In our review of twenty-three studies comprising 18,543 girls and 3,310 boys, we found significant associations of anthropometric status and body composition, with pubertal development in most of the studies.

In the current review, studies reported inconsistencies in the relationship between nutritional status and pubertal development both within and across different countries. For example, in a study in Senegal, no differences were observed in breast development between adolescent girls stunted and not stunted at infancy [19]. However, achievement of breast development in girls was associated with lower likelihood of stunting in Kenya [13], and higher HAZ and BMI in South Africa [30]. Moreover, while a study in Ethiopia reported a negative association between menarche age and BMI [25], a study in Ghana reported a positive association between these variables [24], and there was no association between menarche age and BMI in Uganda [27]. These differences may be attributed to ethnic or racial differences which are unclear in the various studies. It is speculated that ethnicity is a determining factor in physique characteristics and fat patterns [27]. Also, epigenetics may be considered as a role-player in these discrepancies, changing the trend of pubertal development in same and different countries, with changing nutrition and related lifestyle behaviours.

In Zambia, rural boys showed later onset, slower progression and later completion of testicular growth compared to their counterparts in urban settings [20]. The investigators observed that, completion of testicular growth in the slowly growing rural population of adolescents in Zambia, was linked to greater triceps skinfold suggesting that testicular growth is associated with energetic availability even under poor nutritional conditions. The researchers attributed this to the poorer somatic growth of rural adolescents, which reflects in their height and upper arm muscle. They concluded that variation in pubertal onset, progression and maturation in boys is related to energetic status despite genetic factors.

Previous reviews that investigated nutritional status and pubertal development have not included many studies in SSA. In a recent review of 27 studies globally [8], six (6) studies were in SSA. Two (2) of these studies did not report associations of nutritional status with pubertal development [38, 39]. One (1) study compared pubertal development of HIV and non-HIV infected girls. Therefore, only three (3) studies [27, 28, 34] from the previous review that reported associations of pubertal development with nutritional status overlap with the current review of anthropometric status, body composition and pubertal development in SSA.

The previous review included 27 studies that described healthy adolescents born in and after 1998, with the latest publication in 2018 [8]. Researchers identified studies that described age at menarche (AAM) and/or Tanner Stages 2 to 5. There was a similar trend of advanced pubertal development with higher BMI. The meta-analysis showed earlier onset of menarche, B2 and gonadal development stage 2 by Tanner Staging (G2) for overweight compared to thin and normal children, classified by BMI [8]. Likewise, adolescents with normal BMI from the African region experienced later age at menarche compared with adolescents from other WHO regions (Eastern Mediterranean Region, Region of the Americas, South-East Asia Region, Western Pacific Region, and the European Region) [8]. The reviewers suggested that AAM and puberty vary among races and countries. Moreover, the meta-analysis suggested a potential association between overweight/obesity and earlier onset of female puberty, although the relationship was not significant, unlike twenty-one (21) studies in the current review (Table 1), which found significant associations.

In the current review, although the sample in Uganda found no significant association between menarche age and BMI or WC, HC was inversely correlated with menarche age [27]. Likewise, cross-sectional data from the third NHANES III showed that among 10–14 years old females, unit increases in HC is associated with 24% odds of attaining menarche; nevertheless, increases in WC and triceps are associated with 7% and 9% lower odds of menarche, respectively [27]. It is suggested that even though HC is correlated with BMI, HC supports protective metabolic functions.

It is important to note that each of the 23 studies investigated some pubertal milestones. Therefore, comparing pubertal development across the 9 countries was short of complete and accurate comparison for pubertal milestones of boys and girls. For example, only one study investigated the association of HAZ and BMIZ at 5 years, with pubic hair development in adolescents in South Africa. Also, most studies assessed breast development (8 studies) and menstruation (17 studies) for females, while 4 studies assessed genital development for males. There was a relatively smaller proportion of boys, compared to girls, assessed for pubertal development. This reduces the extent of investigations of pubertal development in boys as most studies do not assess all pubertal milestones. Furthermore, there were differences in pubertal assessment methodologies like self (pubertal development scale) and clinical (Tanner staging) assessment of various pubertal milestones, across the studies which challenged comparison across studies.

Pubertal development is stimulated by synergistic effects of increased secretion of gonadal sex steroids, growth hormones (GH) and insulin-like growth factor 1 (IGF-1) [40]. Although the mechanism for metabolic control of puberty is unclear, recent findings have shown relevance of novel neurohormonal and molecular mechanisms that include key cellular energy sensors [41]. Leptin and estrogen levels produced in fat cells have been suggested to play a vital role in stimulation of pubertal development [30]. At the skeletal level, increased IGF-1, with increased insulin, lean body mass/muscle mass/fat mass stimulate chondrocyte maturation and proliferation, and bone mineral accretion increasing bone maturation [42]. This process stimulates pubertal growth spurt. Between the ages of 6 and 8 y, rises in adrenal hormones (like cortisol and dehydroepiandrosterone) and gonadal hormones (like testosterone, estradiol and progesterone) stimulate observable physical changes like body hair [43].

In a state of increased weight, there is increased levels of leptin, insulin, IGF-1 [42] and cortisol [44]. Increased insulin and IGF-1 levels promote increased lean body mass, muscle mass and fat mass, stimulating growth spurt. Increased Leptin and IGF-1 contribute to skeletal development by stimulating chondrocyte maturation and proliferation, bone mineral accretion and increased bone maturation. These developments ultimately lead to the achievement of pubertal growth spurt. Increased Leptin stimulates the hypothalamus to generate luteinizing hormone releasing hormone, the pituitary to secrete gonadotropin, adrenals to secrete adrenal steroids and gonads to secrete sex steroids [42]. The secretion of gonadotropin and sex steroids expedite pubertal development [42]. The sex steroids and skeletal maturation contribute to growth spurt [42]. A serine/threonine protein kinase, mTOR, activated in energy excess, contributes to mediation in the leptin effects of expediting puberty through modulation of the hypothalamic expression of kisspeptin 1 gene, which acts as regulator of reproductive functions [41]. Available evidence suggests that there is significant increase in dehydroepiandrosterone at the time when a child’s BMI is highest compared to when it was lower [42]. This implies that increased body fat may significantly contribute to the activation of adrenal androgen production and the onset of adrenarche.

Moreover, the gluteofemoral fat (lower-body fat) has been inversely associated with AAM, blood glucose, blood pressure and lipid levels [27]. It has also been found to be associated with reduced risk of cardiovascular diseases, all-cause and cardiovascular mortality [27]. While leptin and adiponectin levels have been positively associated with gluteofemoral fat, inflammatory cytokine levels have been inversely linked [45]. It is speculated that gluteo-femoral fat traps excess fatty acids and prevents chronic exposure to elevated lipid levels [27].

On the other hand, in a state of malnutrition, decreased BMI, and reduced fat mass, there are increased levels of cortisol and GH, but reduced levels of leptin, insulin and IGF-1 [42]. Increased cortisol, GH and decreased insulin are associated with decreased lean body mass, muscle mass and fat mass, delaying growth spurt [42]. Moreover, decreased leptin and IGF-1 delay chondrocyte maturation and proliferation, decrease bone mineral accretion and delay bone maturation delaying growth spurt [42]. Consequently, in the hypothalamus, there is inhibition of luteinizing hormone releasing hormone generation and delay in gonadotropin secretion in the pituitary, as a result of decreased leptin levels, delaying pubertal development [42]. In the adrenals and gonads, there is delay in sex steroid secretion delaying the pubertal growth spurt and pubertal development. Leptin deficiency can lead to absence of puberty [42]. Adenosine monophosphate activated protein kinase, a heterotrimeric serine/threonine kinase, interplays with kisspeptin neurons in the metabolic control of puberty, mediating the repressive effects of nutritional deficiency. Sirtuin 1, an energy sensor, represses kisspeptin 1 expression and delays puberty during energy restriction through the epigenetic regulation of arcuate nucleus kisspeptin 1 neurons.

The 6 longitudinal cohort studies allowed for assessment of precise pubertal timing and development. Nevertheless, the cross-sectional nature of 17 studies limited the capture of precise timing of pubertal milestones. Consequently, some reported pubertal timings may have been overestimated or underestimated due to the time of measurement and the possibility of incorrect recall. Specifically, adolescents tend to overestimate as much as underestimate pubertal development [46].

While the NOS provided a structured approach to evaluating the quality of studies included in the review, potential biases in these studies may have influenced our findings. Such biases include those related to subject selection, outcome measurement, and inadequate control for confounders, which could affect the observed associations between nutritional status and pubertal milestones. For example, failure to control for socioeconomic factors may confound results, making it difficult to isolate the specific effects of nutrition. Additionally, publication bias could have skewed our findings toward significant associations. Given these considerations, we have interpreted our results cautiously and recommend that future studies use rigorous designs to minimize bias.

It is noteworthy that the variability in data across studies, particularly due to differences in methodology (e.g., self-report versus clinical assessment), may affect the generalizability of our findings. Self-reported measures could introduce recall bias, while clinical assessments may vary depending on the resources and training available. Similarly, differences in participant characteristics across studies (e.g., age ranges, socio-economic backgrounds, and nutritional contexts) could limit the applicability of our findings to broader populations in SSA. Therefore, our findings should be interpreted with caution.

The review may be biased from the varying quality of the studies in SSA. This distribution suggests potential bias in the review due to the inclusion of lower-quality studies, particularly cross-sectional studies, which are more prone to confounding and other biases. Consequently, the presence of unsatisfactory and high risk of bias studies could weaken the strength of the summarized evidence. Notably, the review assessed associations between different anthropometric and body composition indices versus assessing puberty by different methods across diverse populations. Therefore, the findings may be generalizable to populations in SSA with similar sample characteristics.