Obesity is a global health problem in humans and pets. In 2022, the U.S. Pet Obesity Prevalence Survey found that 59% of dogs were classified as overweight or obese by their veterinarians (1). There is a growing population of overweight and obese bitches being bred (2), especially considering show standards promote an over-conditioned body type in some breeds such as Molossoids, Swiss Mountain and Cattle Dogs, Labrador Retrievers, Water Dogs, Pugs, and Basset Hounds (3, 4). There are numerous studies that have identified adverse health effects of obesity on lifetime health and longevity in dogs (5, 6), such as increasing the risk of insulin resistance (7), altered cardiac structure and function (8), and orthopedic disorders (9). Despite this body of literature, there are still no clearly documented studies on the effects of increased adiposity on the reproductive performance of female dogs, including milk composition in nursing dams and the resulting neonatal health outcomes.

Lactation is the final pathway for the maternal metabolism to influence the neonate, with studies supporting obesity having a negative impact on lactation (10, 11). Overweight and obese women are less likely to initiate lactation, have a delayed onset of lactogenesis II (copious milk production), and have a generally shorter duration of breastfeeding (12, 13), although the exact reason is poorly understood. Overweight and obese mothers have also reported that their infants suckled for longer durations, and they had reduced prolactin responses to suckling at the first 48 h of lactation (14). In sows, obese dams had a higher occurrence of postpartum dysgalactia syndrome, hypothesized to be from an increased catabolic rate from higher maternal fat deposits (15). In dogs, obesity was mentioned as a potential contributor to increased neonatal puppy mortality, in part due to hypogalactia (16, 17); however, there has not been a controlled study to provide clear evidence. The reproductive consequences of canine maternal overweight and obesity, including milk production and composition and neonatal health, are largely unknown (18).

Canine milk is typically higher in protein and fat than human, bovine, or caprine milk (19–21). Milk composition changes in healthy bitches throughout lactation, with some disagreements between the studies. Over the first month of lactation, the total protein of canine milk was shown to either decline (22) or remain unchanged (19), total lipids were unchanged (19, 22), and there was an increase in lactose through the first week (23) and month of lactation (19, 22). Gross energy either increased in the first week (23) or remained unchanged throughout the first month of lactation (19, 22). Additionally, one study found subtle differences in milk composition across dog breeds of similar sizes (21), while another study found no difference between breeds of different sizes (24).

Maternal adiposity has been shown to affect milk composition, although the results are inconsistent. In women, several studies have found that breast milk of obese mothers had higher fat (25–27), calories (27), and protein content (25). In contrast, other studies have described no difference in the overall breast milk fat content (28, 29) despite significant differences in the long-chain fatty acid profiles (28–31), such as increased n-6/n-3 polyunsaturated fatty acids and decreased docosahexaenoic acid (28–31). These changes in altered fatty acid profiles were subsequently shown to be associated with increased infant body mass index (BMI) and weight at birth (28, 29, 31), 2 months (31), 6 months (30), and up to 13 months (29), although these associations were found to generally disappear as children aged (28, 30, 31). Additionally, breast milk of obese mothers also had significantly higher levels of triglycerides (27), leptin (32, 33), and insulin (27, 28, 32, 33), which affected the neonatal intestinal microbiome without having a significant impact on infant weight gain or body fat percentage at 2 weeks of age (33). Despite the large breadth of research in this field, a systematic review concluded that the majority of human studies were too variable and many were of low quality (26).

The impact of maternal adiposity on offspring growth has also been studied in other species. The litter average pre-weaning growth rate was higher in over-conditioned sows despite lower predicted milk yield in the second week of lactation than in under-conditioned sows (15). In mice, offspring from obese dams were hyperphagic from 4 to 6 weeks of age and eventually had increased abdominal fat mass at 3 months and body weight at 6 months compared to control offspring (34). To establish the effect of lactation alone, other murine studies used cross-fostering, i.e., pups born to lean mothers were nursed on obese dams. These fostered pups had increased caloric intake and ended up having higher body weight, body fat, and plasma insulin and leptin concentrations (35, 36). However, one study has only found a sex-dependent increase in perirenal fat (37). Kittens of overweight queens grew faster after weaning and were heavier at 5–6 months of age than kittens from lean dams or from dams whose non-pregnant body condition fluctuated between overweight and lean (38). While normal growth curves have been established for the average neonatal puppy (39) and between different breeds and sizes (40–43), it is yet unknown how maternal adiposity in dogs affects milk composition and early puppy growth.

This study aimed to compare the milk macronutrient composition and puppy growth rate during the neonatal period between overweight and lean bitches. We hypothesized that maternal over-conditioning will result in higher milk fat content and increased milk gross energy, and, in turn, these changes may lead to increased body weight gain in the puppies during the exclusive nursing period (from parturition to day 21 postpartum) compared to lean bitches.

All experimental procedures were approved by the Institutional Animal Care and Use Committee of Virginia Tech (Protocol #22-032). Informed consent was obtained from owners of all participating dogs.

Three bitches required cesarean sections due to dystocia. One bitch (#8) was included for two separate litters; she was lean during her first lactation and overweight 10 months later during her second lactation (Table 1).

Across the 16 litters in our study, 116 puppies were born alive (n = 68 from LE dams; n = 48 from OW dams), and 7 were stillborn (n = 2 from LE dams; n = 4 from OW dams). After eliminating bitches #3 and #9 from the puppy growth analysis, 106 puppies remained (n = 58 from LE dams; n = 48 from OW dams). Three puppies from different litters died within the first week.

This study is the first to examine the effect of increased canine maternal body condition on the neonatal puppy growth rate and milk macronutrient composition, despite its small sample size. Understanding the impact of over-conditioning of the dam during lactation while she is nursing her puppies and providing evidence-based results is essential to guide weight management recommendations for breeding broodbitches.

The dam’s BCS had no effect on puppy birth weight in our study population, although breeds were only partially matched across the maternal BCS groups. This finding is in agreement with feline (38), murine (34), swine (15), and one rat study (54), although the literature is not always consistent. Birth weights of rat pups born from obese dams were lower compared to normal-weight dams (37), but there was no difference in piglet litter weight at birth in over-conditioned sows (15). While the majority of human studies found no difference in infant birth weights from overweight or obese mothers (28, 30, 33, 55), several studies reported significantly increased weight-for-length z-score (28) or weight-for-age z-score (29, 31, 56).

Puppy growth curves (day-to-day body weights) were slightly but significantly affected by the dam BCS group during the first 3 weeks after birth. The literature is also divided and inconsistent about how maternal adiposity affects neonatal weights at different timepoints. Some studies comparing maternal adiposity levels have shown no weight differences in human infants after 2 weeks of age (28, 33), in female mouse pups before 3 months of age (34), or in piglet litter weight at weaning (15). However, a study in women has reported infant BMI and weight differences up to 2 months (31), and one cross-fostering study in rats has found lower birth weight, week 1 body weight, and week 2 body weight in pups nursing on obese dams (37). In contrast, another cross-fostering rat study has found pups from obese rats weighed more at 3 and 24 weeks of age with an increase in percent body fat (54). These cross-fostering studies have demonstrated that maternal adiposity exposure during lactation alone can be a significant determinant of offspring body weight and adiposity. The conflicting conclusions of maternal adiposity’s effect on pre-weaning neonatal weight gain between different species likely reflect many variables involved in growth, as well as differences between species, in experimental design, and maternal diet during lactation. The effects of maternal adiposity are also not limited to the early neonatal period. At 6 months of age, obese mouse dam pups had increased body weights, when hyperphagia was also observed, suggesting that the consequences of maternal adiposity may not emerge until post-weaning (34). When offspring were exposed to obese dams during lactation alone, they also had an increase in percent body fat at 6 months of age (54). We were unable to evaluate post-weaning puppy growth rate between the maternal groups in this study, although it would have been interesting to follow the puppies until adulthood.

Although dogs from more than one breed were included in this study, we selected participants for similar adult body weights. The percentage-based measurements (ADG%, ADGB%, and TGB%) were used to mitigate potential differences in puppy growth curves between the breeds. Nowadays, specific puppy growth charts are available for many dog breeds and for different adult weight classes (42, 43). Clinical practice guidelines also use a generally expected 5–10% daily weight gain for normal neonatal puppy growth rates in the first 3–4 weeks of life, independent of breed (39, 43). Our study has found that ADGg, ADG%, ADGB%, and TGB% were significantly affected by the interaction between the dam BCS and time, with OW dam puppies showing accelerated growth rates in the first 4 days of life. This finding was different from a study in kittens, which showed that kittens from OW queens did not have a significantly greater monthly weight until 4 months of age. However, that study followed kittens until 12 months of age and did not closely examine the changes during the neonatal period (38). A study in pigs has also found that piglets from higher BCS sows had a significantly higher mean daily body weight gain, although these litters also had significantly lower litter sizes compared to lower BCS sows, as piglet litter weights were not different at weaning (15). While our initial hypothesis that puppies from OW litters would be consistently heavier and grow faster was disproven, we did find that puppies from OW dams grew differently, especially during the first few days after birth.

We speculate that either our OW population’s adiposity was not significant enough to exert a more substantial effect, which may only become consistent at an obese maternal body condition (BCS 8–9/9), or the impact of an overweight condition would not be observable until later in a puppy’s life. Such long-term effects may not be detectable in such a small cohort of dogs and masked by many variables that affect body weight and body condition.

In our study, the maternal BCS had no effect on average daily percent gain at the litter level (i.e., LiADG% and LiADGB%) or total percent gain from birth (i.e., LiTGB%), most likely because of the lower statistical degrees of freedom from comparing litter averages instead of individual puppies.

Litter size did not have a significant effect on birth weight in this study. This is in contrast to previous canine literature showing that a larger litter size results in significantly lower individual birth weights (40, 57, 58), although one study found this significance only in medium- and larger-sized breeds (40). The most likely cause for this discrepancy may lie in the small sample size of our study and its limited range of litter sizes (range: 5–11). Litter size also had no effect on any individual puppy growth parameters; however, a significant inverse relationship with LiTGB% showed a 21.5% less total litter weight gain from birth with each additional puppy. Other canine studies have also found an inverse relationship between litter size and growth over time (40, 57, 58).

Birth weight had a positive relationship with puppy growth curves, indicating that puppies born heavier remained heavier overall. This finding is consistent with other studies evaluating puppy growth within the first (23, 40) and second week of life (39, 40), although birth weight was no longer a significant factor at week 3 after birth (40). In contrast, birth weight had no effect on ADGg as a representation of daily body mass gain (in grams), while there was an inverse relationship with ADG% and ADGB%, indicating that the percent daily weight change compared to the previous day or to birth weight was slightly but significantly less in puppies born heavier. This was also true at the litter level with increasing litter average puppy birth weight (LiADG%, LiADGB%, and LiTGB%). Our finding is similar to a human study that found large-for-gestational-age infants grew slowly enough to reach similar body mass to their appropriate-for-gestational-age-sized infant counterparts (59). Another study tracking the puppy growth rate in the first week of life has also found that puppies with lower than normal birth weight had higher growth rates after 2 days of age, which was attributed to compensatory growth (60).

We observed 8 of 116 live-born puppies develop mild swimmer puppy syndrome, all from OW dams and from 3 of the smallest litters with 3, 5, and 6 puppies. We suspect that incidents of swimmer syndrome, in spite of an adequately floored environment, may be related to excessive early weight gain before ambulation, which was seen on the whole study population level in the higher ADG% of OW dam puppies in the first week after birth compared to LE dam puppies despite no difference in ADGg. Interestingly, not all puppies in the same litter (except for litter #9 with a total of 3 puppies) were affected. This highlights that other factors, such as conformation and activity level of the individual puppies, may have played a role. Indeed, factors responsible for the development of swimmer syndrome in puppies have been studied before by Tomihari et al., who found that affected puppies in a guide dog colony were from significantly smaller litter sizes and had significantly higher body weights on days 10 and 28 after birth (61). Small litter size was also a significant risk factor found in a questionnaire study on 115 dogs (62) and in another report on 2,443 puppies (63). While some breed dispositions have been suspected, none have shown significance (63). These findings are unsurprising, as puppies from smaller litters are able to consume more milk with less competition. Further studies in a larger population of overweight/obese lactating bitches are needed to confirm our finding of maternal adiposity as a risk factor for swimmer puppy syndrome.

Milk macronutrients were not significantly affected by the dam BCS group over the first 4 weeks postpartum. Our current results suggest that canine milk macronutrient composition is well regulated and not influenced by maternal body composition. This is not the case in all species. Current human literature has found that breast milk from obese mothers contains either increased milk fat (25, 26) or calories (25). In rhesus macaques, one study has found that milk bioactive factors [e.g., epidermal growth factor (EGF) and transforming growth factor beta 2 (TGF-β2)] were significantly affected by maternal body mass (64). This study has also found correlations between these bioactive factors and milk energy through milk fat % and milk sugar %, though milk protein % was correlated with EGF only (64). It is possible that other milk components that were previously reported to be affected by maternal body composition in other species, such as milk fatty acids (28–31), metabolites (e.g., oligosaccharides and nucleotides) (55), and insulin or leptin (28, 33), could have affected puppy growth. Another explanation is that canine milk macronutrients are not affected by the dam’s BCS until they are obese (8–9/9 BCS), and our OW population was not over-conditioned enough to elicit a clear change. The majority of studies looking at other metabolically relevant milk components in an obese maternal group may give further insights. We were unable to determine milk production and the amount of milk consumed by puppies between OW and LE bitches, which could explain the differences in growth pattern. In clinical practice, puppy weight gain is considered the best estimate for maternal milk production, although this assessment encompasses the amount of milk consumed (not produced) along with all other milk components beyond macronutrients. In a research setting, turnover of body water of the puppies estimated from water kinetics by deuterium oxide administration can be used to estimate the volume of ingested milk and extrapolate it to quantify maternal milk production (19). However, this is impractical and impossible to perform on client-owned bitches.

Litter size did not have a significant impact on any milk components, which is contrary to the inverse relationship found between litter size and milk GE during the first week of lactation by Keller et al. (23). This difference may stem from the aforementioned study’s shorter sampling window or the current study’s limited range of litter sizes.

Milk DM significantly increased by week 4 after parturition. This pattern reflected the changes in milk fat and CP and was not seen in the only currently published longitudinal study’s reported values (19). Milk sugar also significantly increased over the first 4 weeks postpartum. This pattern is similar to that seen by Keller et al. in the first week postpartum (23) and is also in agreement with other canine studies that followed bitches until at least 4 weeks after birth (19, 22).

The changes in milk GE over time rightfully reflect that GE is calculated from milk CP, fat, and sugar, as it significantly increased over the first 4 weeks postpartum. This pattern is similar to the study of Keller et al. in the first week postpartum (23). In contrast to our findings, current canine longitudinal studies showed no difference in milk GE over the course of lactation (19, 22). This difference may be the result of dog breeds used (21) or a subtle contribution from the overweight bitches that only followed trends. On an energy basis, sugar’s contribution to GE decreased slightly toward the end with a corresponding slight increase in energy from fat. The contribution of crude protein to milk GE was relatively stable over time.

The numerical values of the different milk macronutrients, DM and GE, were either comparable or somewhat different from those published for canine milk, even when compared with samples collected at similar timepoints in lactation (19, 23). These discrepancies may be explained by differences in methodology, breeds, or husbandry (including nutrition) between studies. Furthermore, although we collected a homogenized milk sample from the available front and hind teats, we were unable to collect samples to standardize fore and hind milk within each gland or to standardize the timing of collection in relation to the puppies’ most recent nursing.

The main limitations of this study stem from using client-owned animals; however, this variety indeed represents a real-world population to which recommendations can be immediately applied. With our variety of six different dog breeds, dam weights ranged from 20.5 to 40.5 kg. A study by Zhang et al. has found that there can be breed-related differences in canine milk composition (21); however, their results are difficult to compare with our data since their sampling window was wide and many of these milk macronutrients do change throughout lactation. Puppy birth weights also ranged between 167.3 g and 561.3 g, and hence, we used several percentage-based alternative parameters (ADG%, ADGB%, and TGB%) to monitor and compare puppy weight gain. Growth curve equations have been developed to monitor canine neonatal puppies in the past, impacted significantly by breed, sex, or birth weight alone (39), but this single growth equation lost accuracy after day 12 when the growth of the puppies slowed during the neonatal period. Total milk production and consumption are clear unmeasured contributors to puppy growth; however, this parameter was not feasible to collect in our client-owned bitches due to a lack of standardization, variable teat conformation and thus feasibility to completely milk out the gland, and discouragement of unnecessary medications to stimulate milk let-down.

As with studies involving client-owned dogs, there are several variables beyond the control of the investigators, including but not limited to the owner’s home environment and the time owners are able to monitor puppies and ensure the dam is nursing them. For the health of the litters, minor interventions were advised to the owners. While puppies requiring milk supplementation were excluded from the data set, some puppies falling behind were advised to have more nursing time or access to fuller teats. The five diagnosed swimmer puppies from two OW litters were intervened on primarily by physical therapy and limiting total daytime nursing times. Litter #9 was excluded from the data analysis due to the intervention of significant nursing limitations and outlier litter size. While these minor interventions may have affected our results, owners and breeders would do these easy adjustments within their usual husbandry practices. Analyzed litter sizes ranged from 5 to 11 puppies, not lending to a full range of litter sizes; however, studying the effect of litter size was not the primary goal of the study. Having a more uniform litter size made our study population more homogenous and allowed for a more accurate determination of the impact of maternal body condition. As discussed previously, future studies would benefit from a larger sample size, a consistent diet, maternal groups with more distinct adiposity levels (i.e., lean vs. obese [BCS 8–9/9]), and the possibility of analyzing other milk biomarkers that have been shown to be affected by maternal adiposity in other species.

Maternal overweight condition affected the growth curve of the puppies subtly but significantly. Puppies of over-conditioned dams gained significantly more during the first few days after birth. Birth weight was also a significant determinant; heavier-born puppies remained heavier overall while their percent daily gain was slightly less. Litter size did not affect individual puppy growth. Milk macronutrient composition did not differ between lean and overweight bitches, so milk consumption rather than milk macronutrient composition could have been the primary determinant of puppy and litter growth differences between the maternal groups. The role of other milk components not examined in the study cannot be ruled out. The small study sample size may have been a limiting factor in elucidating differences in milk macronutrient composition between overweight and lean bitches. Alternatively, differences in milk composition may not be seen until a truly obese maternal body condition (BCS 8–9/9) has been reached. Swimmer syndrome was diagnosed in a few puppies of overweight dams only, highlighting an important consequence of overweight maternal condition combined with small litter size on neonatal puppy health.