Objective: To study the clinical effect of nano-fat mixed granule fat transplantation in the treatment of cicatricial facial depression and atrophy, and to explore the related experimental mechanism. Methods: (1) From January 2012 to April 2018, 105 patients conforming to the inclusion criteria, with cicatricial facial depression and atrophy deformity who needed facial fat transplantation, were admitted to our unit. Their medical records were analyzed retrospectively. According to the patients′ wishes, 54 patients (12 males and 42 females) aged 10-59 years received traditional transplantation of pure autologous granule fat collected from abdomen/thigh and were included in simple transplantation group, while the other 51 patients (14 males and 37 females) aged 7-63 years received transplantation of autologous nano-fat mixed granule fat collected from abdomen/thigh and were included in mixed transplantation group. The treatment satisfaction of patients 3 and 6 months after operation was assessed by the facial fullness, symmetry, scar, and complications using self-made scales and photo data before and after operation. Six months after operation, the patients were assessed whether they needed to undergo a second operation, and the rate of second operation was calculated. During the second operation, the fat of patients transplanted in the first operation was collected, and the morphology of adipocytes and microangiogenesis was observed under a scanning electron microscope. (2) Adipose-derived stem cells (ADSCs) were isolated and cultured from abdominal fat of a 4-week-old male Sprague-Dawley (SD) rat. The 5th passage of cells were selected to observe cell morphology after cultured for 14 days, observe expression of vimentin and cytokeratin-18 by immunofluorescence method, identify osteogenic and adipogenic differentiation, and detect rates of CD29 and CD44 positive cells by flow cytometer (n=3). Eighteen 4-week-old male SD rats were divided into ADSCs transplantation group, simple scar group, and blank control group according to the random number table, with 6 rats in each group. Rats in ADSCs transplantation group and simple scar group were subcutaneously injected with 1 mL bleomycin which was dissolved in phosphate buffered saline (PBS) with a mass concentration of 1 mg/mL at the back to establish scar models. After 3 hours, rats in ADSCs transplantation group were injected with 1×10⁶ ADSCs suspended in 0.1 mL PBS at the same injection site, while rats in simple scar group were injected with 0.1 mL PBS. Rats in blank control group were injected with the same doses of PBS in the same place at the same two time points mentioned above. After continuous injection for 28 days in each group, the full-thickness skin tissue of the injected area of all rats was collected to observe the collagen fibers by Masson staining and expressions of α-smooth muscle actin (α-SMA) and transforming growth factor β₁ (TGF-β₁) by immunohistochemistry, and the positive cells were counted. Data were processed with Mann-Whitney U test, χ² test, one-way analysis of variance, and least significant difference test. Results: (1) Compared with the preoperative condition, the facial fullness and symmetry of patients in simple transplantation group were better in 3 months after operation, with scar color closer to the surrounding skin, and the filling volume of patients in this group decreased in 6 months after operation as compared with that in 3 months after operation. In mixed transplantation group, the facial fullness and symmetry of patients were better in 3 and 6 months after operation as compared with the preoperative condition, with scar color and texture closer to the surrounding skin, and the filling volume in 6 months after operation was not obviously reduced as compared with that in 3 months after operation. Fat liquefaction and subcutaneous nodule formation occurred respectively in 1 patient in simple transplantation group within 3 months after operation. The treatment satisfaction of patients in mixed transplantation group was significantly higher than that in simple transplantation group in 3 and 6 months after operation (Z=-2.566, -3.084, P<0.05 or P<0.01). Six months after operation, the second operation rate of patients in mixed transplantation group was 7.84% (4/51), which was significantly lower than 22.22% (12/54) in simple transplantation group (χ²=4.199, P<0.05). At the second operation, compared with those of simple transplantation group, the cells of fat transplanted in the first operation of patients in mixed transplantation group were more plump, without collapse or dryness, and the cells were closely arranged, with smaller gap; the tubular and the cord-like microvascular structure on the cell surface were more abundant, and the cell gap was full of network-like microvascular structure that grew into the adipose tissue. (2) The fifth passage of cells isolated and cultured from rat fat grew adherently to the wall, with long fusiform or spindle shape, showing shoal-of-fish-like growth. Vimentin and cytokeratin-18 were highly expressed in the cells. Cells showed osteogenic and adipogenic differentiation ability by induction. The positive expression rates of CD29 and CD44 were higher than 90.00%. The cells were identified as ADSCs. After 28 days of injection, the collagen fibers in the dermis of skin tissue at the injection area of rats in blank control group were finely arranged. In simple scar group, a large amount of collagen was deposited in the dermis of skin tissue at the injection area of rats, the fiber bundles were thick and loosely unevenly arranged, and a large number of inflammatory infiltration and scattered muscle fibers were observed. In ADSCs transplantation group, the collagen fibers in the dermis of skin tissue at the injection area of rats were thicker than those of blank control group, with still neat arrangement, and a small amount of scattered muscle fiber and inflammatory infiltration was observed. After 28 days of injection, the expression of α-SMA in ADSCs transplantation group was mainly in microvessels in the dermis of skin tissue at the injected area of rats, and the number of α-SMA and TGF-β₁ positive cells was (49±12) and (63±10) cells per 20-fold field of view, respectively, which was similar to (35±16) and (44±17) cells per 20-fold field of view of blank control group (P>0.05), all significantly less than (135±13) and (121±23) cells per 20-fold field of view of simple scar group (P<0.05). Conclusions Compared with those of autologous simple granule fat transplantation, autologous nano-fat mixed granule fat transplantation has better filling fullness in the treatment of patients with scar facial depression and atrophy. The filling effect lasts longer, and the improvement of scar texture is more obvious. As showed in the rat scar model experiment, the mechanism may be that ADSCs inhibit the expressions of α-SMA and TGF-β₁, thus inhibiting the formation of scar.