There currently exists no evidence in the literature to suggest that the use of weightlifting belts during training improves long-term unbelted performance in any way.
Given the prevalence of their use, this finding should be surprising. The premise behind the behaviour, of course, is the assumption that belts support the erector spinae in limiting spinal flexion, thereby mitigating injury. However, this notion is invalidated by the observation that activation of the erector spinae is greater during belted peformance: weightlifting belts force the spinal extensors to work harder—not less. And whilst this might ostensibly appear positive, the strength of the erector spinae is not the only limiting factor in spinal strength. Furthermore, a study of 245 elite powerlifters revealed greater rates of injury to the lumbar spine amongst athletes wearing belts, as compared with those who did not, suggesting that lifters likely overestimate the support that belts provide. Indeed, since trained weightlifters and powerlifters can lift, on average, only around 5% more whilst wearing a belt, and since belts make no difference to the lifting strength of untrained subjects, it is not entirely clear whether weightlifting belts offer any inherent advantage whatsoever, or whether the perceived advantage they afford is the consequence of athletes' being trained to rely on them. It is entirely possible, even probable, that biomechanical changes that occur when a belt is worn force a pattern of motor recruitment, and hence a pattern of distinct muscular development, that is optimised for belted lifting.
Without further research comparing the performances of athletes trained entirely with and without weightlifting belts, or examining the hypertrophy of the core muscles—transversus abdominus, multifidus, diaphragm, and pelvic floor—as a consequence of both training states, any conclusions we might draw necessarily amount to conjecture. However, what is clear is that (1) belts are unnecessary for elite-level strength development and performance; (2) maximal lifts can be performed safely without a belt; (3) activation of the transversus abdominis and internal oblique is positively and significantly associated only with unbelted lifting; (4) belts result in higher rates of spinal injury amongst athletes who are most practiced in their use; (5) belts have little or no effect on power, velocity, or range of motion; and (46) belts offer little or no advantage to untrained individuals.
In view of the evidence, there is currently no scientific justification for the use of weightlifting belts, other than for competition efforts by athletes who have routinely used them in training. It is probable that belts offer no advantage to strength development, and may even hinder it.