Abstract
Background
Telogen effluvium (TE) is characterized by diffuse hair loss following stressful events such as childbirth, prolonged surgery or anesthesia, and severe febrile illnesses, as well as intentional or unintentional rapid weight loss. However, literature regarding the relationship between TE and weight loss is limited.
Objective
To describe the clinical characteristics of TE induced by weight loss and assess their quantitative relationship.
Methods
This retrospective study was performed on patients diagnosed with TE induced by weight loss in our institution, between June 2006 and March 2021. Information on age, sex, weight before weight loss, reduced weight, and weight loss period was collected.
Results
This study enrolled 140 patients with TE, of whom 30 were men and 110 were women, with a mean age of 34.57±13.89 years. The mean weight loss percentage was 15.21±7.18% and mean weight loss rate was 3.54±2.85 kg/months. The mean weight loss rate in men (5.03±3.92 kg/month) was significantly faster than that in women (3.14±2.35 kg/month) (p=0.026). The mean weight loss percentage was mostly higher in 10’s and became significantly smaller with age (p=0.004). There were no significant differences by the causes of weight loss.
Conclusion
TE induced by weight loss occurred at a mean weight loss percentage and mean weight loss rate of approximately 15.21% and 3.54 kg/months, respectively. Women and older adults are especially vulnerable to TE, even if the degree of weight loss is not more severe than in men and young adults.
Keywords: Alopecia, Hair, Weight loss
INTRODUCTION
Hair is an essential component of the skin and is of great cosmetic importance. Because it is associated with beauty and self-image, the loss of scalp hair can cause significant distress to patients. The medically significant and actionable causes of diffuse hair loss include diffuse-type alopecia areata, anagen effluvium, and telogen effluvium (TE)1.
TE was first described in 1961 by Albert Kligman2 who studied the response of hair follicles to physiological stress. Unlike the hairs on our bodies (and mammalian fur), which are mostly in the telogen (rest) phase, our scalp hairs are mostly in the anagen (growth) phase (approximately 90%)1. Kligman2 suggested that following a stressful event, anagen hair follicles of the human scalp undergo acceleration of their hair cycle and enter the telogen phase prematurely (through unknown mechanisms). This leads to diffuse hair loss about 2–3 months after the precipitation event, which prompts the patient to seek medical advice. The list of stressful events for TE includes childbirth, prolonged surgery or anesthesia, severe febrile illnesses, acute psychiatric conditions, acute physical or emotional trauma, and idiopathic causes2. Since then, several drugs, especially mepesulfate3, oral contraceptives4, triparanol5, and thioureas6, have been reported to induce TE. Fortunately, the acute form of TE usually resolves spontaneously with full hair regrowth.
Intentional or unintentional rapid weight loss is another important cause of TE. Several case reports have described TE induced by weight loss7. The degree of TE induced by weight loss may vary depending on the weight loss. Massive weight loss is likely to incur a pronounced TE, whereas a person with less weight loss may have a milder TE that is less marked, and therefore, less noticeable hair loss. However, there is currently a lack of literature on the quantitative relationship between TE and weight loss, such as how much weight loss is necessary or whether rapid weight loss is crucial. This study aimed to describe the clinical characteristics of TE induced by weight loss and to assess the quantitative relationship between them.
MATERIALS AND METHODS
Study design
This retrospective analysis of electronic medical records included all patients who underwent intentional or unintentional weight loss with a diagnosis of TE in our institution, between June 2006 and March 2021.
The clinical and demographic information for each patient was recorded, including age, sex, weight before weight loss, reduced weight, and period of weight loss. The possible causes of weight loss, duration from weight loss to hair loss, and period of recovery from hair loss were recorded. This study was approved by the Institutional Review Board (IRB) of Kyung Hee University Hospital in Gangdong (IRB No. 2023-12-019).
Outcome
The primary outcome measure was the amount of weight loss that caused hair loss. The percentage of weight loss was calculated as follows: (Reduced Weight)/(Weight Before Weight Loss) × 100 (%). The rate of weight loss was calculated as follows: (Reduced Weight)/(Weight Loss Period) (kg/month). In addition, we aimed to find a correlation between the degree of weight loss and the onset of hair loss or recovery period. The secondary outcome was to identify whether there was a difference in the degree of weight loss, duration from weight loss to hair loss and recovery period from hair loss according to sex, age, and cause of weight loss.
Statistical analysis
Continuous data are presented as mean ± standard deviation. Descriptive statistical methods were used for the variables. Student’s t-test for paired data was used to compare continuous variables and the χ2 test was used for qualitative variables. A 2-sided p-value <0.05 was considered statistically significant. All analyses were performed using the SPSS version 20 for Windows (IBM Corp., Armonk, NY, USA).
RESULTS
Demographics and clinical characteristics
In total, 444 patients were diagnosed with TE during the study period. Among them, patients who underwent chemotherapy, were pregnant, had a chronic disease, or had an operation history without weight loss were excluded. Ultimately, 140 patients with TE were enrolled in this study, of whom 30 were men and 110 were women, with a mean age of 34.57±13.89 years (range, 15–80 years) (Table 1). Of the 140 patients, 66 patients (47.1%) underwent weight loss on purpose. Thirteen patients (9.3%) had several underlying diseases, such as diabetes mellitus, herniated intervertebral disk, ulcerative colitis, and malignancy, including gastric cancer, colon cancer, thyroid cancer, and gallbladder cancer. At last, 11 patients (7.9%) had operative history before diffuse hair loss. Considering the duplicated cases among underlying diseases and operative histories, 20 patients were identified as experiencing hair loss due to unintentional weight loss. The cause of weight loss was not recorded for the remaining 54 (38.6%) patients.
Table 1. Demographic and clinical characteristics of patients.
| Characteristics | All (n=140) | |||
|---|---|---|---|---|
| Age (yr) | 34.57±13.89 (15–80) | |||
| Sex | ||||
| Male | 30 (21.4) | |||
| Female | 110 (78.6) | |||
| Cause of weight loss | ||||
| Diet | 66 (47.1) | |||
| Underlying disease | 13 (9.3) | |||
| Malignancy | 5 | |||
| Gastric cancer | 1 | |||
| Colon cancer | 1 | |||
| Thyroid cancer | 2 | |||
| Gallbladder cancer | 1 | |||
| Diabetes mellitus | 2 | |||
| Herniated intervertebral disk | 2 | |||
| Ulcerative colitis | 1 | |||
| Dental disease | 1 | |||
| Polycystic ovary syndrome | 1 | |||
| Pharyngitis | 1 | |||
| Operative history | 11 (7.9) | |||
| Gastrectomy | 3 | |||
| Colectomy | 2 | |||
| Thyroidectomy | 2 | |||
| Extended cholecystectomy | 1 | |||
| Kidney transplant | 1 | |||
| Hip arthroplasty | 1 | |||
| Unknown | 1 | |||
| Unknown cause | 54 (38.6) | |||
| Degree of weight loss | ||||
| Percentage* (%) | 15.21±7.18 (3.85–35.56) | |||
| Rate† (kg/month) | 3.54±2.85 (0.29–15) | |||
| Duration from weight loss to hair loss (mo) | 1.12±1.36 (0.0–6.0) | |||
| Recovery period (mo) | 4.83±3.04 (0.5–16.0) | |||
Values are presented as number (range) or number (%).
*Percentage of Weight Loss = (Reduced Weight)/(Weight Before Weight Loss) × 100.
†Rate of Weight Loss = (Reduced Weight)/(Weight Loss Period).
Primary outcome
The mean percentage of weight loss was 15.21±7.18% (range, 3.85–35.56%) and the mean rate of weight loss was 3.54±2.85 kg/month (range, 0.29–15 kg/month). The patients began to lose their hair an average of 1.12 months after undergoing weight loss, and all patients improved by an average of 4.83 months during follow-up without any treatment. However, whether the patients gained weight was not observed again. There was also no significant correlation between the degree of weight loss and the onset of hair loss or recovery period.
Comparison of the degree of weight loss
1) By sex
Of the 140 patients, 30 were men and 110 were women. The mean age was 31.33±9.40 years for men and 35.45±14.79 years for women (Table 2). The mean percentage of weight loss in men (16.45±7.01%) was higher than that in women (14.87±7.19%), but it was not significant. However, the mean rate of weight loss in men (5.03±3.92 kg/month) was significantly faster than that in women (3.14±2.35 kg/month) (p=0.016). There was no significant difference between the 2 groups in the duration from weight loss to hair loss or recovery period.
Table 2. Comparison of the degree of weight loss according to sex.
| Characteristics | Male (n=30) | Female (n=110) | p-value | |
|---|---|---|---|---|
| Age (yr) | 31.33±9.40 | 35.45±14.79 | 0.068 | |
| Degree of weight loss | ||||
| Percentage† (%) | 16.45±7.01 | 14.87±7.19 | 0.286 | |
| Rate‡ (kg/month) | 5.03±3.92 | 3.14±2.35 | 0.016* | |
| Duration from weight loss to hair loss (mo) | 0.98±1.28 | 1.16±1.39 | 0.536 | |
| Recovery period (mo) | 5.00±2.66 | 4.78±3.15 | 0.712 | |
Values are presented as mean ± standard deviation.
*p<0.05.
†Percentage of Weight Loss = (Reduced Weight)/(Weight Before Weight Loss) × 100.
‡Rate of Weight Loss = (Reduced Weight)/(Weight Loss Period).
2) By age
The number of patients according to the age group of the 140 patients is shown in Table 3. Except for a group in their 80s with only one patient, the mean percentage of weight loss in each age groups (10’s to 70’s) were 18.39±7.76, 16.75±7.09, 16.65±7.66, 12.09±5.20, 10.22±5.89, 11.25±3.59, and 9.24±4.02%, respectively. In patients in their 10s, the mean percentage of weight loss was mostly higher and became significantly smaller with age (p=0.004). The mean rate of weight loss in each age groups (10’s to 70’s) were 3.74±1.33, 3.44±2.59, 4.37±4.02, 3.27±2.38, 1.88±1.14, 1.54±0.82, and 4.33±4.04 kg/month, respectively; however, there was no significant difference (p=0.170). In addition, there was no significant difference between the groups in the duration from weight loss to hair loss or recovery period.
Table 3. Comparison of the degree of weight loss according to age.
| Characteristics | 10’s (n=16) | 20’s (n=45) | 30’s (n=34) | 40’s (n=27) | 50’s (n=10) | 60’s (n=4) | 70’s (n=3) | 80’s (n=1) | p-value | |
|---|---|---|---|---|---|---|---|---|---|---|
| Degree of weight loss | ||||||||||
| Percentage† (%) | 18.39±7.76 | 16.75±7.09 | 16.65±7.66 | 12.09±5.20 | 10.22±5.89 | 11.25±3.59 | 9.24±4.02 | 13.46 | 0.004* | |
| Rate‡ (kg/month) | 3.74±1.33 | 3.44±2.59 | 4.37±4.02 | 3.27±2.38 | 1.88±1.14 | 1.54±0.82 | 4.33±4.04 | 7.00 | 0.170 | |
| Duration from weight loss to hair loss (mo) | 1.08±1.44 | 0.93±1.15 | 1.07±1.33 | 1.08±1.38 | 1.56±1.59 | 2.67±3.06 | 1.33±1.16 | 1.00 | 0.592 | |
| Recovery period (mo) | 5.85±3.93 | 4.51±3.12 | 5.02±3.35 | 5.04±1.99 | 4.00±3.54 | 3.50±1.73 | 4.50±2.12 | 7.00 | 0.772 | |
Values are presented as mean ± standard deviation.
*p<0.05.
†Percentage of Weight Loss = (Reduced Weight)/(Weight Before Weight Loss) × 100.
‡Rate of Weight Loss = (Reduced Weight)/(Weight Loss Period).
3) By cause
With the exception of 54 patients whose causes were not clearly stated, patients who lost weight for their own purpose through diet were classified into the intentional group (n=66), and those who lost weight due to chronic disease or surgery were classified into the unintentional group (n=20, considering duplicated causes) (Table 4). The mean age of patients in the intentional group (29.97±10.47 years) was significantly younger than that in the unintentional group (44.70±18.74 years) (p=0.003). The mean percentage of weight loss in the intentional group (17.09±7.39%) was higher than that in the unintentional group (14.34±8.04%), but it was not significant and was thought to be due to age differences. In addition, there were no significant differences in the degree of weight loss, duration from weight loss to hair loss and recovery period from hair loss between the groups.
Table 4. Comparison of the degree of weight loss according to cause.
| Characteristics | Intentional (n=66) | Unintentional (n=20) | p-value | |
|---|---|---|---|---|
| Age (yr) | 29.97±10.47 | 44.70±18.74 | 0.003* | |
| Sex | 0.590 | |||
| Male | 14 (21.2) | 4 (20) | ||
| Female | 52 (78.8) | 16 (80) | ||
| Degree of weight loss | ||||
| Percentage† (%) | 17.09±7.39 | 14.34±8.04 | 0.182 | |
| Rate‡ (kg/month) | 4.18±3.35 | 4.06±3.26 | 0.893 | |
| Duration from weight loss to hair loss (mo) | 0.95±1.18 | 1.23±1.18 | 0.412 | |
| Recovery period (mo) | 5.25±3.58 | 4.25±2.17 | 0.139 | |
Values are presented as number (%) or mean ± standard deviation.
*p<0.05.
†Percentage of Weight Loss = (Reduced Weight)/(Weight Before Weight Loss) × 100.
‡Rate of Weight Loss = (Reduced Weight)/(Weight Loss Period).
DISCUSSION
TE is a common cause of diffuse, non-scarring hair loss that can be triggered by various factors. In 1993, Headington8 classified TE according to 5 different mechanisms as follows: (1) immediate anagen release, (2) delayed anagen release, (3) short anagen release, (4) immediate telogen release, and (5) delayed telogen release. Immediate anagen release is the most common type, and follicles that would normally complete a longer cycle by remaining in the anagen phase, prematurely enter the telogen phase. In delayed anagen release, some follicles remain in the prolonged anagen phase rather than normally cycling into the telogen phase, and when they are finally released from the anagen phase, clinical signs of increased shedding are observed. Short anagen phase is associated with decreased hair length. In immediate telogen release, hair follicles normally programmed for the release of club hair after an interval of usually 100 days after the end of the anagen phase are prematurely stimulated to cycle into the anagen phase. Finally, in delayed telogen release, hair follicles remain in the prolonged telogen phase rather than being shed and recycled into the anagen phase; when teloptesis finally sets in, increased shedding of club hair is observed.
Several reports have described that postfebrile, postpartum, psychogenic, chronic disease and certain drugs can cause TE2-6. Moreover, there has been a case series review of TE as a side effect of dramatic weight loss. Goette and Odom7 observed several patients with TE subsequent to dramatic weight loss secondary to a diet regimen started 2 to 5 months previously. In their study, 9 patients experienced profuse hair loss 2 to 5 months after starting a vigorous weight reduction program resulting in weight loss of 11.7–24.75 kg and telogen counts of 25%–50% were observed. However, no large studies have investigated TE induced by weight loss. Therefore, we retrospectively examined 140 patients to determine the association between TE and weight loss.
In the case series of Goette and Odom7, 9 patients were observed, comprising 8 women and one man. Interestingly, compared to our study, both studies found that women had a higher incidence of TE, which may be because women are more interested in managing their appearance or weight. Goette and Odom7 reported that the range of weight loss was 11.7–24.75 kg and they did not mention the original weight of the patients. Therefore, we confirmed the degree of weight loss using the original weight and duration of weight loss by calculating the percentage and rate of weight loss. The mean percentage of weight loss was 15.21±7.18% and the mean rate of weight loss was 3.54±2.85 kg/month. As a result, if someone loses approximately 15% of his or her weight or lose 3.5 kg per month, TE can develop. We also compared the degree of weight loss according to sex, age, and cause. Based on sex, there was no significant difference between the mean percentage of weight loss, but the mean rate of weight loss in men (5.03±3.92 kg/months) was significantly faster than that in women (3.14±2.35 kg/months). In other words, women are more vulnerable to TE than men even if they lose less weight per month. In addition, our study identified that the mean percentage of weight loss was mostly higher in patients in their 10s and became significantly lower with age. In other words, older adults can experience TE even if they do not lose much weight. However, in the case of the elderly, it is difficult to distinguish whether this was due to weight loss or a natural phenomenon of aging. Finally, between the intentional and unintentional groups, we confirmed that there was no significant difference in the degree of weight loss, onset of hair loss, or recovery period. However, interestingly, the mean age of patients in the intentional group (29.97±10.47 years) was significantly younger than that in the unintentional group (44.70±18.74 years). In other words, the younger the patient, the more likely they are to be diagnosed with TE by intentionally dieting and losing weight.
Goette et al.7 suggested that rigorous caloric restriction with a subsequent inadequate energy supply to the hair matrix is the cause of the precipitation of TE of the crash dieter. We also agree that the cause for the precipitation of TE is probably not direct weight loss but rather the restriction of caloric intake during dieting. Bullough9 and Bullough and Eisa10 demonstrated that epidermal mitotic activity is decreased by underfeeding and nearly ceases with chronic inhibition in mice. Because the hair matrix is known to have the highest rate of cell turnover in human tissues, it is not surprising that a drastic reduction in caloric intake might adversely affect the cellular kinetics of the hair matrix and precipitate the catagen phase. During the ensuing telogen phase, resting hair in the affected follicles is gradually shed. This hypothesis appears to be supported by the observation of gradual regrowth of normal healthy hair after normalization of caloric intake and maintenance of reduced body weight.
A limitation of this study is its retrospective nature. Other medical histories that could have caused diffuse hair loss were not recorded properly. Also, since the degree of weight loss was recorded based on the patient’s memories and statements, there is a possibility of recall bias. Furthermore, because the grade of hair loss was not recorded, we obtained the mean percentage and rate of weight loss; however, we could not determine whether hair loss occurred more severely as the degree of weight loss increased. Finally, most patients showed an improvement in TE without treatment, but there is a lack of data on whether TE improved with or even without weight gain.
In conclusion, TE induced by weight loss occurred at a mean percentage of weight loss of approximately 15.21% or a mean rate of weight loss of approximately 3.54 kg/month. Women and older adults are especially vulnerable to TE, even if the degree of weight loss is not more severe than in men and young adults. Therefore, evaluation of weight loss is essential for patients complaining of diffuse hair loss, and TE should always be considered for accurate diagnosis and proper treatment selection, even if weight loss is not severe in women and elderly individuals.
Footnotes
FUNDING SOURCE: This study was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (No. 2019R1A2C1089788).
CONFLICTS OF INTEREST: The authors have nothing to disclose.
DATA SHARING STATEMENT: The data that support the findings of this study are available from the corresponding author, upon reasonable request.
References
- 1.Chien Yin GO, Siong-See JL, Wang ECE. Telogen effluvium - a review of the science and current obstacles. J Dermatol Sci. 2021;101:156–163. doi: 10.1016/j.jdermsci.2021.01.007. [DOI] [PubMed] [Google Scholar]
- 2.Kligman AM. Pathologic dynamics of human hair loss. I. Telogen effuvium. Arch Dermatol. 1961;83:175–198. doi: 10.1001/archderm.1961.01580080005001. [DOI] [PubMed] [Google Scholar]
- 3.Field JB, Attyah AM, Ramsay GD, Levitt H. The clinical intoxication induced by a heparinoid. Am J Med Sci. 1961;241:637–644. doi: 10.1097/00000441-196105000-00013. [DOI] [PubMed] [Google Scholar]
- 4.Griffiths WA. Diffuse hair loss and oral contraceptives. Br J Dermatol. 1973;88:31–36. doi: 10.1111/j.1365-2133.1973.tb06668.x. [DOI] [PubMed] [Google Scholar]
- 5.Winkelmann RK, Perry HO, Achor RW, Kirby TJ. Cutaneous syndromes produced as side effects of triparanol therapy. Arch Dermatol. 1963;87:372–377. doi: 10.1001/archderm.1963.01590150088016. [DOI] [PubMed] [Google Scholar]
- 6.Lindback K. Toxic, allergic and myxedematoid symptoms in the treatment of thyrotoxicosis with antithyroid substances. Acta Med Scand. 1946;124:265–281. [Google Scholar]
- 7.Goette DK, Odom RB. Alopecia in crash dieters. JAMA. 1976;235:2622–2623. [PubMed] [Google Scholar]
- 8.Headington JT. Telogen effluvium. New concepts and review. Arch Dermatol. 1993;129:356–363. doi: 10.1001/archderm.129.3.356. [DOI] [PubMed] [Google Scholar]
- 9.Bullough WS. The effect of a restricted diet on mitotic activity in the mouse. Br J Cancer. 1949;3:275–282. doi: 10.1038/bjc.1949.31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Bullough WS, Eisa EA. The effects of a graded series of restricted diets on epidermal mitotic activity in the mouse. Br J Cancer. 1950;4:321–328. doi: 10.1038/bjc.1950.30. [DOI] [PMC free article] [PubMed] [Google Scholar]