation, including the type and duration of seizures, the length of the postictal period, any recent infections or fevers, recent antibiotic treatments, associated symptoms, history of vaccinations and immunizations, prior episodes of FS or epilepsy diagnoses, and other neurological conditions [4,18,19].
Trace elements like zinc (Zn) and selenium (Se) are vital for many physiological functions, especially cellular metabolism, antioxidant defense, and the immunological response [20-23]. These components are essential for preserving homeostasis. The high concentration of zinc in the brain and its possible function in synaptic neurotransmission have made zinc status in children with febrile seizures a significant topic of discussion [24]. Because zinc facilitates the inhibitory impact of NMDA receptors, it controls the rate-limiting enzyme in GABA production and inhibits excitatory neuronal firing [25,26].
According to [27-29], selenium (Se) is a trace element that is vital to human health and plays important roles in growth, development, metabolism, immunological function, and antioxidative defense. Selenium is a key factor in the development of epilepsy, according to recent research. The potential involvement of selenium in the pathophysiology of febrile seizures has not been well studied [30-32]. Therefore, the goal of the current study is to evaluate how serum zinc and selenium levels affect the etiology and risk of febrile seizures in children.
To determine the impact of serum zinc and selenium on the etiopathogenesis of children with febrile seizures, a case control study involving 80 children with complex febrile seizures (CFS) and 40 healthy children will be carried out at al-Azhar University Hospital between August 2024 and June 2025.
Ethics Approval and Consent to Participate
According to the authors, the study complies with the Declaration of Helsinki of the World Medical Association. The entire study protocol was authorized by the Al-Azhar University Faculty of Medicine local committee (Research/AZ.AST./ PHY003/5/228/1/2024). All participants received explanations of the benefits, potential risks, and each step of the process. Each participant signed an informed consent form prior to participating in the study.
Patients’ Selection Criteria
The second control group consisted of 40 healthy children who were brought to the pediatric outpatient clinic for normal follow-up and 40 children who went to the pediatric emergency clinic with complex febrile seizures (CFS). Children who shared the same gender, height, and weight as the case group were chosen to form the control group.
Diagnosis of CFS
Additionally, according to the revised Japanese guidelines for the therapy of FS, CFS is defined as FS with one or more of the following three characteristics: focal symptoms, prolonged (≥15 min) duration, and recurrence within 24 hours [33]. Even if the patient's temperature was below 38.0 °C at admission, they were considered febrile cases and included in this study if their temperature was ≥38.0 °C within 24 hours before or following the beginning of seizures.
Exclusion Criteria
The case and control groups were created by excluding patients with congenital metabolic disorders, epilepsy, seizures without fever, chronic diseases, asphyxia history, evidence of central nervous system infection, central nervous system structural abnormalities, developmental delay, electrolyte imbalances, or physical signs of systemic disease. The study excluded children whose socioeconomic situation varied substantially (for instance, based on parental education level, income level, or reliance on social assistance programs). Children who reported severe weight loss or appetite loss within the previous three months, or who were diagnosed with malnutrition because of inadequate or unbalanced diet,were not included in the study. Children who took supplements of any vitamins, minerals, or trace elements, such as zinc or selenium, were also not allowed to participate in the study. The study did not include children with any chronic conditions that could impact the metabolism of zinc and selenium, such as inflammatory bowel illnesses, persistent infections, or malabsorption syndromes.
All the Patients were Subjected to
Complete history, including name, age, sex, height, weight, habits, family history, medical history, etc. Clinical examination, including local and general examinations, including blood examination results following seizure onset, body temperature at admission, presence of convulsive seizures, classification of CFS, total number and type of anticonvulsants used, duration of convulsive seizures, duration of impaired consciousness, and blood examination results. Parental or observer interviews, emergency transport records, referrals, and/or medical records were used to establish the exact moment at which seizures began. Hospitalization was assessed by selecting and extracting the data from the final seizure in cases where there were numerous seizures. White blood cell, hemoglobin, platelet, and liver function tests included ALT and AST, while renal function testing included urea and creatine levels (Cr). The Agilent 7500 CE ICP-MS equipment was used to evaluate serum selenium samples using the inductively coupled plasma mass spectrometry method, while the Shimadzu AA6800 device was used to test serum zinc level samples using the atomic absorption spectrophotometer method. These FS patients had blood samples taken for zinc and selenium during the first hour following the seizure.
Sample Size Estimation
To determine the patient sample size, we created an independent case-control study with one control per case. According to a 2025 study by [34], FS affects 2% to 5% of children and is the most prevalent kind of seizure in childhood. To do this, 80 experimental subjects and 40 control subjects would need to be studied. The Type I error probability associated with this null hypothesis test is 0.05. A chi-squared, uncorrected statistic was used to evaluate this null hypothesis.
Statistical Analysis
Statistical package for social science (SPSS) version 17.0 (Chicago, USA) was used to code, tabulate, and analyze the data. A mean ± standard deviation (SD) for parametric data and a median (range) for non-parametric data were used to compute descriptive statistics for the anthropometric measurements and laboratory data. The receiver operating characteristic (ROC) curves, Pearson correlation (r), Student T test (t), Mann-Whitney test (U), Fisher's exact test (FET), and chi-squared (χ2) measures. All analyses were deemed statistically significant if the P value was less than 0.05.
The study's population flow chart includes 126 children who arrived at the pediatric emergency department at Al-Azhar University Hospital's Faculty of Medicine with complex febrile seizures (CFS). 120 children took part in the study and were statistically evaluated, 80 of whom were CFS patients and 40 of whom were healthy controls. Six patients were eliminated from the study (2 patients refused consent, and 4 did not fit the inclusion criteria), (Figure 1).

Figure 1. Flow chart of the patients studied.
Also, age and sex did not significantly differ between CFS patients and healthy controls in our study (P>0.05). Family history of FS and family history of epilepsy were significantly different between CFS patients and healthy controls (P<0.05); nevertheless, 17 (21.25%) of CFS patients had a family history of FS, and 32 (40%) had a family history of epilepsy, (Table 1).
|
Groups |
CFS patients N=80 |
Healthy controls N=40 |
X2 |
P value |
|
Age/month Mean ± SD |
24.29 ± 12.87 |
27.38 ± 10.41 |
t= -1.412 |
0.161 |
|
Sex Male Female |
41 (51.25) 39 (48.75) |
17 (42.5) 23 (57.5) |
0.818 |
0.366 |
|
Family history of FS |
17 (21.25) |
0 (0.0) |
9.903 |
0.002* |
|
Family history of epilepsy |
32 (40) |
0 (0.0) |
21.818a |
<0.001* |
Complex Febrile Seizures (CFS), Chi-Squared (χ2), Standard deviation (SD) * Significant for P value <0.05
Table 1. Demographic data among CFS patients and Healthy controls.
Acute gastrointestinal infection (AGE) was detected in 25 (31.25%) of CFS patients, whereas upper respiratory tract infection (URTI) was detected in 55 (68.75%) CFS patients. Additionally, there was a significant difference between CFS patients and healthy controls in terms of infection (P<0.001).
Additionally, the mean number of convulsive seizures, duration of seizures, and time from the start of a seizure to complete consciousness recovery were 17.15 ± 41.36, 12.28 ± 6.94, and 209.55 ± 249.44 among CFS patients, respectively (Table 2).
|
Groups |
CFS patients |
Healthy controls |
X2 |
P value |
|
|
N=80 |
N=40 |
|
|
|
Infection No URTI AGE |
0 (0.0) 55 (68.75) 25 (31.25) |
40 (100) 0 (0.0) 0 (0.0) |
120.00 |
<0.001* |
|
Convulsive seizure duration (min) Mean ± SD |
17.15 ± 41.36 |
--- ± --- |
--- |
--- |
|
Duration of seizure(min) Mean ± SD |
12.28 ± 6.94 |
--- ± --- |
---- |
---- |
|
Time from seizure onset to full recovery of consciousness (min) Mean ± SD |
209.55 ± 249.44 |
--- ± --- |
--- |
---- |
Upper respiratory tract infections (URTI), Acute Gastroenteritis (AGE), Complex Febrile Seizures (CFS) Chi-Squared (χ2), Standard deviation (SD) * Significant for P value <0.05
Table 2. Comparison between CFS patients and Healthy controls regarding clinical data.
Moreover, hemoglobin, WBC count, platelet, MPV, neutrophil, lymphocyte, and CRP did not significantly differ between CFS patients and healthy controls (P>0.05) (Table 3).
|
Groups |
CFS patients N=80 |
Healthy controls N=40 |
t |
P value |
|
|
Mean ± SD |
Mean ± SD |
|
|
|
Hemoglobin (g/dL) |
11.73 ± 1.18 |
11.25 ± 1.31 |
1.919 |
0.059 |
|
WBCs count (109/L) |
10.81 ± 3.44 |
10.95 ± 3.53 |
-0.204 |
0.839 |
|
Platelet (109/L) |
296.28 ± 58.84 |
299.43 ± 64.93 |
-0.258 |
0.797 |
|
MPV (f/L) |
7.39 ± 0.74 |
7.35 ± 0.77 |
0.221 |
0.826 |
|
Neutrophil (109/L) |
9.29 ± 2.56 |
9.16 ± 2.82 |
0.229 |
0.820 |
|
Lymphocyte (109/L) |
3.07 ± 1.02 |
3.27 ± 1.02 |
-1.003 |
0.319 |
|
CRP (mg/L) |
20.91 ± 9.64 |
22.52 ± 10.27 |
-0.828 |
0.411 |
Table 3. Laboratory examinations among CFS patients and Healthy controls.
Complex Febrile Seizures (CFS), White Blood Cells (WBCS), Mean platelet volume (MPV), C-Reactive Protein (CRP) Student T test (t), Standard deviation (SD) * Significant for P value <0.05 According to data from the current investigation, serum selenium levels did not significantly differ between CFS patients and healthy controls (P=0.214). However, CFS patients' serum zinc levels were substantially lower (92.18 ± 15.47) than those of healthy controls (134.93 ± 17.06) (P<0.001) (Table 4, Figure 2,3).
|
Groups |
CFS patients N=80 |
Healthy controls N=40 |
U |
P value |
|
Serum selenium levels (μg/L) Mean ± SD |
50.86 ± 9.06 |
53.53 ±11.78 |
1.256 |
0.214 |
|
Serum zinc levels (μg/dL) Mean ± SD |
92.18 ± 15.47 |
134.93 ± 17.06 |
13.341 |
<0.001* |
Table 4. Comparison between CFS patients and Healthy controls regarding Serum selenium levels and Serum zinc levels. Complex Febrile Seizures (CFS), Standard deviation (SD), Mann-Whiteny U test (U), * Significant for P value <0.05

Figure 2. Serum selenium levels among CFS patients and Healthy controls.

Figure 3. Serum zinc levels among CFS patients and Healthy controls.
ROC curve study showed that the sensitivity was 91.9%, specifically 79.4% at AUC 0.580 (P=0.156), and the cutoff limit for serum selenium levels was >49.9. Additionally, the cutoff value for serum zinc levels was >91.1 with a 96% sensitivity, specifically 89.8% at AUC 0.974 (P<0.001), which may have played a crucial role in the etiopathogenesis of children with complex febrile seizures in comparison to selenium (Table 5, Figure 4).
|
Variables |
AUC |
Cutoff level |
Std. Error |
Sensitivi- ty % |
Specificity % |
P value |
Asymptotic 95% Confidence Interval |
|
|
Lower Bound |
Upper Bound |
|||||||
|
Serum selenium levels (μg/L) |
0.580 |
>49.9 |
0.064 |
91.9 |
79.4 |
0.156 |
0.454 |
0.706 |
|
Serum zinc lev- els (μg/dL) |
0.974 |
>91.1 |
0.011 |
96.0 |
89.8 |
<0.001* |
0.951 |
0.996 |
AUC: Area Under Curve * for significant P value
Table 5. The Cutoff level for Serum selenium levels and Serum zinc levels.

Figure 4. the ROC curve analysis for Serum selenium levels and Serum zinc levels for detection etiopathogenesis of children.
In our study, the correlation between the cumulative percentage of patients with persistently impaired consciousness and the time from seizure onset. It shows that the median time from seizure onset to full recovery of consciousness was 186 minutes (range: 1–1100 minutes), and that 12.83%, 26%, and 11% of the patients, respectively, experienced impaired consciousness for more than 6, 8, and 12 hours (Figure 5).

Figure 5a. Distribution of time from seizure onset to full recovery of consciousness in patients with CFS. The horizontal axis shows the duration of convulsive seizures, separated by 200 min intervals. The vertical axis shows the number of patients.

Figure 5b. Association between duration of seizure in patients with CFS. The horizontal axis shows the duration of convulsive seizures, separated by 10 min intervals. The vertical axis shows the number of patients.

Figure 5c. Association between convulsive seizure duration in patients with CFS. The horizontal axis shows the duration of convulsive seizures, separated by 10 min intervals. The vertical axis shows the number of patients.
There was a substantial negative link between serum selenium levels and neutrophils in our study, while there was a significant positive correlation with age, WBC count, and lymphocyte. However, blood selenium levels did not significantly correlate with hemoglobin, platelet, MPV, CRP, length of convulsive seizures, duration of seizure, or time from seizure commencement to full return of consciousness (P>0.05), (Table 6). Additionally, blood zinc levels and age and WBC count had a substantial positive correlation (P<0.05), while serum zinc levels and neutrophils had a significant negative correlation (P=0.022). The duration of convulsive seizures, the time from the start of the seizure to complete consciousness, lymphocytes, hemoglobin, platelets, MPV, and CRP, however, did not significantly correlate with blood zinc levels (P>0.05), (Table 6).
|
|
Serum selenium levels (μg/L) |
Serum zinc levels (μg/dL) |
|
|
Age (months) |
r |
0.240 |
0.414 |
|
p value |
0.032* |
<0.001* |
|
|
Convulsive seizure duration (min) |
r |
-0.042 |
-0.102 |
|
p value |
0.710 |
0.369 |
|
|
Duration of seizure(min) |
r |
0.042 |
0.005 |
|
p value |
0.710 |
0.966 |
|
|
Time from seizure onset to full recovery of consciousness (min) |
r |
0.002 |
-0.044 |
|
p value |
0.987 |
0.697 |
|
|
WBCs count (109/L) |
r |
0.581 |
0.432 |
|
p value |
<0.001* |
<0.001* |
|
|
Neutrophil (109/L) |
r |
-0.328 |
-0.256 |
|
p value |
0.003* |
0.022* |
|
|
Lymphocyte (109/L) |
r |
0.288 |
-0.115 |
|
p value |
0.010* |
0.311 |
|
|
Hemoglobin (g/dL) |
r |
-0.005 |
0.009 |
|
p value |
0.963 |
0.939 |
|
|
Platelet (109/L) |
r |
0.065 |
-0.086 |
|
p value |
0.564 |
0.448 |
|
|
MPV (f/L) |
r |
-0.158 |
-0.187 |
|
p value |
0.162 |
0.097 |
|
|
CRP (mg/L) |
r |
-0.157 |
-0.206 |
|
p value |
0.165 |
0.066 |
|
White Blood Cells (WBCS), Mean platelet volume (MPV), C-Reactive Protein (CRP) , r: correlation coefficient * Significant for P value
Table 6. Correlation between Serum selenium levels and Serum zinc levels and other variables studied.
For individuals with FSs, determining the precise type of seizure may be crucial to creating a suitable treatment approach [35,36]. This is the first article that describes the specific aspects of acute-phase CFS, to the best of our knowledge, despite the recent publication of major research on the clinical features of febrile seizures [37,38].
We found no significant difference in hemoglobin, WBC count, platelet, MPV, neutrophil, lymphocyte, and CRP between CFS patients and healthy controls. Regardless of the median NLR value, [39] observed no correlation between NLR and FS type. Notably, the FS and fever groups did not differ statistically significantly in WBC count, neutrophil ratio, NLR, PLR, or CRP [40,41]. According to [42] found no discernible differences in white blood cell (WBC) counts between the group experiencing febrile seizures and the group not experiencing seizures [43]. Alongside their fever, children who experienced febrile seizures also showed a lower lymphocyte count than those who did not. After about four to five hours, an inflammatory response may be the cause of this spike. On the other hand, it could potentially be connected to the circulation of toxic substances in the blood [44]. According to [42], it was also not surprising that neutrophilia and lymphopenia, which are linked to elevated cortisol levels, caused the NLR ratio to rise. Therefore, for many individuals with FSs, a CBC test and accurate NLR measures might not be required.
Convulsive seizure means, seizure duration, and time from seizure beginning to complete recovery of consciousness were 17.15 ± 41.36, 12.28 ± 6.94, and 209.55 ± 249.44 for our CFS patients, respectively. The results of earlier studies on seizure length and recovery time are different from this. According to [45], 90.2% of seizures lasted less than five minutes, and 93% of them resulted in unconsciousness lasting less than thirty minutes. According to [46,47], the median recovery time and duration of FS were 18 minutes and 2.5 minutes, respectively. According to [48], 6.8% of convulsive seizures in CFS lasted more than two hours, with a median duration of 30.5 minutes.
Because of their effects on synaptogenesis, ion channels, neurotransmitters, oxidative stress, and the immune system, zinc and selenium are implicated in the etiopathogenesis of seizures [49]. It has been shown that the action of cytokines causes plasma zinc levels to drop quickly during infection. [50]
The zinc levels in the FS group were likewise lower than those in the group of children with fever but no seizures [21,51]. However, research by [52,53] found no significant difference in serum zinc levels between the children with FS and the group with fever but no seizures. According to our study, children with CFS had considerably higher serum zinc levels than children in good health. When FSs occur, the body releases zinc to raise the seizure threshold, lower oxidative stress, and restore equilibrium. In children, it is noteworthy that a further decrease in serum zinc levels during an infection may increase neuronal excitability, hence raising the risk of seizures. This procedure emphasizes Zinc’s possible protective function in FS management. [54] found that, in comparison to the placebo group, zinc sulfate supplementation significantly decreased the frequency of FS recurrence in children with normal serum zinc levels. Other research and our findings indicated that zinc supplementation might help avoid FSs. Supplementing with zinc sulfate at a dose of 2 mg/kg/day (maximum 50 mg) during infection periods or seasons where infections are common may lower the risk of seizures in children aged 6–60 months, when FSs are most common. Zinc supplementation may be specifically targeted at children who have a family history of FSs, seizures at low-grade fever, malnutrition or micronutrient deficiencies, a genetic mutation that predisposes them to FSs, recurrent FSs, or significantly lower serum zinc levels during infections.
During times of infection, when inflammatory biomarkers rise, serum selenium levels fall [55,56]. Serum selenium levels' function in FSs is poorly known because there aren't many studies on the subject and most of them use tiny patient samples. Selenium levels in CFS patients and healthy children did not significantly differ in our investigation. Likewise, [57] did not discover any statistically significant difference in serum selenium levels between children who had FSs and those who had a fever but no vomiting.
Other recent research has demonstrated that selenium is a key factor in the development of epilepsy. Selenium's potential involvement in the pathophysiology of febrile seizures has been the subject of several investigations [30-32]. Regarding this issue, Ataş and Poyrazoğlu (2023) discovered that serum selenium levels did not significantly rise during seizure activity, but they did significantly drop after infection. But according to, [58] children with FSs had lower selenium levels than those who had fever but no seizures. This could be because, to preserve oxidative balance during FS, the body may turn to other antioxidant mechanisms such glutathione, vitamin E, or vitamin C. Additionally, some people are more susceptible to selenium due to genetic differences that impact selenium metabolism. Although there have been no notable changes in serum selenium levels during FSs, it is possible that some melanoproteins or brain tissue may still have altered selenium levels. Thus, a better understanding of selenium's connection to seizures may be possible through cellular-level analyses that look at its impacts. To achieve a more effective diagnostic approach for FS cases, future research involving more patients from multiple centers should investigate the combined use of clinical evaluations and biomarkers. This approach should focus on comparing serum levels of zinc and selenium before and after seizures, taking genetic variations into account. These studies will improve our comprehension of the connection between FSs and these trace elements and could aid in the creation of new treatment approaches.
According to our research, CFS patients had substantially lower serum zinc levels than healthy controls. Superior serum selenium had a sensitivity of 91.9%, or 79.4%, whereas the cutoff limit for serum zinc levels was >91.1, with a sensitivity of 96%, or 89.8%. Finally, zinc may be crucial in preventing febrile seizures by lowering oxidative stress and raising the seizure threshold. Supplementing with zinc may help shield kids from the negative consequences of febrile seizures and enhance the overall effectiveness of treatment.
Consent for Publication
all authors have read and revised well for the manuscript and agree to publish.
Availability of Data and Material
All data supporting the study are presented in the manuscript or available upon request.
Competing Interests
There is no conflict of interest.
Funding
The author received no financial support for this article's research, authorship, and publication.
Acknowledgments
Not applicable
Authors' Information (optional)
Not applicable