Study characteristics
We retrieved 1642 studies. After removing duplicates, 890 studies were evaluated. By reading the titles and abstracts and excluding irrelevant articles and reviews, the remaining 346 articles were further evaluated. After reading the full text, a total of 20 studies were included in the final meta-analysis. Seventeen were case-control studies[25-41], and three were cross-sectional studies[42-44]. The publication years ranged from 2005 to 2021, with a total of 3642 participants. The NOS results showed that the methodological quality was generally high. The vast majority of participants were young and middle-aged women, including people in Asia, Europe, and North America. Among them, Egypt, Turkey, and Iran included three studies; India and Iran included two studies; and Bangladesh, China, Jordan, Nepal, Pakistan, Saudi Arabia, and the United States included one study each. The included studies evaluated serum ferritin level, serum iron level, and prevalence of ferritin deficiency. The subjects were patients with acute and chronic telogen effluvium. We performed a subgroup analysis of different hair loss types to explore the similarities and differences in serum iron and ferritin levels. The subjects with acute and chronic telogen effluvium were placed in separate subgroups, and those with unclassified telogen effluvium were placed in the telogen effluvium group. The literature search flowchart is shown in Figure 1. Table 1 presents the characteristics of the included studies.
Table 1
Baseline characteristics of the included studies.
Author (year)
|
Type of study
|
Sample size(P/C)
|
country
|
Sex
(P/C)
|
Age(P/C)
(mean ± SD, y)
|
Type
|
Diagnosis
|
Data collected
|
NOS
|
Alizadeh et al.2021
|
Case-control
|
83/83
|
Iran
|
Female
|
36.78±5.88
|
CTE
|
History, physical examination, pull test,dermoscopy
|
serum ferritin, prevalence of ferritin deficiency
|
7
|
Cheng et al.2021
|
Cross-Sectional
|
193/183
|
China
|
NA/Female
|
28.37±5.11/29.11±5.46
|
TE, ATE, CTE
|
History, physical examination, pull test,dermoscopy
|
serum iron, serum ferritin
|
7
|
Chisti et al.2012
|
Case-control
|
100/100
|
India
|
Female
|
26.6±7.25/26.83±9.97
|
ATE, CTE
|
History, physical examination
|
serum ferritin
|
7
|
Elethawi et al.2012
|
Case-control
|
38/25
|
Iraq
|
Female
|
30.15±6.5
|
CTE
|
History, physical examination, pull test
|
serum ferritin
|
7
|
Ertug et al.2018
|
Case-control
|
455/196
|
Turkey
|
Female
|
29.01±8.71/31.29±8.7
|
TE
|
History, physical examination, pull test
|
serum iron
|
7
|
Fatani et al.2015
|
Cross-Sectional
|
160/425
|
Saudi Arabia
|
Female
|
28.94±10.6/34.24±11.4
|
TE
|
History, physical examination
|
serum ferritin
|
6
|
Hamad et al.2010
|
Case-control
|
34/26
|
Egypt
|
Female
|
22.95±7.31/19.06±9.82
|
TE
|
History, physical examination, pull test, trichogram
|
serum iron
|
7
|
Hasseeb et al.2020
|
Case-control
|
53/26
|
Iraq
|
Female
|
Range 18–45
|
ATE, CTE
|
History, physical examination
|
serum iron, serum ferritin
|
7
|
Hodeib et al.2017
|
Case-control
|
40/20
|
Egypt
|
Female
|
28.10±4.65/27.45±6.41
|
TE, ATE, CTE
|
History, physical examination, pull test,dermoscopy
|
serum iron, serum ferritin
|
7
|
Karadag et al.2011
|
Case-control
|
63/50
|
Turkey
|
Female
|
29.1±11.9/28.4±9.4
|
TE, ATE, CTE
|
History, physical examination, dermoscopy
|
serum ferritin
|
7
|
Karim et al.2010
|
Case-control
|
30/30
|
Bangladesh
|
Female
|
25.4 ±7/24.8 ±5.6
|
TE
|
History, physical examination, pull test
|
serum ferritin
|
7
|
Moeinvaziri et al.2009
|
Case-control
|
30/30
|
Iran
|
Female
|
28.1±8.5/29.4±8.2
|
CTE
|
History, physical examination, trichogram
|
serum iron, serum ferritin, prevalence of ferritin deficiency
|
7
|
Naser et al.2021
|
Cross-sectional
|
60/60
|
Iraq
|
Female
|
32.6±6.47/41.3±4.59
|
CTE
|
History, physical examination, pull test
|
serum ferritin, prevalence of ferritin deficiency
|
7
|
Obaidat et al.2005
|
Case-control
|
72/30
|
Jordan
|
Female
|
26±9.2/32±11.6
|
CTE
|
History, physical examination, pull test
|
serum ferritin, prevalence of ferritin deficiency
|
7
|
Olsen et al.2010
|
Case-control
|
381/76
|
USA
|
Female
|
Range 18-65
|
TE
|
History, physical examination
|
serum ferritin, prevalence of ferritin deficiency
|
7
|
Pradhan et al.2018
|
Case-control
|
60/60
|
Nepal
|
Female
|
33.93±11.02/34.97±10.42
|
TE
|
History, physical examination
|
serum ferritin
|
7
|
Rasheed et al.2013
|
Case-control
|
80/40
|
Egypt
|
Female
|
29.8 ± 9.3/30.8 ± 8.56
|
TE
|
History, physical examination, pull test, trichogram, dermoscopy
|
serum ferritin
|
7
|
Sarkar et al.2013
|
Case-control
|
40/40
|
India
|
Female
|
24.28 ± 6.17/23.11±4.67
|
CTE
|
History, physical examination, pull test, trichogram
|
serum ferritin, prevalence of ferritin deficiency
|
7
|
Ullah et al.2019
|
Case-control
|
50/50
|
Pakistan
|
Female
|
26±7.61/31.88±6.62
|
TE
|
History, physical examination, pull test
|
serum ferritin
|
7
|
Yavuz et al.2018
|
Case-control
|
40/30
|
Turkey
|
NA/Female
|
NA
|
CTE
|
NA
|
serum iron
|
7
|
Abbreviations: P, patients; C, control; TE, telogen effluvium; ATE, acute telogen effluvium; CTE, chronic telogen effluvium; NA, not applicable.
Main results
Serum ferritin level in telogen effluvium
Ten studies measured serum ferritin level in telogen effluvium[27,28,30,31,34,35,37,39,43,44]. Significant heterogeneity was observed (P < 0.00001, I2 = 93%), and a random-effects model was applied. These results showed that serum ferritin levels were lower in the telogen effluvium group than in the control group (P < 0.00001, SMD = −1.09, 95% CI = −1.50 to −0.69). Four studies measured serum ferritin level in acute telogen effluvium[26,28-30]. Heterogeneity was observed (P = 0.04, I2 = 63%). A random-effects model was used. The acute telogen effluvium group had a lower serum ferritin level than the control group (P = 0.03, SMD = −0.53, 95% CI = −1.00 to −0.06). Nine studies measured serum ferritin level in chronic telogen effluvium[26,28-30,32,33,36,40,42]. Heterogeneity was observed (P = 0.0001, I2 = 75%). A random-effects model was used. The chronic telogen effluvium group had lower serum ferritin level than the control group (P < 0.0001, SMD = −0.73, 95% CI = −1.06 to −0.40). Serum ferritin levels were lower in patients with different types of telogen effluvium than in the normal population. The results are shown in Figure 2.
Serum iron level in telogen effluvium
Three studies had a measurement of serum iron level in telogen effluvium[27,28,39]. Heterogeneity was found (P < 0.00001, I2 = 93%), and the random-effects model was applied. These results showed that serum iron level was reduced in the telogen effluvium compared to the control group (P = 0.03, SMD =−1.09, 95% CI =−2.06 to −0.11). Only two studies had a measurement of serum iron level in acute telogen effluvium[28,29]. No significant heterogeneity was observed (P = 0.52, I2 = 0%). To maintain the consistency of the results, we used the random-effects model. The acute telogen effluvium group had significantly lower serum iron level than the control group (P = 0.001, SMD = −0.77, 95% CI = −1.25 to −0.30). We also switched the fixed-effects model analysis, which is consistent with the random-effects model. Three studies had a measurement of the serum iron level in chronic telogen effluvium[28,29,32]. Heterogeneity was observed (P = 0.04, I2 = 68%). A random-effects model was applied. The chronic telogen effluvium group had significantly lower serum iron level than the control group (P = 0.005, SMD = −0.84, 95% CI = −1.42 to −0.26). In patients with different types of telogen effluvium, serum iron levels were lower than those in the normal population. The results are shown in Figure 3.
Serum ferritin and iron levels in acute and chronic telogen effluvium
Four studies had a measurement of serum ferritin level in acute and chronic telogen effluvium[26,28-30]. Only two studies had a measurement of serum iron level in acute and chronic telogen effluvium[28,29]. No significant heterogeneity was observed, and the fixed-effects model was applied. There was no significant difference in serum ferritin level between acute and chronic telogen effluvium (P = 0.62, SMD = 0.08, 95% CI = −0.23–0.38), nor were serum iron levels (P = 0.26, SMD = 0.26, 95% CI = −0.19–0.72). The results are shown in Figure 4.
Prevalence of ferritin deficiency and telogen effluvium
Six studies on chronic telogen effluvium have reported the prevalence of ferritin deficiency, which has not been publicly reported in patients with acute telogen effluvium. Due to the small number of included studies, we directly used a random-effects model for analysis. Three studies reported the prevalence of ferritin deficiency and chronic telogen effluvium with ferritin level below 10–15 μg/L (ng/mL)[25,32,41]. There was no significant difference between the chronic telogen effluvium and control groups (P = 0.51, SMD = 1.57, 95% CI = 0.41–5.95). Two studies reported the prevalence of ferritin deficiency and chronic telogen effluvium with ferritin levels of 20 μg/L[33,36], 30 μg/L[32,42], 40 μg/L[25,41], and ≤ 70 μg/L[41,42]. The prevalence of ferritin deficiency in the chronic telogen effluvium group was significantly higher than that in the control group at serum ferritin levels of 20 μg/L (P = 0.0002, SMD = 4.32, 95% CI = 1.99–9.38) and 30 μg/L (P < 0.00001, SMD = 18.62, 95% CI = 8.28–41.90). No significant difference was observed between the chronic telogen effluvium and control groups with ferritin levels of 40 μg/L (P = 0.43, SMD = 1.22, 95% CI = 0.74–2.00) and ≤ 70 μg/L (P = 0.54, SMD = 1.97, 95% CI = 0.23–16.74). In patients with chronic telogen effluvium, the prevalence of iron deficiency was higher than that in the general population with ferritin levels at thresholds of 20 μg/L and 30 μg/L, and the difference was statistically significant. The results are shown in Figure 5.
Sensitivity analysis
The contribution of each study to the pooled estimate was evaluated to assess the sensitivity of the analyses. Yavuz et al. had an undue influence on the summary SMD under the serum iron level in acute telogen effluvium[38]. This was excluded from the analysis. Other studies did not substantially change the pooled point estimates. We also converted the random- and fixed-effects models in all studies. No substantial changes were observed in these data. These results confirm the reliability of the present study.
Publication bias
Funnel plots were used to evaluate the publication bias. In the study of serum ferritin and iron levels in patients with telogen effluvium, the funnel plot was symmetrical, suggesting that the overall publication bias was small (Figure 6). In the studies on the prevalence of serum ferritin deficiency, because only two to three studies were included in each subgroup, the number was relatively small, and no further publication bias testing was performed.