This doesn't say anything about which providers are the majority in Europe, which is what OP's study is about. Your findings (non-US companies are "more likely") and OPs findings (US companies are the majority) are both compatible.