Topographic and morphometric characterization of the humerii nutrient foramina and their implications for fracture repair and healing

Abstract

Fractures of the humerus are common on the midshaft of the bone, often causing injury to the nutrient artery. Successful fracture repair and healing requires preservation of the blood supply to the long bones which is conveyed through the nutrient foramina (NF). The topography of long bone NF varies in different populations. These variations can affect the preservation of blood supply to long bones during fracture repair management. The current study aimed to determine the topography and morphometry of the NF of the humerus in different populations of South Africa including the South African Africans (SAA), South Africans of European descent (SAED), and South Africans of mixed ancestry (SAMA). The study examined 596 dry humerii from the three South African populations, sourced from Raymond A. Dart Collection of Modern Human Skeletons. The parameters examined included the presence, number, location, position, size and direction of the NF, and foramina index (FI). The NF were present in 97.1% of the humerii. Majority of bones (76.8%) evinced a single NF with a diameter equal to or larger than 1.27 mm. The number of NF varied across the different population groups (p = 0.000), with SAA having more humerii presenting with a single NF and SAED having more humerii with two NF. The position of NF varied within and across populations (p = 0.002). Males in SAED had a higher mean FI on both the right (p = 0.030) (effect size = 0.258) and left (p = 0.022) (effect size = 0.421) sides than females. SAED had a lower mean FI than SAMA (p = 0.002) (effect size = 0.384). The location of NF varied across different populations (p = 0.000), with SAA having more NF located on the anteromedial surface and medial border, and SAED having more NF located on the lateral border (p = 0.000). NF were directed towards the distal ends of the shafts in 99.8% of bones and towards the proximal end in 0.2% of bones. The topography and morphometry of the nutrient foramina of the humerus are variable in the South African populations. Knowledge of the NF variations may aid in the management of humerus fractures.

Introduction

Humeral fractures are a significant cause of the loss of function of the upper limb, leading to reduced quality of life, productivity, and lower earning potential¹. The humerus is mostly vulnerable during sports and automobile injuries, and overall, humeral fractures account for about 3% of all bone fractures in the human body^1,2. Most humeral fractures occur on and around the midpoint of the shaft with the possibility of damaging the nutrient artery^3,4. The humerus receives about 90% of its diaphyseal blood supply from a nutrient artery that enters the shaft through a nutrient foramen (NF) located just below the midpoint of the bone^5,6. The NF is an external opening to the nutrient canal that conveys nutrient vessels and nerves⁷. Besides its role in fracture healing and repair, the nutrient artery also assists in the bone metabolic processes, haematopoiesis, and the survival of the osteocytes and osteoblasts throughout life^8–11.

Fractures passing through the NF have a poor prognosis due to the disturbance of the blood supply to the long bone^4,12,13. Interrupted blood supplies can result in the deprivation of the osteoblasts and osteocytes of essential nutrients and oxygen leading to poor bone remodelling and growth and subsequent non-union of fractures. Most humeral fractures are treated surgically with high success rates¹⁴. However, non-unions still account for about 20–30% of cases^15–17. These arise due to surgical techniques which, most of the time, interfere with the nutrient artery and the foraminal area, leading to poor vascularisation of the fracture area⁵. Knowledge of the number, size, location, position, and direction of the NF is critical to the preservation of blood supply to the long bones during fracture repair^{3,12,13,18,19}.

The topographical anatomy of the NF varies in different populations, and these variations may present challenges in managing humeral fractures²⁰. While there have been few studies conducted on the topography of the NF on the humerus^5,6,10,12,18, there are no such reports on the humerus in South African populations. There is evidence that the skeletal biology of South Africans, like limb proportion ratios and foramina sizes, is different from other regions^21–23. Given that South Africa has three dominant populations that show considerable post-cranial skeletal variations^24,25, it is essential to investigate the topography of the NF in South Africans. The aim of the current study was to determine the topographical anatomy and morphometric variations of the humeral NF in South African Africans (SAA), South Africans of European descent (SAED), and South Africans of mixed ancestry (SAMA). Topographical and morphometric knowledge of the NF may aid in managing humeral fractures and lead to reduced cases of non-union fractures.

Materials and methods

Study sample, inclusion and exclusion criteria

This cross-sectional study was conducted on the dry humerus bones of individuals obtained from the Raymond A. Dart Collection of Modern Human Skeletons housed in the School of Anatomical Sciences, Health Science Faculty at the University of the Witwatersrand²⁶. The current study was undertaken in compliance with, and as per the requirements outlined within the National Health Act (No. 61 of 2003) on the use of human remains for research and teaching purposes. Ethical approval was granted through the Human Research Ethics Committee (medical): ethics waiver number W-CBP-220504-01 and all methods were performed in accordance with the relevant guidelines and regulations of this ethics committee. The specimens utilised in the current study were donated from both the government of South Africa and individual’s family members for the purposes of research and teaching and hence the informed consent and approval to use the skeletal remains was obtained from the University of the Witwatersrand Human Research Ethics Committee (medical). This study analysed a total of 596 humerii of known demographics (age, sex, side of the body, and population affinity) from adult skeletons (age range 21 to 65 years) representing three of the largest South African population groups. The current study population includes South African Africans (SAA, also called South African Black), South Africans of European descent (SAED, also called South African White), and South Africans of Mixed ancestry (SAMA, also called South African Coloured) individuals according to the official South African classification of these population groups and recent publication terminologies^7,27–33. The SAA population represent a diverse history constituting multiple tribes (e.g., Sesotho, Setswana, isiXhosa, and isiZulu) that are genetically closely related and all descendants of Bantu-speakers³⁴. The SAED population were migrants from Western Europe and identified to be morphologically distinct to South Africa when compared to other European nationalities²¹. The SAMA are a self-identified mixed ancestry population group that stems from a multitude of parental populations extending from but not limited to Africa (Khoe-San, Bantu-speakers), Asia, and Europe³⁵. Historically the colonisation of South Africa and the onset of Apartheid led to the sociocultural identities highlighted above mainly due to acts of endogamy exacerbated by the social and physical restrictions imposed during this period³⁶.

The sample consisted of 248 humerii of the SAA (118 females and 130 males), 210 humerii of the SAED (96 females and 114 males), and 138 humerii of the SAMA (56 females and 82 males). Humerii that presented with trauma, gross pathological deformities, congenital deformities, evidence of any genetic diseases, surgical interventions, and postmortem damage were excluded from the study.

Topography and morphometry of the humerus nutrient foramina

The nutrient foramina of the humerus were examined with the aid of a magnifying glass and identified as marked grooves with a foramen piercing the diaphysis¹⁰. All foramina that could admit a 25 gauge (0.515 mm outer diameter) hypodermic needle were considered viable, and were counted and recorded as nutrient foramina¹². The size of the nutrient foramen was then classified as dominant or accessory (secondary). A dominant-sized nutrient foramen was defined as any viable foramen that permits an 18 gauge (1.27 mm outer diameter) hypodermic needle and secondary-sized nutrient foramen was any viable foramen that did not allow the passage of the 18 gauge hypodermic needle¹². The direction of the nutrient foramen and its associated nutrient canal was confirmed by inserting a hypodermic needle through the foramen, and its orientation was noted as being directed toward the proximal or distal end of the humerus. Plain radiography was undertaken to confirm if the hypodermic needles were inserted into the nutrient foramina as previously described by Mazengenya and Billings³¹. The location of the dominant nutrient foramina was described according to the anatomical borders and surfaces of the diaphysis of the long bone in which they were identified. The borders and surfaces of the humerus were described according to Moore et al.³⁷ which include the anteromedial surface (AMS), anterolateral surface (ALS), posterior surface (PS), anterior border (AB), medial border (MB), and lateral border (LB). The position of the nutrient foramen was expressed as foraminal index (FI) calculated using the Hughes formula³⁸: FI = DNF∕TL×100, whereby DNF is the distance from the apex of the head of the humerus to the most distal edge of the dominant nutrient foramen measured using Vernier calipers (Mitutoyo Japan, accuracy = 0.02 mm, resolution = 0.01 mm, repeatability = 0.01 mm), TL is the total length of the humerus (measured from the most proximal aspect of the humeral head to the most distal point of the trochlea using a laboratory osteometric board (Paleo-Tech Concepts, accuracy = 0.02 mm, resolution = 0.01 mm, repeatability = 0.01 mm)).

Data management and statistical analysis

All data were managed in Microsoft Excel Office 365 (Microsoft Corporation) and analysed using Stata/IC software (Version 12). Intra- and interobserver reproducibility errors were calculated using the Lin’s concordance coefficient³⁹. For intraobserver error, the primary observer took all measurements from 10 randomly selected humeri (5 males, 5 females) and repeated the same measurements two weeks later. There was no significant difference between the two measuring occasions (RC = 1.000, 95%CI for RC = 1.000 upper and RC = 1.000 lower). The primary and secondary observers took the same measurements using 10 randomly selected humerii (5 males, 5 females) four weeks apart. No significant differences were observed between the two observers (RC = 0.999, 95%CI for RC = 0.976 upper and RC = 0.932 lower). The categorical data were presented as frequencies and percentages. Chi-square tests were used to test for association between nutrient foramina parameters (number, location, position, and direction) and sex, side of the body, and population affinity. The initial alpha value of ≤ 0.050 was utilised for the Chi-square test, which was adjusted using the Bonferroni correction method (p ≤ 0.007) for the location of dominant NF, and p ≤ 0.002 for the location of accessory NF). The morphometric data were presented as mean, range, and standard deviations. The Shapiro-Wilk test and Histograms were used to test the normality of the morphometric data. The morphometric data (FI) was parametric in the SAA and non-parametric in the SAED and SAMA. For the parametric data, Student’s t-test was used to determine significant differences in mean values for comparison of sidedness and biological sex (e.g., right side comparisons between males versus females). The Wilcoxon rank-sum (Mann-Whitney U) test was used to determine significant median differences between the sides of the body and sex for the non-parametric data. For significant differences between sides of the body or sexes, the effect size was calculated using Cohen’s d (effect sizes: small effect = 0.2, medium effect = 0.5, large effect ≥ 0.8). The Kruskal Wallis test was used to determine significant median differences across population groups. An initial alpha (p) value of ≤ 0.050 was utilised for the Mann-Witney U test, Student’s t-test, and Kruskal Wallis test. To identify differences among population groups, the Kruskal Wallis test was further interpreted using Bonferroni corrections for multiple comparisons, considered significant at a p-value of ≤ 0.017 due to the three pairwise comparisons conducted.

Results

Overall, the NF was present in 97.1% of humerii while the NF was not observed (absent) in 2.9%. The presence of the NF was 99.6% on the SAA humerii, 97.6% on the SAED humerii, and 92% on the SAMA humerii. There were no statistically significant differences in the presence of NF between sexes in the SAA (p = 0.293), SAED (p = 0.243), and SAMA (p = 0.731) population groups. There were no statistically significant differences in the presence of NF between sides in the SAA (p = 0.316), SAED (p = 0.174), and SAMA (p = 0.116) population groups.

The number of NF (dominant and accessory) did not vary between sexes in the SAA (p = 0.499), SAED (p = 0.337), and SAMA (p = 0.497). The number of NF (dominant and accessory) did not vary between sides of the body in the SAA (p = 0.418), SAED (p = 0.200), and SAMA (p = 0.244). Most (76.8%) of the humerii had a single NF. The number of NF showed variations across populations (p = 0.000). The SAA had more humerii with a single NF whereas the SAED had a higher number of humerii with two NF (Table 1). The highest number of NF observed was four and identified in an SAED female on the right humerus (Fig. 1).

Table 1.

The distribution of the number of the nutrient foramina on the humerus in South African populations.

Population	SAA					SAED					SAMA
Sex	Male		Female		Total n (%)	Male		Female		Total n (%)	Male		Female		Total n (%)
Side	R	L	R	L		R	L	R	L		R	L	R	L
No of NF
0	–	–	1	0	1 (0.4%)	3	1	1	–	5 (2.38%)	4	2	4	1	11 (7.97%)
1	56	53	52	50	211 (85.08%)	42	34	27	29	132 (62.86%)	34	34	22	25	115 (83.33%)
2	9	11	6	9	35 (14.11%)	12	21	19	18	70 (33.33%)	3	5	1	2	11 (7.97%)
3	–	1	–	–	1 (0.4%)	–	1	–	1	2 (0.95%)	–	–	1	–	1 (0.72%)
4	–	–	–	–	–	–	–	1	–	1 (0.48%)	–	–	–	–	–
Total sample	65	65	59	59	248 (100%)	57	57	48	48	210 (100%)	41	41	28	28	138 (100%)

–, Absent; SAA, South African Africans; SAED, South Africans of European descent; SAMA, South Africans of Mixed Ancestry; M, male; F, female; R, right; L, left.

Fig. 1.

A photograph of the right humerii showing the variations in the number of the nutrient foramina. Hypodermic needles inserted in the nutrient foramina indicate the viability and direction of the NF. (A) Shows a single nutrient foramen. (B) Shows two nutrient foramina. (C) Shows four nutrient foramina.

The foramina index (FI) of the NF humerus ranged from 33.03 to 80.63% in all populations combined (Fig. 2). The position of the NF showed variations across populations (p = 0.003). Across the populations, the SAED had a significant lower mean FI than the SAMA (p = 0.002) (effect size = 0.384), whereas the mean FI between the SAED and SAA (p = 0.141), and SAA and SAMA (p = 0.211) were not significant. In the SAED, males had a significant higher mean FI on both the right (p = 0.030) (effect size = 0.258) and left (p = 0.022) (effect size = 0.421) sides than in females (Fig. 3). There were no significant differences between mean FI in the SAA (p = 0.601) and SAMA (p = 0.384) populations with specific reference to sidedness and sex (Fig. 3).

Fig. 2.

A graph showing the distribution of nutrient foramina position range on the humerus in South African populations, based on foraminal indices (FI). The vertical lines on the graph indicate the complete range of FI values. Mean values for each group are represented by (times) for South African Africans (SAA), (filled black triangle) for South Africans of European descent (SAED), and (open circle) for South Africans of Mixed Ancestry (SAMA). A statistically significant mean FI difference of p = 0.002 is denoted by * in the SAED, effect size = 0.346.

Fig. 3.

The graph shows the distribution of Foraminal Index (FI) by side, population group and sex for different population groups of South Africa. Box-plots represent the distribution of the FI between sides and sex: diamond shape = outliers, p-values derived from Mann-Whitney comparison test.

Except for the humerii with absent NF, the remaining humerii (97.1%) had at least one dominant-sized NF (Table 2). The SAED had a significant higher number of accessory NF than the SAA (p = 0.000) and SAMA (p = 0.000). There was no significant difference in the number of accessory NF between the SAA and SAMA (Table 3).

Table 2.

The distribution of the location of the dominant and accessory nutrient foramina on the humerus in South African populations.

Population	Sex	Side	Dominant nutrient foramina							Accessory nutrient foramina
			NUL	AB	ALS	AMS	LB	MB	PS	NUL	AB	ALS	AMS	LB	MB	PS	LB, PS, AMS
SAA	M	R	–	–	2	38	3	21	1	56		1	4	–	3	1	–
		L	–	1	2	36	4	20	2	53	–	–	4	2	4	2	–
	F	R	1	1	1	40	–	16	–	53	–	–	4	1	–	1	–
		L	–	–	–	40	1	17	1	50	–	–	2	1	2	4	–
Total (n)			1	2	5	154	8	74	4	212	–	1	14	4	9	8	–
Population %			0.40	0.81	2.02	62.10	3.23	29.84	1.61	85.48	–	0.40	5.65	1.61	3.63	3.23	–
SAED	M	R	3	1	1	26	4	20	2	45	–	–	6	2	4	–	–
		L	1	2	1	31	4	18	–	35	–	2	12	4	2	2	–
	F	R	1	1	1	27	5	12	1	28	1	5	7	3	3	–	1
		L	–	1	3	30	6	5	3	29	–	–	14	2	2	1	–
Total (n)			5	5	6	114	19	55	6	137	1	7	39	11	11	3	1
Population %			2.38	2.38	2.86	54.29	9.05	26.20	2.86	65.23	0.48	3.33	18.57	5.24	5.24	1.43	0.48
SAMA	M	R	4	–	1	27	1	8	–	38	–	–	1	1	1	–	–
		L	2	–	2	33	1	3	–	36	–	–	4	1	–	–	–
	F	R	4	–	–	17	–	6	1	26	–	–	1	–	1	–	–
		L	1	1	–	20	–	6	–	26	–	1	1	–	–	–	–
Total (n)			11	1	3	97	2	23	1	126	–	1	7	2	2	–	–
Population %			7.97	0.72	2.17	70.29	1.45	16.67	0.72	91.30	–	0.72	5.07	1.45	1.45	–	–

–, Absent; SAA, South African Africans; SAED, South Africans of European descent; SAMA, South Africans of Mixed Ancestry; M, male; F, female; R, right; L, left; AMS, anteromedial surface; ALS, anterolateral surface; PS, posterior surface; AB, anterior border; MB, medial border; LB, lateral border. NUL = Absent nutrient foramina.

Table 3.

The adjusted statistical values following the Bonferroni correction for multiple comparisons of the location of the dominant and accessory nutrient foramina on the humerus in South African populations.

Surface	Dominant nutrient foramina							Accessory nutrient foramina
	NUL	AB	ALS	AMS	LB	MB	PS	NUL	AB	ALS	AMS	LB	MB	PS	LB, PS, AMS
Population
SAA	0.002	0.337	0.651	0.718	0.116	0.040	0.722	0.000	0.398	0.061	0.003	0.125	0.946	0.035	0.398
SAED	0.610	0.104	0.546	0.010	0.001*	0.777	0.176	0.000	0.175	0.007	0.000*	0.010	0.140	0.577	0.175
SAMA	0.000	0.472	0.877	0.013	0.033	0.007*	0.264	0.039	0.583	0.388	0.026	0.259	0.111	0.066	0.583

–, Absent; SAA, South African Africans; SAED, South Africans of European descent; SAMA, South Africans of Mixed Ancestry; M, male; F, female; R, right; L, left; AMS, anteromedial surface; ALS, anterolateral surface; PS, posterior surface; AB, anterior border; MB, medial border; LB, lateral border. NUL = Absent nutrient foramina. Adjusted significant statistical difference at p ≤ 0.007 for dominant nutrient foramina, and p ≤ 0.002 for accessory nutrient foramina are denoted by *.

The location of the NF did not vary between sexes in the SAA (dominant p = 0.183, accessory p = 0.515), SAED (dominant p = 0.110, accessory p = 0.497), and SAMA (dominant p = 0.240, accessory p = 0.633). The location of the NF did not vary between sides of the body in the SAA (dominant p = 0.842, accessory p = 0.379), SAED (dominant p = 0.576, accessory p = 0.093), and SAMA (dominant p = 0.367, accessory p = 0.249). The variable locations of the dominant NF were 63% on the AMS, 26.3% on the MB, 5% on the LB, 2.4% on the ALS, 1.9% on the PS, and 1.4% on the AB (Fig. 4). The location of the NF showed variations across populations (p = 0.000). The location of the NF on the LB was significantly higher in the SAED (p = 0.000), whereas the location of the NF on the MB was significantly lower in the SAMA (p = 0.007). Half (50%) of the accessory NF were located on the AMS, and their number was significantly higher in the SAED than in the SAA (p = 0.000) and SAMA (p = 0.000) population groups (Tables 2 and 3). The number of the accessory NF located on the AMS showed no significant differences between the SAA and SAMA (Table 3) population groups.

Fig. 4.

A photograph of the humerii showing the variations in the location of the dominant and accessory nutrient foramina. Hypodermic needles indicated viability of the nutrient foramen. (A) Illustrates a single dominant nutrient foramen located on the posterior surface of the left humerus. (B) Illustrates a single dominant nutrient foramen located on the anteromedial surface of the right humerus. (C) Shows two nutrient foramina: an accessory nutrient foramen (proximal) located on the anterior surface, and the dominant nutrient foramen (distal) located on the anteromedial surface.

The direction of the NF and its associated nutrient canal did not vary between sexes in the SAA (p = 0.293), SAED (p = 0.327), and SAMA (p = 0.731). The direction of the NF and its associated nutrient canal did not vary between sides of the body in the SAA (p = 0.316), SAED (p = 0.237), and SAMA (p = 0.116). The NF and nutrient canals were directed toward the distal ends in 99.8% of humerii (Fig. 1) and toward the proximal end in 0.2% (Table 4). The single proximally directed NF and nutrient canal was observed in an SAED male on the right humerus.

Table 4.

The direction of the number of the nutrient foramina on the humerus in the South African populations.

Population	SAA					SAED					SAMA
Sex	Male		Female		Total n (%)	Male		Female		Total n (%)	Male		Female		Total n (%)
Side	R	L	R	L		R	L	R	L		R	L	R	L
Direction
Distal	65	65	58	59	247 (100%)	53	56	47	48	204 (99.5%)	37	41	24	27	129 (100%)
Proximal	–	–	–	–	–	1	–	–	–	1 (0.5%)	–	–	–	–	–
Total sample	65	65	58	59	247 (100%)	54	56	47	48	205 (100%)	37	41	24	27	129 (100%)

–, Absent; SAA, South African Africans; SAED, South Africans of European descent; SAMA, South Africans of Mixed Ancestry; M, male; F, female; R, right; L, left.

Discussion

The current study examined the topography and morphometry of the nutrient foramina of the humerus in three South African populations. The results support previous findings that the humerus predominantly has a single dominant-sized NF^10,13,19, but cases of up to four NF are not surprising^19,20. Contrary to the current findings of 20.1% of bones with accessory NF, a previous study utilising undetermined populations reported that 7% of the humerii showed only secondary-sized (accessory) NF¹². Interestingly, the number of NF on the humerii showed significant population-specific variations between the studied population groups. However, previous studies on lower limb bones in the same population groups did not show any population specificity with reference to the number of NF^7,31,40. The presence of numerous NF on a single bone indicates abundant blood supply to the bone to support bone metabolism. Clinically, numerous NF offer alternative regions for harvesting vascularised osteo-periosteal flaps. A vascularised osteo-periosteal flap is a segment of the bone that is harvested together with the overlying periosteum and its underlying blood vessels. These flaps become crucial in orthopaedic and reconstructive surgeries where maintaining the blood supply to the bone is crucial for healing and repair^41,42.

Few (2.9%) humerii in the current study presented with no NF. The absence of NF on long bones has been widely reported^10,20,40 and in these instances the periosteal vessels supply blood to the diaphysis^8–10. The reported absence of NF could be due to the methodology used in the characterization of the NF. Nearly all the previous studies on the humerus used the 18–25 gauge (1.27–0.515 mm outer diameter) hypodermic needles to characterize the patency of the NF^{9,10,12,13,19,20}. Thus, NF smaller than this, for example, those in the range of 26–34 gauge (0.464–0.159 mm outer diameter), are likely to be missed and recorded as absent¹¹. Using a magnifying glass may lead to identifying smaller NF that may be difficult to locate without visual aids⁷.

During embryonic development, the growth of long bones like the humerus follows the ‘growing-end theory’ that one end of the bone grows faster than the other, and the nutrient canal (that carries nutrient vessels) is always directed away from the growing end to avoid the nutrient artery being pulled by the faster-growing end^43,44. This then determines the direction of the NF post embryonic development. In the current study, the majority of the NF and their associated canals were directed distally (towards the elbow) away from the growing proximal end, similar to previous study in Indian populations²⁰. In the bones where the NF does not obey the ‘growing-end theory’, it is suggested that the bone may have undergone a unique ossification during maturation in which the opposite end (for example, distal humerus) may have grown faster than the typical proximal humerus²⁰.

The position of the NF helps determine the parts of the bone that may be affected by non-union of fractures and delayed healing due to avascular necrosis^12,45. The position of the NF on the humerus in the current study falls within the previously established humerus FI ranges of 15–83%^{10,11,13,19,20}. The mean FIs of the humerus in the current study were in the distal half of the middle third of the diaphysis as follows: 57.36% in the SAA, 58.83% in the SAMA and 55.95% in the SAED. Similar findings were observed in the Indian (57.60%)¹⁰, Southern Brazilian (55.20%)¹¹, and Spanish (57.70%)¹³ populations. Thus, transverse fractures in this region of the humerus may rupture the nutrient artery and result in delayed healing due to decreased bone vascularisation, a complication that can prevent the union of fracture segments¹³. The position of the NF showed significant variations within and across South African populations in the present study. Males had significantly higher mean FI than females in the SAED population only. Across the populations, the SAED had a significantly lower mean FI than the SAMA. In the humerus, non-union fractures are common in the distal half of the middle third (FI 49–66%), suggesting that in individuals whereby the FI is low (due to NF occupying the proximal half of the middle third of the shaft), fractures that occur distal and furthest from the NF may be difficult to heal due to inadequate blood supply. Although periosteal arteries supplement nutrient arteries in supplying blood to diaphysis, their effect is low compared to the latter, leading to delayed union of fractured bones. The majority of fractures on the humerus occur in the middle third (midshaft) of the diaphysis and often are transverse in nature⁴⁶. Midshaft fractures are associated with delayed union and or non-union due to various factors including disruption of the blood supply and poor immobilisation due to antagonistic actions of the flexors and extensors and also between the proximally attached muscles versus distally attached muscles^14,46,47. Disruption of the arterial blood supply to the shaft during transverse midshaft fractures is common⁴⁶, this is because the humeral nutrient artery in the majority of cases arises from the brachial artery and then enters the humeral shaft through the nutrient foramen located on the anteromedial surface of the midshaft humeral diaphysis^6,18.

Knowledge of the location of the NF on the surface of the diaphysis aids in determining the safe area for surgical procedures to avoid iatrogenic arterial injuries and to determine safe zones for the harvesting of viable vascularised osteo-periosteal flaps for bone grafting^7,41,45,48. Notably, the lateral (LS) and anterior (AS) surfaces of the distal humerus are preferred areas for harvesting vascularised osteo-periosteal flaps since they receive arterial blood supply from the radial and middle collateral branches of the profunda brachii artery⁴². In the current study, the anteromedial surface (AMS) of the humerus housed the majority of the dominant NF. Similarly, various previous studies reported the AMS of the humerus as the most prevalent location of the NF, with frequencies ranging between 60 and 89.7%^6,10–12,19. According to Campos et al.¹³, in situations where the humeral diaphysis had two NF, the dominant-sized NF were found on the AMS while the accessory NF were located on posterior surface (PS) or anterolateral surface (ALS) only. However, in the current study both the dominant and accessory NF occupied all the humeral surfaces but with variable frequencies, similar to other reports^20,49.

Unlike previous studies on the NF of the lower limb bones that reported no population differences between the three major SA populations^7,31,40, the current study found significant population-specific variations regarding the location of the NF on the humeral diaphysis. The location of the dominant NF on the AMS was associated more with the SAMA populations. The accessory NF, on the other hand, were mainly on the AMS more so in the SAED than the SAA and SAMA. Location of NF on the AMS which is the flexor surface of the humerus, may have functional connotations indicating that the diaphyseal surface subjected to high work demands receives more blood supply⁵⁰. Various factors have been attributed to differences in the skeletal biology in various populations, including changes in secular trends as a result of changes in nutrition and genetic constitution, geographic area, extreme labour environments, changes in socioeconomic status, and the advent of technology^51,52.

Clinically, population-specific variations necessitate preoperative angiograms to detect any variant topographical presentations that may complicate surgical procedures. In addition, precise location of the NF and the associated nutrient artery is vital during open reduction/internal fixation of humeral fractures in order to guide the placement of screws thereby sparing the vascular structures¹⁴. The common surgical approach during open reduction of fractures of the proximal humeral shaft utilises the anterolateral plane through the deltopectoral interval and the brachialis muscle split^6,14,46,47 while distal humeral fractures are surgically managed through the posterior approach. This anterolateral approach avoids the anteromedial surface of the humerus thereby sparing the humeral nutrient artery, but the posterior approach may disturb accessory nutrient arteries, branches of the profunda brachii which often enter the distal humeral shaft on the posterior surface^6,14.

Conclusion

The current study has shown topographical and morphometric variations of the NF of the humerus across the South African population groups. So far, this is the largest sample-sized study involving three population groups from the same geographical area indicating population-specific variations of the NF parameters on the humerus. The topographic and morphometric information of the NF on the humerus obtained herein may improve the management of humeral fractures and also aid in the safe harvesting of osteo-periosteal flaps during bone grafting.

Author contributions

PM made substantial contributions to the conception or design of the work. PM, BB, ATM performed the acquisition, analysis, or interpretation of data; PM, AM, PM, BB drafted the work or revised it critically for important intellectual content; PM, PM, ATM, BB approved the version to be published; andagree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Data availability

The datasets generated and/or analysed during the current study are not publicly available due to institutional ethics board restrictions but are available from the corresponding author on reasonable request.

Declarations

Ethical approval

The study was conducted under the ethical clearance waiver W-CBP-220504-01, obtained from the University of the Witwatersrand Ethics Committee. The study was undertaken as per the requirements of the Human Tissue Act (No. 65 of 1983) and the National Health Act (No. 61 of 2003) on using human specimens for research and teaching purposes.

Competing interests

The authors declare no competing interests.